LIST OF CONTRIBUTORS

Solomon Appel Robert H. Ashton Reza Barkhi

Metropolitan College of New York, New York, NY, USA Fuqua School of Business, Duke University, Durham, NC, USA Pamplin College of Business, Virginia Polytechnic Institute and State University, Blacksburg, VA, USA School of Management, University of Michigan-Dearborn, MI, USA College of Business Administration, San Diego State University, San Diego, CA, USA Department of Accounting, University of Arkansas at Little Rock, AR, USA Zicklin School of Business, CUNY – Baruch College, New York, NY, USA Belk College of Business, University of North Carolina at Charlotte, NC, USA College of Business and Economics, West Virginia University, Morgantown, WV, USA RSM Erasmus University, Department of Financial Management, Rotterdam, The Netherlands

Mohamed E. Bayou Chee W. Chow

Cynthia M. Daily Harry Z. Davis Nabil Elias Arron Scott Fleming

Frank G. H. Hartmann

vii

viii

LIST OF CONTRIBUTORS

Fred A. Jacobs Frances Kennedy James M. Kohlmeyer, III Leslie Kren John Y. Lee Michael S. Luehlfing Adam S. Maiga

School of Accountancy, Georgia State University, Atlanta, GA, USA Department of Accountancy and Legal Studies, Clemson University, SC, USA College of Business, East Carolina University, Greenville, NC, USA School of Business, University of Wisconsin, Milwaukee, WI, USA Lubin School of Business, Pace University, Pleasantville, NY, USA School of Professional Accountancy, Louisiana Tech University, LA, USA School of Accounting, Florida International University, Miami, FL, USA I.H. Asper School of Business, University of Manitoba, Winnipeg, Canada School of Professional Accountancy, Louisiana Tech University, Ruston, LA, USA School of Business, Wayne State University, Detroit, MI, USA Department of Accountancy and Legal Studies, Clemson University, SC, USA College of Commerce, National Chengchi University, Taipei, Taiwan

William W. Notz Thomas J. Phillips, Jr.

Alan Reinstein Lydia Schleifer Anne Wu

EDITORIAL BOARD
Thomas L. Albright University of Alabama Jacob G. Birnberg University of Pittsburgh Germain B. Boer Vanderbilt University William J. Bruns, Jr. Harvard University Peter Chalos University of Illinois, Chicago Donald K. Clancy Texas Tech University Robin Cooper Emory University Srikant M. Datar Harvard University Antonio Davila Stanford University Alan S. Dunk University of Canberra Nabil S. Elias University of North Carolina, Charlotte Kenneth J. Euske Naval Postgraduate School ix Eric G. Flamholtz University of California, Los Angeles George J. Foster Stanford University Eli M. Goldratt Avraham Y. Goldratt Institute John Innes University of Dundee Larry N. Killough Virginia Polytechnic Institute Thomas P. Klammer University of North Texas Carol J. McNair U.S. Coast Guard Academy James M. Reeve University of Tennessee, Knoxville Karen L. Sedatole Michigan State University George J. Staubus University of California, Berkeley Lourdes White University of Baltimore Sally K. Widener Rice University

STATEMENT OF PURPOSE AND REVIEW PROCEDURES
Advances in Management Accounting (AIMA) is a professional journal whose purpose is to meet the information needs of both practitioners and academicians. We plan to publish thoughtful, well-developed articles on a variety of current topics in management accounting, broadly defined. AIMA is to be an annual publication of quality-applied research in management accounting. The series will examine areas of management accounting, including performance evaluation systems, accounting for product costs, behavioral impacts on management accounting, and innovations in management accounting. Management accounting includes all systems designed to provide information for management decisionmaking. Research methods will include survey research, field tests, corporate case studies, and modeling. Some speculative articles and survey pieces will be included where appropriate. AIMA welcomes all comments and encourages articles from both practitioners and academicians.

REVIEW PROCEDURES
AIMA intends to provide authors with timely reviews clearly indicating the acceptance status of their manuscripts. The results of initial reviews normally will be reported to authors within eight weeks from the date the manuscript is received. Once a manuscript is tentatively accepted, the prospects for publication are excellent. The author(s) will be accepted to work with the corresponding Editor, who will act as a liaison between the author(s) and the reviewers to resolve areas of concern. To ensure publication, it is the author’s responsibility to make necessary revisions in a timely and satisfactory manner.

xi

EDITORIAL POLICY AND MANUSCRIPT FORM GUIDELINES
1. Manuscripts should be type written and double-spaced on 81/200 Â 1100 white paper. Only one side of the paper should be used. Margins should be set to facilitate editing and duplication except as noted: a. Tables, figures, and exhibits should appear on a separate page. Each should be numbered and have a title. b. Footnote should be presented by citing the author’s name and the year of publication in the body of the text; for example, Ferreira (1998) and Cooper and Kaplan (1998). 2. Manuscripts should include a cover page that indicates the author’s name and affiliation. 3. Manuscripts should include, on a separate lead page, an abstract not exceeding 200 words. The author’s name and affiliation should not appear on the abstract. 4. Topical headings and subheadings should be used. Main headings in the manuscript should be centered, secondary headings should be flush with the left-hand margin. (As a guide to usage and style, refer to the William Strunk, Jr., and E.B. White, The Elements of Style.) 5. Manuscripts must include a list of references, which contain only those works actually cited. (As a helpful guide in preparing a list of references, refer to Kate L. Turabian, A Manual for Writers of Term Papers, Theses, and Dissertations.) 6. In order to be assured of anonymous review, authors should not identify themselves directly or indirectly. Reference to unpublished working papers and dissertations should be avoided. If necessary, authors may indicate that the reference is being withheld for the reason cited above. 7. Manuscripts currently under review by other publications should not be submitted. Complete reports of research presented at a national or regional conference of a professional association and ‘‘State of the Art’’ papers are acceptable. 8. Four copies of each manuscript should be submitted to John Y. Lee at the address given below the Guideline 11. xiii xiv

EDITORIAL POLICY

9. A submission fee of $25.00, made payable to Advances in Management Accounting, should be included with all submissions. 10. For additional information regarding the type of manuscripts that are desired, see ‘‘AIMA Statement of Purpose.’’ 11. Inquires concerning Advances in Management Accounting may be directed to either of the editors: Marc J. Epstein Jones Graduate School of Administration Rice University Houston, TX 77251-1892 John Y. Lee Lubin School of Business Pace University Pleasantville, NY 10570-2799

INTRODUCTION
This volume of Advances in Management Accounting (AIMA) begins with a paper by Ashton on various models of value creation that have been proposed for supporting value-based management. Balanced Scorecard, the Baldridge Quality Award Criteria, the Service-Profit Chain, and the Skandia Intellectual Capital Model are among them. Similarities and differences among value-creation models are noted, their potential for guiding the identification of value drivers and performance measures for value-based management is assessed, and critical management issues that must be addressed if such models are to contribute to long-run value creation are explored. The substantial body of research evidence linking intangible value drivers to financial outcomes is reviewed, and some directions for further research are offered. This will become a valuable source for management accounting researchers, including doctoral students, in their research in this area. The next paper by Chow, Kohlmeyer, and Wu addresses the issues of innovation and risk. Innovation is the key to competitive advantage, and attaining innovation often requires taking on higher-than-usual levels of risk, but managers often emphasize safe, short-term results over more risky, long-term outcomes. As a result, a major challenge to firms is increasing employees’ willingness to adopt risky yet more profitable alternatives. This study uses an experiment to test how the level of performance standard, per se, affect employees’ propensity to take on (more) risky projects. Using participants from the U.S. and Taiwan to represent higher vs. lower individualism national cultures, it also examines the effects of national culture on employee actions. The findings are consistent with expectations from combining goal and prospect theories that a specific high standard motivates greater risk taking than a low standard. They find only limited difference between the U.S. and Taiwanese samples’ individualism/collectivism scores, which may help to explain the lack of significant differences between their reactions to the performance standard treatment. The paper by Elias and Notz tests the effects of two different organizational cultures on budgetary conflict. Budgetary conflict is perceived by conflicting parties as a zero sum game, or distributive – one party’s gain is the other party’s loss. They propose and test the effects of an empowering xv xvi

INTRODUCTION

organizational culture (EOC) in contrast to the traditional organizational culture (TOC). They hypothesize that an EOC would produce more integrative conflict resolution than the typical TOC. Using a laboratory experiment they confirmed their hypotheses that the EOC produces more integrative budget negotiation outcomes, greater convergence, and greater satisfaction with the outcome than TOC. The next paper by Kren and Maiga examines subordinate–superior information asymmetry as an intervening variable linking budgetary participation and slack. The results indicate two offsetting effects of participation on slack. A significant negative indirect relation between participation and slack was found to act through information asymmetry. Thus, managers reveal private information during the budget process, reducing information asymmetry that subsequently reduces budget slack. These results provide evidence about the inability of past research to confirm a consistent direct relation between budget participation and budget slack. The paper by Hartmann investigates whether acknowledgment of different types of uncertainty may explain the following apparently conflicting research findings: Research on budget-based performance evaluation traditionally predicts that the use of accounting performance measures (APM) in complex, dynamic, and uncertain situations results in dysfunctional managerial attitudes and behaviors. Although this suggests that such situations require the use of subjective performance measures (SPM), empirical evidence is inconclusive, as APM, rather than SPM, have been found to also have a negative effect on managerial ambiguity. This suggests that APM may be more, rather than less, appropriate than SPM in situations of high uncertainty. He develops hypotheses that predict differential interactions between the environmental uncertainty and task uncertainty and APM and SPM on managerial ambiguity. These hypotheses are tested using survey data from 250 managers in 11 organizations. Tests using moderated regression analysis provide support for the existence of different interactions between uncertainty and the use of performance measures, and provide reconciliation for the opposing findings in the extant literature. In the next paper, Bayou and Reinstein address the problem of the ineffective pricing methods available for smaller firms. The few management accounting pricing methods in the management accounting literature are ineffective in helping small firms use their idle capacity during lingering economic recessions, and some of these methods may even worsen this problem. Extending the traditional break-even-cost-volume-profit model, they derive a more effective pricing method, the break-even-full-capacity-utilization (BEFCU) model, to handle this problem. To demonstrate its practicality,

Introduction

xvii

the authors apply the BEFCU model to an actual job shop. This model integrates certain strategies based on built-in flexibility in commitments with suppliers and customers and maintaining a mode of conservatism in accounting for plant assets. The paper by Davis, Appel, and Lee provide the evidence that even when Murphy’s Law is objectively untrue, because of sampling bias, people perceive the law as true, and this perceptual bias has a far-reaching implication in management accounting research. A corollary to Murphy’s Law is: ‘‘The other lane always moves faster than my lane.’’ A manager who is aware of this perceptual bias will try to structure her budget cutbacks and all other ‘‘negative compensations’’ in such a way that her employees perceive that the cutback applies to everyone, not just to themselves. The findings of their study support the wisdom that, whenever managers implement managerial plans that will be perceived as ‘‘negative,’’ the plan should be implemented all at once. Spreading the implementation over a period of time produces more discontent on the part of the personnel affected. The findings lend credence to a generalization that people’s discontent is minimized when the number of observations (and thus the number of chances for forming a negative perception) of undesirable events is minimized. In the next paper, Maiga and Jacobs use structural equation modeling to investigate the impact of ABC implementation factors (management support, clarity and consensus of ABC objectives, nonaccounting ownership, and training) on quality, cost, and cycle time improvements. Overall, the results of the structural analyses support the theoretical model indicating that ABC implementation factors influence quality, cost and cycle time, and partial support for the relations among quality, cost and cycle time improvement and their effect on financial performance. The relationships are further analyzed within the context of ABC implementation stage, adoption of advanced manufacturing practices, industry characteristics, and plant size to determine if these contextual factors impact the model constructs and the relationships between the variables in the theoretical model. The results show that these contextual factors do not affect the model constructs. However, they affect the model relations. The paper by Kennedy and Schleifer examines how performance measurement can both encourage and hinder team performance. It then proposes a team performance measurement system using ratios and activity-based management that seek to encourage innovation and empowerment while maintaining a system of control. In the next paper, Fleming and Barkhi examine the influence of the psychological factor of social comparison over APM in a compensation experiment. The results of this study are consistent

xviii

INTRODUCTION

with social comparison theory in that CEO director-subjects award greater pay and shield the compensation of the CEO when firm accounting performance is below average. They also find shielding is mitigated when subjects are informed that the decision of the amount of compensation awarded will be revealed to the public. In a research note, Phillips, Daily, and Luehlfing address the consistency of readability levels before and after the recent changes in professional examinations for management accountants. We believe the 11 articles in Volume 16 represent relevant, theoretically sound, and practical studies the discipline can greatly benefit from. These manifest our commitment to providing a high level of contributions to management accounting research and practice. Marc J. Epstein John Y. Lee Editors

VALUE-CREATION MODELS FOR VALUE-BASED MANAGEMENT: REVIEW, ANALYSIS, AND RESEARCH DIRECTIONS
Robert H. Ashton
ABSTRACT
Models of value creation that have been proposed for supporting valuebased management are described and analyzed, including the Balanced Scorecard, the Baldrige Quality Award Criteria, the Deming Management Method, the Service-Profit Chain, and the Skandia Intellectual Capital Model. These models are compared, their potential for guiding the identification of value drivers and performance measures for valuebased management is assessed, and management issues that must be addressed if such models are to contribute to long-run value creation are explored. These issues include causally linking value drivers to each other and to financial outcomes, the extent to which the models take a dynamic, or whole-system, view of value creation, and whether multiple value drivers should be explicitly weighted and combined to form a ‘‘value index.’’ Finally, the substantial body of research evidence linking intangible value drivers to financial outcomes is reviewed, and some directions for further research are offered.

Advances in Management Accounting, Volume 16, 1–62 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16001-9

1

2

ROBERT H. ASHTON

INTRODUCTION
Adopting the perspective of value-based management, this paper analyzes several value-creation models that have appeared in the literature and that have guided both the practice of management and research in management accounting. Value-based management focuses on defining and implementing management strategies having the highest potential for creating shareholder value, identifying value drivers and aligning management processes that support value creation, and designing performance measurement and incentive systems that reflect value creation (Ittner & Larcker, 2001). The growing interest in value-based management reflects profound changes in the competitive business environment – involving, for example, technology, globalization, customer demands, greater attention to quality and service, and increased emphasis on business forms such as partnerships and alliances. These changes have led to dissatisfaction with traditional, transactions-based financial measurement and reporting systems for managing the firm (e.g., Johnson & Kaplan, 1987; Dixon, Nanni, & Vollmann, 1990; Eccles, 1991; Eccles & Pyburn, 1992). One result of the dissatisfaction with traditional measurement and reporting has been an increased interest in identifying and measuring ‘‘intangible value drivers’’ such as the skills and knowledge contained within the firm, the operational, customer-related, and innovation processes through which the firm manages today and prepares for tomorrow, and the relationships the firm has developed with key external stakeholders. The interest in firms’ potential, or capacity, for creating long-run shareholder value that ultimately will be realized is evidenced not only by proposals for more comprehensive measurement systems for internal use (e.g., Kaplan & Norton, 1992), but also by proposals for expanded external disclosures (e.g., American Institute of Certified Public Accountants (AICPA), 1994; OECD, 2000; Danish Ministry of Science & Technology, 2003). This interest is also supported by surveys of financial analysts’ information use (e.g., Previts, Bricker, Robinson, & Young, 1994; Dempsey, Gatti, Grinnell, & Cats-Baril, 1997; Low & Siesfeld, 1998) and by academic research documenting positive associations between non-financial disclosures and market outcomes (e.g., Maines et al., 2002, 2003). Moreover, the impact of these developments on management practice and management accounting research is being felt across a wide range of developed and developing economies (e.g., Epstein & Manzoni, 2002, 2004; Rejc, 2005). Because research in management accounting has tended to follow changes in management practice, much of today’s management accounting research

Value-Creation Models for Value-Based Management

3

involves a strategic focus on long-run value creation through the identification, measurement, management, and reporting of the drivers of firm value. A generic value-based management accounting framework involving six (largely sequential) steps is shown in Fig. 1. Briefly, the framework involves (1) choosing value-enhancing organizational objectives, (2) selecting strategies consistent with those objectives, (3) identifying performance variables, or value drivers, that create value, (4) selecting performance measures that reflect those value drivers, along with performance targets and action plans, (5) evaluating managerial performance and the success of

Overall Objective: Increase Shareholder Value

Identify Specific Organizational Objectives

Develop Strategies and Select Organizational Design

Identify Value Drivers

Develop Action Plans, Select Measures, and Set Targets

Evaluate Performance

Fig. 1.

Value-Based Management Accounting Framework. Source: Ittner and Larcker (2001).

4

ROBERT H. ASHTON

action plans, and (6) continuously assessing the overall value-based management framework and making modifications as needed (Ittner & Larcker, 2001). Steps (3)–(6) generally involve some model of value creation that guides the identification of value drivers and performance measures and their use in managing the firm. Many value-creation models have been proposed, each with a particular perspective on the drivers and measures considered essential for long-run value creation. The principal value-creation models are described and analyzed in this paper, beginning with the ‘‘quality management’’ prescriptions of Deming (1982) and including other quality- and customer-oriented models (the Baldrige Quality Award Criteria and the Service-Profit Chain), the Balanced Scorecard, and the Skandia Intellectual Capital Model. The basic features of these models are described, their potential for supporting value-based management is assessed, and research evidence that bears on the link between intangible value drivers and financial outcomes is reviewed. The paper is organized as follows. The value-creation models are described in the next two sections, where key features of each model are noted and similarities and differences among them are discussed. The following section examines several issues that arise in the use of such models by management, including the importance of understanding causal relationships that link value drivers to each other and to financial outcomes, the extent to which the models take a dynamic, or whole-system, view of value creation, and whether the measures associated with underlying value drivers should be explicitly weighted and combined to form a ‘‘value index.’’ Finally, research evidence that links many of the value drivers included in such models to financial outcomes at both the firm and market levels is reviewed, and some directions for additional research are suggested.

VALUE-CREATION MODELS
Models that involve mainly a quality orientation or a customer orientation are described first. Much overlap exists between quality and customer models because a principal rationale for improved quality is its positive effects on customer-related outcomes. This is followed by models that emphasize intellectual capital. Finally, the balanced scorecard and related strategy map are described, with emphasis on the fact that recent explications of the latter incorporate important aspects of both quality and customer models and intellectual capital models.

Value-Creation Models for Value-Based Management

5

Quality- and Customer-Oriented Models Deming Management Method Many important aspects of today’s value-creation models can be traced at least as far back as the writings of W. Edwards Deming. Among other contributions now often known collectively as the Deming Management Method (Walton, 1986), Deming developed a list of 14 Points that he considered ‘‘principles of transformation’’ for management practice. The 14 Points, largely oriented toward business processes, are a set of interrelated prescriptions that serve as guidelines for quality management practices. For the most part, they are loosely stated and contain such prescriptions as ‘‘Break down barriers between departments,’’ ‘‘Eliminate the need for inspection on a mass basis by building quality into the product,’’ and ‘‘Remove barriers that rob [management and employees] of their right to pride of workmanship’’ (Deming, 1982, pp. 16–17).1 Deming also described a set of cause-and-effect linkages – the Deming ‘‘Chain Reaction’’ – beginning with improved quality and ending with improvement in various organizational outcomes. The linkages in the Deming Chain Reaction involved hypothesized relationships among internal processes, costs, customer and market variables, and employee and organizational outcomes, as shown in Fig. 2. The elements of the chain reflected a strong emphasis on internal processes and were heavily oriented toward product quality, consistent with much of the total quality management (TQM) focus of the 1980s (e.g., Garvin, 1987). Anderson, Rungtusanatham, and Schroeder (1994) have articulated a theory of quality management based on concepts underlying Deming’s 14 Points, the Deming Chain Reaction, and other Deming contributions.
Costs decrease because of less rework, fewer mistakes, fewer delays, snags; better use of machine-time and materials.

Improve quality

Productivity improves

Capture the market with better quality and lower price

Stay in business

Provide jobs and more jobs

Fig. 2.

The Deming ‘‘Chain Reaction.’’ Source: Deming (1986).

6

ROBERT H. ASHTON

Internal and External Cooperation Organizational System Process Management

Continuous Improvement

Visionary Leadership

Process Outcomes

Customer Satisfaction

Learning

Employee Fulfillment

Causal Direction

Feedback Mechanism

Fig. 3.

Model Underlying the Deming Management Method. Source: Anderson et al. (1994).

Anderson et al. analyzed the content and evolution of the Deming Management Method, based on Deming’s writings and those of others, observations of practice, and a Delphi study involving a panel of Deming experts from academe and practice.2 Seven concepts underlying Deming’s work were identified – (1) visionary leadership, (2) internal and external cooperation, (3) learning, (4) process management, (5) continuous improvement, (6) employee fulfillment, and (7) customer satisfaction – and a theory of causal linkages among these concepts was proposed (Fig. 3). The breadth of the concepts and practices under the TQM umbrella expanded significantly from Deming’s early contributions. Juran (1989), for example, distinguished between ‘‘Little Q’’ and ‘‘Big Q’’ quality-improvement projects: ‘‘Little Q’’ projects focused narrowly on products, manufacturing processes, end purchasers of current products, and the costs associated with deficient products. In contrast, ‘‘Big Q’’ projects concerned services as well as products, support processes as well as manufacturing processes, internal as well as external customers, and a broader range of costs. As TQM continued to evolve, it became more oriented toward overall firm strategy. Included were customer and supplier relationships, employee empowerment, cross-functional training and product design, commitment to a ‘‘quality philosophy’’ (involving specific practices such as benchmarking, continuous improvement, and a lean and open organization), and an increased focus on measurement and analysis of organizational performance

Value-Creation Models for Value-Based Management

7

(Powell, 1995; Cua, McKone, & Schroeder, 2001; Douglas & Judge, 2001). TQM came to be seen as an integrated, systematic, organizationwide strategy for improving product and service quality y neither a program nor a specific tool or technique [but] a shift in both thinking and organizational culture y (Waldman, 1994, p. 511).

The original, process-oriented perspective of Deming’s work strongly influenced subsequent TQM developments, especially the Baldrige Quality Award Criteria, and both Deming’s work and elements of the Baldrige Model are reflected in later models of value creation. Baldrige Quality Award Criteria The Malcolm Baldrige National Quality Award was introduced in 1987 to promote quality awareness and practices and to publicize the quality achievements of U.S. companies. A major impetus for its introduction was widespread concern about the eroding competitive position of the U.S. Some of the major building blocks for the Baldrige Model were the quality prescriptions of Deming and others. A new framework for quality assessment was created, however, ‘‘above the views of particular practitioners or prescribers of specific quality systems’’ (Bell & Keys, 1998, p. 54). While the Baldrige Quality Award Criteria are perhaps best known for their role in evaluating applicants and selecting recipients of the Baldrige Quality Award, their principal use has been as a source of information about achieving performance excellence (Bemowski & Stratton, 1995). Only a few hundred companies have actually applied for the Baldrige Award, but more than 2 million copies of the performance criteria underlying the award were requested during the first 10 years of its existence (Flynn & Saladin, 2001). Most U.S. states and several countries have established quality awards based closely on the Baldrige Award, extending its impact even further. Thus, the Baldrige Model has evolved from an initial emphasis on promoting and recognizing quality management practices to a comprehensive framework for improving organizational performance, and is often used as a model for performance improvement (Flynn & Saladin, 2001). One analysis concluded that the Baldrige Model provides the most complete description in the world of what an organization capable of consistently delivering superior value to customers should look like (Gale, 1994, p. 323, emphasis in original).

The Baldrige Model incorporates ‘‘core values and concepts,’’ performance criteria and subcriteria, and a weighting scheme for the criteria and subcriteria. The core values and concepts, described as ‘‘embedded beliefs and

8

ROBERT H. ASHTON

Organizational Profile: Environment, Relationships, and Challenges

2 Strategic Planning 1 Leadership 3 Customer and Market Focus

5 Human Resource Focus 7 Business Results 6 Process Management

4 Information and Analysis

Fig. 4.

Baldrige Criteria for Performance Excellence Model. Source: Baldrige National Quality Program (2004).

behaviors found in high-performing organizations’’ (Baldrige National Quality Program, 2004), are: (1) visionary leadership, (2) customer-driven excellence, (3) organizational and personal learning, (4) valuing employees and partners, (5) agility, (6) focus on the future, (7) managing for innovation, (8) management by fact, (9) public responsibility and citizenship, (10) focus on results and creating value, and (11) systems perspective. The model, shown in Fig. 4, involves seven categories of performance criteria. Note that a Leadership category is included (unlike the other models described here). In fact, the model ‘‘starts’’ with leadership – based on the view that in an ‘‘assessment scheme based on cause/effect thinking, the most fundamental place to look for ‘cause’ is the place where vision, direction, and culture are created – senior leadership’’ (Bell & Keys, 1998, p. 55). Moreover, a ‘‘leadership triad’’ – consisting of the Leadership, Strategic Planning, and Customer and Market Focus categories – is said to drive a ‘‘results triad’’ – consisting of the Human Resource Focus, Process Management, and Business Results categories. The remaining category, Information and Analysis, serves as a foundation for the entire framework, while the Organizational Profile, which is not a part of the model per se, explicitly recognizes the importance to performance of the broad context in which the firm operates.

Value-Creation Models for Value-Based Management
Criteria and Subcriteria Leadership Organizational Leadership Public Responsibility and Citizenship Strategic Planning Strategy Development Strategy Deployment Customer and Market Focus Customer and Market Knowledge Customer Satisfaction and Relationships Information and Analysis Measurement and Analysis of Organizational Performance Information Management Human Resource Focus Work Systems Employee Learning and Motivation Employee Well-Being and Satisfaction Process Management Value Creation Processes Support Processes Business Results Customer-Focused Results Product and Service Results Financial and Market Results Human Resource Results Organizational Effectiveness Results Governance and Social Responsibility Results Total Points Point Values 120 70 50 85 40 45 85 40 45 90 45 45 85 35 25 25 85 50 35 450 75 75 75 75 75 75 1000

9

Fig. 5.

Performance Criteria and Point Values for the Baldrige Quality Award. Source: Baldrige National Quality Program (2004).

The model’s seven categories of performance criteria entail 19 first-level subcriteria, 32 second-level subcriteria, and several dozen items at an even more detailed level of analysis. Importance weights, or ‘‘point values,’’ are assigned to the criteria and first-level subcriteria for the purpose of selecting winners of the Baldrige Quality Award from each year’s pool of applicants, with 55% of the total points attached to value drivers and 45% to results (see Fig. 5).3 Interviews with Baldrige examiners reveal that companies scoring highest on the Baldrige criteria have both widespread deployment (horizontally and vertically) and extensive alignment of quality practices throughout the company (Garvin, 1991), as well as a determined focus on selected practices (Brown, 2003). Similarly, focus and alignment have been identified by Kaplan and Norton (2001a) as the two most critical issues in successful implementation of the Balanced Scorecard.

10

ROBERT H. ASHTON

Service-Profit Chain Linkages among employee- and customer-related value drivers and financial outcomes were proposed by Heskett, Jones, Loveman, Sasser, and Schlesinger (1994) in the Service-Profit Chain (see Fig. 6). The Service-Profit Chain brings together marketing and operations perspectives on value creation by linking customer-perceived quality with operational (or internal) quality processes (Soteriou & Zenios, 1999). Heskett et al. (1994, pp. 164–165) were explicit about the kinds of linkages envisioned:
The links in the chain (which should be regarded as propositions) are as follows: Profit and growth are stimulated primarily by customer loyalty. Loyalty is a direct result of customer satisfaction. Satisfaction is largely influenced by the value of services provided to customers. Value is created by satisfied, loyal, and productive employees. Employee satisfaction, in turn, results primarily from high-quality support services and policies that enable employees to deliver results to customers.

The Service-Profit Chain was later expanded by Heskett, Sasser, and Schlesinger (1997) to include additional concepts – for example, employee capability and output quality – and to introduce the ‘‘customer value equation,’’ expressed as the relationship between benefits and costs (both financial and non-financial) from the customer’s perspective (see Fig. 7). Heskett et al. (1997) described linkages among elements of the Service-Profit Chain for several companies, including the positive association between employee satisfaction and customer satisfaction (called the ‘‘satisfaction mirror’’). They also noted that ‘‘the elements of the service profit chain constitute a form of the balanced scorecard’’ (p. 210) that is especially relevant for service organizations. Their notion of ‘‘customer value’’ that links internal (employeerelated) variables to external (customer-related) variables – and ultimately to financial outcomes – provides a structured approach to understanding the determinants of customer profitability (Shapiro, Rangan, Moriarty, & Ross, 1987; Bellis-Jones, 1989; Foster & Gupta, 1994; Foster, Gupta, & Sjoblom, 1996; Kaplan & Narayanan, 2001; Zeithaml, Rust, & Lemon, 2001; Guilding & McManus, 2002), customer equity (Blattberg & Deighton, 1996; Lemon, Rust, & Zeithaml, 2001), and customer lifetime value (Rust, Lemon, & Zeithaml, 2000). The potential usefulness of the Service-Profit Chain for managing business processes is illustrated by Rucci, Kirn, and Quinn (1998), three executives at Sears, Roebuck and Company. The Sears approach, called the Employee-Customer-Profit Chain, emphasizes three key stakeholder groups – employees, customers, and investors – reflecting the premise that Sears wants to be ‘‘a compelling place to work,’’ ‘‘a compelling place to shop,’’ and ‘‘a compelling place to invest.’’ Based on data collected from employees

Value-Creation Models for Value-Based Management

Operating Strategy and Service Delivery System

Internal Service Quality

Employee Retention Employee Satisfaction Employee Productivity External Service Value Customer Satisfaction Customer Loyalty

Revenue Growth

Profitability retention repeat business referral

service concept: results for customers

workplace design job design employee selection and development employee rewards and recognition tools for serving customers

service designed and delivered to meet targeted customers needs

Fig. 6.

Links in the Service-Profit Chain. Source: Heskett et al. (1994).

11

12

Internal Operating strategy and service delivery system
Loyalty

External Service concept Target market

Customers
Satisfaction Productivity & Output quality

Revenue growth

Employees
Capability

Service value
Customer Value Equation:
Value to Customer =
(Results Produced for Customer + Service Process Quality) (Price to Customer + Service Acquisition Costs)

Satisfaction

Loyalty

Profitability

Service quality

ROBERT H. ASHTON

Fig. 7.

Service-Profit Chain. Source: Adapted from Heskett et al. (1997).

Value-Creation Models for Value-Based Management
A Compelling Place to Work A Compelling Place to Shop

13
A Compelling Place to Invest

Customer Recommendations

Attitude About the Job Employee Behavior
Attitude About the Company Employee Retention

Service Helpfulness Return on Assets Customer Impression Merchandise Value Customer Retention Operating Margin Revenue Growth

5 unit increase in employee attitude

DRIVES

1.3 unit increase in customer impression

DRIVES

0.5% increase in revenue growth

Fig. 8.

Employee-Customer-Profit Chain at Sears. Source: Rucci et al. (1998).

and customers, internal financial data, and related statistical analyses, Rucci et al. (1998) constructed a model of employee–customer–profit relationships that specifies value drivers and causal linkages among them (see Fig. 8). The model and related analyses allow statements such as ‘‘y a 5 point improvement in employee attitudes will drive a 1.3 point improvement in customer satisfaction, which in turn will drive a 0.5% improvement in revenue growth’’ (Rucci et al., 1998, p. 91). Such statements are indicative of the deep level of knowledge claimed by the Sears executives with respect to the value drivers incorporated in the model:
We understand the several layers of factors that drive employee attitudes, and we know how employee attitudes affect employee retention, how employee retention affects the drivers of customer satisfaction, how customer satisfaction affects financials, and a great deal more. We have also calculated the lag time between a change in any of those metrics and a corresponding change in financial performance, so that when we see a shift in, say, employee attitudes, we know not only how but also when it will affect results. Our [model] makes the employee-customer-profit chain operational because we manage the company on the basis of these indicators, with remarkably positive results (Rucci et al., 1998, p. 84, emphasis in original).

Epstein and Westbrook (2001) describe two other companies that have developed employee–customer–profit models that are similar in spirit to the Sears model. The Canadian Imperial Bank of Commerce (CIBC) developed models for four distinct customer groups that link leadership, employee commitment, customer loyalty, customer behavior, and profit.

14

ROBERT H. ASHTON

They found that ‘‘a 5% increase in employee commitment yields a 2% increase in customer loyalty, which increases profitability by $72 million annually’’ (p. 42). BFI, a large waste disposal company, focusing on customer dissatisfaction resulting from breaches in service dependability (e.g., missed pickups), modeled the path from service dependability to profit. One result: ‘‘A 2-point gain in service dependability led to a 1-point gain in overall customer satisfaction, which led to a 1% decline in customer defection, which produced a pretax increase in profits of $41 million’’ (p. 45). The CIBC, BFI, and Sears applications demonstrate that the Service-Profit Chain provides an integrative framework for understanding the links among investments in service quality, employee actions, customer perceptions, and behaviors – and how all of these translate into financial performance. Skandia Intellectual Capital Model Skandia, an insurance and financial services company with headquarters in Stockholm, has devoted substantial resources to understanding, measuring, and reporting what might be considered the ultimate non-financial value driver – intellectual capital. A working definition of intellectual capital that Skandia has used is:
The knowledge, skill and technologies Skandia uses to create a competitive edge. This includes accessible knowledge and the applied experiences of all employees, and the organizational structure, technology and professional systems within a firm. Intellectual capital is the soft and intangible part of a company’s value (Oliver, 1996, p. 6).

Skandia was the first company to appoint a Director of Intellectual Capital (Leif Edvinsson), and a series of internal studies of the company’s ‘‘hidden value’’ was begun in 1992.4 A prominent feature of the approach is the Skandia Business Navigator, shown in Fig. 9. The Navigator is depicted as a house. The foundation of the house is termed the renewal and development focus, and is oriented toward the future. The walls consist of the customer focus and the process focus, both oriented toward the present, while the roof reflects the financial focus, oriented toward the past. The human focus constitutes the center of the house. The Navigator is similar in many respects to the Balanced Scorecard (described later): The financial and customer dimensions are common to the two frameworks, and the Navigator’s process and renewal and development dimensions correspond closely to the Balanced Scorecard’s internal business processes and learning and growth dimensions, respectively. However, the

Value-Creation Models for Value-Based Management

15

HISTORY

FINANCIAL FOCUS

INTELLECTUAL CAPITAL

TODAY

CUSTOMER FOCUS

HUMAN FOCUS

PROCESS FOCUS

TOMORROW

RENEWAL & DEVELOPMENT FOCUS

Fig. 9.

Skandia Business Navigator. Source: Adapted from Skandia (1994).

Navigator has a fifth dimension – the human focus – whereas the Balanced Scorecard includes the human focus as only one of several components within its learning and growth dimension. Moreover, the Skandia framework highlights intellectual capital – comprising the human, customer, process, and renewal and development dimensions – as the ultimate value driver. The critical role of intellectual capital is developed further in a related framework known as the Skandia Value Scheme (Edvinsson, 1997; Edvinsson & Malone, 1997), shown in Fig. 10. It maintains that a firm’s market value results from its financial capital and its intellectual capital. Of course, financial capital and intellectual capital cannot literally be summed to get market value, and no claim is made that the difference between market value and financial capital is a measure of a firm’s intellectual capital at a particular point in time. Instead, financial capital, which generally is quantified and disclosed in financial reports, and intellectual capital, which generally is not, are viewed from a conceptual standpoint as the two major types of value drivers for a firm. The Skandia Value Scheme posits that intellectual capital is composed of narrower classes of value drivers, which in turn are composed of even narrower classes, and so on, in an effort to make the conceptual notion of intellectual capital less abstract. First, intellectual capital is seen as resulting from human capital and structural capital. Human capital includes knowledge, skills, and experience, while structural capital includes value drivers

16

ROBERT H. ASHTON

Market value

Financial capital

Intellectual capital

Monetary capital

Physical capital

Human capital

Structural capital

Customer capital

Organizational capital

Innovation capital

Process capital

Intellectual property

Intangible assets

Fig. 10.

Skandia Value Scheme. Source: Adapted from Edvinsson (1997).

that are internal to the firm (e.g., processes, routines, databases, and ‘‘culture’’) and external to the firm (e.g., relationships with customers, suppliers, and alliance partners). The two major components of structural capital are customer capital and organizational capital. Organizational capital, in turn, results from both organizational processes that apply existing knowledge and innovations that generate new knowledge. Innovation capital includes intellectual property (intellectual capital that is legally protected) and intangible assets that may be quantified and disclosed in financial reports. The Skandia Value Scheme has been further refined to produce the Intellectual Capital Distinction Tree (Roos, Roos, Edvinsson, & Dragonetti, 1998) which offers a more detailed breakdown of human capital and structural capital, the two principal determinants of a firm’s intellectual capital (see Fig. 11). Human capital is divided into competence, attitude, and ‘‘intellectual agility,’’ which together capture a wide range of variables at the individual level. Structural capital is divided into relationships (with customers, alliance partners, and other stakeholders), organization (e.g., processes and ‘‘culture’’), and renewal and development, which together capture a wide range of variables at the organizational and market levels. The original goal of Skandia’s intellectual capital efforts was to communicate to the outside world (mainly the capital markets) the company’s

Value-Creation Models for Value-Based Management

17

Total Value

Financial Capital

Intellectual Capital

Human Capital

Structural Capital

Competence • Knowledge • Skills

Attitude • Motivation • Behavior • Conduct

Intellectual Agility • Innovation • Imitation • Adaptation • Packaging

Relationships

Organization

Renewal and Development

• Infrastructure • Customers • Processes • Suppliers • Culture • Alliances • Shareholders • Other Stakeholders

Fig. 11.

Intellectual Capital Distinction Tree. Source: Roos et al. (1998).

‘‘hidden value’’ by putting metrics on intellectual capital (Bartlett & Mahmood, 1996) and, toward that end, one result was the publication of several Intellectual Capital Supplements to their annual and interim financial reports (e.g., Skandia, 1994, 1995a, 1995b, 1996a, 1996b, 1997, 1998). It was soon realized, however, that the Skandia Navigator could also be used as ‘‘a tool to steer the organization’’ (Oliver, 1996, p. 9) and to support a new business model and organizational structure that were widely viewed as highly innovative (e.g., Goldman Sachs Global Equity Research, 2000).5 Key elements of the Skandia Intellectual Capital Model, especially the distinction between human capital and structural capital, have strongly influenced subsequent work on value creation, including the most recent exposition of the Balanced Scorecard. Balanced Scorecard and Strategy Map The Balanced Scorecard was developed during a period of intense interest in TQM principles. Several authors in the late 1980s and early 1990s commented on the limitations of financial performance measurement systems for decision making and control in an environment of dramatically increased quality-consciousness. While some of the concern stemmed from issues

18

ROBERT H. ASHTON

surrounding traditional standard costing systems (Berliner & Brimson, 1988; Fisher, 1992; Maisel, 1992), perspectives provided by the quality movement were critical (Beischel & Smith, 1991; Eccles, 1991). Interest in performance measurement systems that combined financial and nonfinancial (especially operational) measures led to development of the Balanced Scorecard, and the related interest in linking such measures to each other and to firm strategy subsequently led to the Strategy Map.

Balanced Scorecard The Balanced Scorecard (Kaplan & Norton, 1992), shown in Fig. 12, emphasizes goals and measures within four perspectives: (1) Financial: How do we look to shareholders? (2) Customer: How do customers see us? (3) Internal Business: What must we excel at? (4) Innovation and Learning: Can we continue to improve and create value?6 Hoque (2003) argues explicitly that a firm following TQM principles needs a performance management system such as the Balanced Scorecard to achieve continuous improvement. Moreover, he develops specific linkages among TQM activities, TQMrelated performance measures, and the four scorecard perspectives.7 Nonfinancial measures – especially in the internal business processes and
How do we look to shareholders? Financial Perspective Goals Measures

How do customers see us? Customer Perspective Goals Measures

What must we excel at? Internal Business Perspective Goals

Innovation and Learning Perspective Can we continue to improve and create value? Goals Measures

Fig. 12.

Balanced Scorecard. Source: Kaplan and Norton (1992).

Value-Creation Models for Value-Based Management

19

learning and growth perspectives – are the key components of the linkages developed. While development of the Balanced Scorecard was influenced by the TQM focus of the 1980s, the advantages of using a balanced set of measures, including non-financial measures, to evaluate organizational performance had been appreciated for quite some time. Drucker (1954, p. 87), for example, maintained that management must balance the firm’s objectives:
There are few things that distinguish competent from incompetent management quite as sharply as the performance in balancing objectives y Objectives in the key areas are the ‘‘instrument panel’’ necessary to pilot the business enterprise.

To achieve this balance, Drucker argued that firms need performance measures in eight areas: (1) market standing, (2) innovation, (3) productivity, (4) physical and financial resources, (5) profitability, (6) manager performance and development, (7) worker performance and attitude, and (8) public responsibility.8 A second early example of the recognition of the importance of using a balanced set of performance measures was the Tableau de Bord (‘‘dashboard’’ or ‘‘instrument panel’’), developed in France several decades ago (Epstein & Manzoni, 1997, 1998; Lebas, 1994). The Tableau de Bord incorporated non-financial measures that were tailored to responsibility centers and/or the firm as a whole at operational, management, and strategic levels, evolving initially from a reporting emphasis to a managerial tool (Bourguignon, Malleret, & Norreklit, 2004).9 Kaplan (1998) has remarked that the Tableau de Bord has had little impact on managerial ´ practice, but a survey by Gehrke and Horvath (2002) suggests that it has – at least in France. The Balanced Scorecard has been developed more extensively than the other value-creation models discussed here and has been adopted by a large and diverse set of companies, including non-profits.10 Its popularity can be explained by at least two factors. One is the insight of its developers to embed the scorecard within the firm’s strategic management system. The other is the recognition that the scorecard can enable a firm to communicate the corporate vision, promote organizational alignment, and create shared understanding – providing a ‘‘line of sight’’ from the activities of individual employees to overall firm objectives and performance. Thus, the Balanced Scorecard is more than a measurement framework, as Kaplan and Norton’s (1996c) first book suggests by devoting roughly equal space to ‘‘measuring strategy’’ and ‘‘managing strategy.’’ They argue that the Balanced Scorecard can serve as the cornerstone of a firm’s strategic management system by supporting four management processes that ‘‘contribute to linking

20

ROBERT H. ASHTON

long-term strategic objectives to short-term actions’’ (Kaplan & Norton, 1996d, p. 75):  Translating the vision: Clarifying the vision, gaining consensus  Communicating and linking: Communicating and educating, setting goals, linking rewards to performance measures  Business planning: Setting targets, aligning strategic objectives, allocating resources, establishing milestones  Feedback and learning: Articulating the shared vision, supplying strategic feedback, facilitating strategy review and learning Kaplan and Norton (2001a) describe the experiences of several early adopters of the Balanced Scorecard, focusing especially on its use for strategic management purposes. When used effectively, it was observed, the scorecard allowed companies to describe and communicate the strategy in understandable and actionable ways, align all resources and activities to the strategy, and establish organizational linkages across business units, shared services, and individuals. Thus, the management systems in which successful scorecards were embedded had three distinct characteristics – strategy, focus, and organization – leading Kaplan and Norton to the notion of a ‘‘strategy-focused organization’’ and to the Strategy Map, a refinement and extension of the Balanced Scorecard. Strategy Map Strategy-focused organizations are said to embody five principles: (1) translating the strategy to operational terms, (2) aligning the organization to the strategy, (3) making strategy everyone’s everyday job, (4) making strategy a continual process, and (5) mobilizing change through executive leadership (Kaplan & Norton, 2001a). While these principles, working together, can enable a company to effectively implement its strategy, implementation cannot be achieved without first describing and communicating the strategy. Thus, the Strategy Map was developed as a ‘‘logical and comprehensive architecture for describing strategy’’ (Kaplan & Norton, 2001a, p. 10).11 Strategy Maps maintain the original four perspectives of the Balanced Scorecard, tailor each perspective according to the firm’s (or business unit’s) particular situation, and posit detailed cause-and-effect linkages within and across the four perspectives – and in turn to strategic financial themes. The generic Strategy Map depicted in Fig. 13 illustrates the broad financial themes of revenue growth, cost management, and asset utilization, and it encompasses broad customer value propositions related to product leadership, customer intimacy, and operational excellence. Broad themes are

Value-Creation Models for Value-Based Management

Improve Shareholder Value Financial Perspective Revenue Growth Strategy
Increase Customer Value Customer Profitability Customer Retention Product Leadership Customer Intimacy Shareholder Value ROCE

Productivity Strategy
Improve Cost Structure Cost per Unit Improve Asset Utilization Asset Utilization

Build the Franchise New Revenue Sources

Customer Acquisition

Customer Perspective
Price Quality

Customer Value Proposition
Product/Service Attributes Time Functionality

Operational Excellence Image Brand

Relationship Service Relationships

Customer Satisfaction “Increase Customer Value” (Customer Management Processes) “Achieve Operational Excellence” (Operational Processes) “Be a Good Corporate Citizen” (Regulatory and Environmental Processes)

Internal Perspective

“Build the Franchise” (Innovation Processes)

Learning and Growth Perspective

A Motivated and Prepared Workforce
Strategic Competencies Strategic Technologies Climate for Action

Fig. 13.

Generic Template for a Balanced Scorecard Strategy Map. Source: Kaplan and Norton (2001a). 21

22

ROBERT H. ASHTON

also shown for the internal perspective and the learning and growth perspective, and hypothesized causal linkages connect all of these themes and perspectives to the ultimate objective of increasing shareholder value. Strategy Maps have come to occupy a central position in Kaplan and Norton’s thinking:
[W]e have learned how the Balanced Scorecard, initially proposed to improve the measurement of an organization’s intangible assets, can be a powerful tool for describing and implementing an organization’s strategy y. We now realize that the strategy map, a visual representation of the cause-and-effect relationships among the components of an organization’s strategy, is as big an insight to the executives as the Balanced Scorecard itself (Kaplan & Norton, 2004b, p. 9).

They argue that Strategy Maps add a level of granularity that improves clarity and focus, effectively illustrate the temporal dynamics of strategy, and provide a uniform and consistent way to describe strategy. As the Strategy Map has continued to evolve, the most critical differences between it and the initial formulation of the Balanced Scorecard concern the internal perspective and the learning and growth perspective, which together ‘‘drive the strategy.’’ As illustrated in Fig. 14, the key roles of three types of internal business processes are further developed and clarified – processes related to operations management, customer management, and innovation. The internal perspective also features a substantially increased emphasis on a fourth set of internal processes – regulatory and social processes related to the environment, safety and health, employment, and the community. Earlier versions of the Balanced Scorecard had sometimes been criticized for what some saw as insufficient concern with stakeholders other than shareholders. The learning and growth perspective has been developed even more extensively than the internal perspective, especially with respect to linking intangible assets to strategy and ultimately to financial outcomes. The development of the Balanced Scorecard’s learning and growth perspective over time is shown in Fig. 15. The most recent version classifies intangible assets into three categories:  Human capital: Skills, knowledge, competences  Information capital: Information systems, databases, networks, technology infrastructure  Organization capital: Culture, leadership, alignment, teamwork The similarity of these three categories to key elements of the Skandia Intellectual Capital Model is apparent. The Strategy Map goes further than the Skandia Model, however, by considering ways of establishing bridges

Value-Creation Models for Value-Based Management

Productivity Strategy

Growth Strategy

Financial Perspective
Improve Cost Structure Increase Asset Utilization

Long-Term Shareholder Value
Expand Revenue Opportunities Enhance Customer Value

Customer Value Proposition

Customer Perspective

Price

Quality

Availability

Selection

Functionality

Service

Partnership

Brand Image

Product/ Service Attributes

Relationship

Operations Management Processes

Customer Management Processes
Selection Acquisition Retention Growth

Innovation Processes
• Opportunity ID • R&DPortfolio • Design/Develop • Launch

Regulatory and Social Processes
• Environment • Safety and Health • Employment • Community

Internal Perspective

Supply Production Distribution Risk Management

Learning and Growth Perspective

Human Capital Information Capital Organization Capital
Culture Leadership Alignment Teamwork

Fig. 14.

Balanced Scorecard Strategy Map. Source: Kaplan and Norton (2004b). 23

24
A. Learning and Growth Perspective in 1996
Staff Competencies
Strategic skills Training levels Skill leverage

ROBERT H. ASHTON

Technology Infrastructure
Strategic technologies Strategic databases Experience capture Proprietary software Patents, copyrights

Climate for Action
Key decision cycle Strategic focus Staff empowerment Personal alignment Morale Teaming

B. Learning and Growth Perspective in 2001
Strategic Competencies Strategic Technologies Climate for Action

Skills • Strategic skill coverage ratio

Knowledge Sharing • Best practice sharing

Infrastructure

Applications

Awareness • Understanding of strategy

Alignment • Goals aligned with Balanced Scorecard

Readiness • Average tenure (key positions)

Motivation • Morale (satisfaction) • Suggestion program (empowerment)

• Strategic technology coverage

C. Learning and Growth Perspective in 2004 CREATING ALIGNMENT
Strategic Job Families Strategic IT Portfolio Organization Change Agenda

CREATING READINESS

Human Capital • Skills • Training • Knowledge

+

Information Capital • Systems • Databases • Networks

+

Organization Capital • Culture • Alignment • Leadership • Teamwork

Fig. 15.

Balanced Scorecard: Learning and Growth. Source: Kaplan and Norton (1996c, 2001a, 2004b).

from these three types of capital to the internal perspective of the scorecard and, in turn, to the customer and financial perspectives. Human, information, and organization capital are linked to the rest of the scorecard via strategic job families, the strategic IT portfolio, and the organization change agenda, respectively. These three bridges provide operational platforms for achieving the organizational ‘‘readiness’’ (Kaplan & Norton, 2004a) that intangible assets can provide.

COMPARING THE MODELS
All of these models involve value drivers in three fundamental (and overlapping) categories: people, processes, and relationships. People include

Value-Creation Models for Value-Based Management

25

employees and managers inside the firm, of course, but also customers and suppliers outside the firm and other stakeholders as well. Processes include those with both short- and long-run time frames, for example, quality assurance processes, customer relationship processes, human resource processes, R&D and innovation processes, and control and learning processes. Relationships include those linking the firm to both its immediate competitive environment (involving suppliers, distributors, end users, alliance partners, competitors, etc.) and its broader social and natural environments (e.g., the community, regulators, and interest/pressure groups). People, processes, and relationships are viewed as sources of a firm’s ‘‘value potential’’ or ‘‘value capacity.’’ The function of management that is implied by these models involves both the transformation of intangible sources of value into value potential (via people, processes, and relationships) and the transformation of value potential into realized value (via transactions). Tangible sources of value (physical and monetary capital) are still considered essential for value creation, but the emphasis of these models is on intangible sources of value. The models emphasize the capacity of such sources for value creation, and less attention is paid to value realization in the form of cash, profit, or shareholder return. All of the models incorporate a combination of financial and nonfinancial measures, lead and lag measures, ‘‘hard’’ and ‘‘soft’’ measures, and measures that focus on both the short run and the long run. All of the models envision using such measures in the implementation, rather than the formulation, of firm strategy. All entail measures of value drivers that are both internal and external to the firm. And all make clear the importance of using measures to foster improvement instead of just control. None of the models is a stand-alone approach to management, but must be embedded in and supported by other systems – including organizational design systems, incentive and reward systems, executive development systems, and knowledge management and organizational learning systems. All of the models reflect measurement approaches, as opposed to valuation approaches, to understanding value creation. Measurement approaches focus on underlying dimensions of performance, allowing measures to be expressed in a variety of units (e.g., dollars, employee satisfaction scores, customer retention percentages, defect rates, response times). By focusing on underlying dimensions of performance at a disaggregate level, measurement approaches facilitate the construction of causal chains that link fundamental value drivers with financial outcomes. Valuation approaches, in contrast, restrict measures to a monetary unit, emphasizing aggregate financial outcomes. Several

26

ROBERT H. ASHTON

valuation approaches exist, including comprehensive models such as Economic Value Added (Stewart, 1991; Ehrbar, 1998) and Shareholder Value Added (Rappaport, 1986, 1998) that provide aggregate measures such as economic profit, and more specialized models that emphasize the importance of intangibles or intellectual capital and that provide aggregate measures such as ‘‘knowledge earnings’’ (Lev & Mintz, 1999; Mintz, 2000; Webber, 2000).12 Note that measurement approaches to value creation are more comprehensive than valuation approaches, as the summary measures resulting from the latter can be included among the financial measures of the former. While many similarities exist among these value-creation models, they differ in the types of value drivers, and therefore the types of measures, that are emphasized. For example, the Skandia Model takes more of an ‘‘insideout’’ view of value creation, while the Balanced Scorecard takes more of an ‘‘outside-in’’ view. The difference can be thought of in the context of the SWOT framework (strengths, weaknesses, opportunities, threats), which plays a dominant role in much strategic thinking. Following Porter (1980), the relevance for strategy of the SWOT framework’s external elements (opportunities and threats) is often emphasized more than that of its internal elements (strengths and weaknesses), even though managers have greater influence over the latter. The Porterian view – which emphasizes the bargaining power of the firm’s suppliers and customers, the threat of new entrants and substitute products, and the actions of competitors – can be contrasted with the more recent resource-based view of the firm, which emphasizes internal resources and capabilities. The resource-based view maintains that resources and capabilities create strategic advantage to the extent they are valuable, rare, inimitable, and non-substitutable (e.g., Wernerfelt, 1984; Barney, 1991, 1995; Foss, 1997). Thus, the Porterian, or external, view is basically outside-in, while the resource-based, or internal, view is basically inside-out. Both the Skandia Model and the Balanced Scorecard clearly recognize the importance of both outside-in and inside-out perspectives on value creation, but the Balanced Scorecard has somewhat more of an outside-in flavor while the Skandia Model clearly emphasizes inside-out. This distinction must be tempered, however, by recognition that recent expositions of the Balanced Scorecard include a larger role for intangibles and intellectual capital in its learning and growth dimension. In Kaplan and Norton (2001a) this is reflected in the learning and growth categories of strategic competencies, strategic technologies, and ‘‘climate for action,’’ and in Kaplan and Norton (2004b) it is reflected in the categories of human capital, information capital, and organization capital.

Value-Creation Models for Value-Based Management

27

Because value-creation models differ in their relative emphasis on different types of value drivers, the types of measures employed will likely differ among models as well. The Service-Profit Chain, for example, strongly emphasizes operations- and customer-related drivers of quality products and services, but does not emphasize research and development or learning systems that promote firm-wide innovations. This has implications for the types of non-financial measures that a Service-Profit Chain approach to value creation would likely emphasize. And while the Service-Profit Chain can be viewed as a detailed expansion of the Balanced Scorecard’s customer dimension (as well as customer-related elements of its internal business processes dimension), the Balanced Scorecard includes a more comprehensive set of value drivers across the entire value chain (Porter, 1985), and thus entails a broader (more ‘‘balanced’’) set of non-financial measures. Similarly, because the Baldrige Model emphasizes leadership as a key value driver, measurement systems based on the Baldrige Model would likely emphasize measures of leadership effectiveness more strongly than would measurement systems based on other models.

MANAGEMENT ISSUES
The principal goal of value-creation models such as these is to support management activities aimed at long-run shareholder value creation, in part through guiding the identification and measurement of tangible and intangible value drivers. Achieving this goal will require that management address several complex issues that arise in the formulation and use of such models. This section considers several important issues including the need to understand causal linkages among value drivers and outcomes, the extent to which these value-creation models take a dynamic, or whole-system, view of value creation, and whether multiple value drivers should be explicitly weighted and combined to form a ‘‘value index.’’ Causal Linkages Perhaps the most critical issue in using any value-creation model for management purposes is the extent to which it embodies chains of causeand-effect relationships that link measures of intangible value drivers to each other and, ultimately, to financial outcomes such as profit, cash, and share return. Causal relationships could be estimated statistically based on careful empirical analysis of historical data, or they could be estimated

28

ROBERT H. ASHTON

subjectively based on careful reasoning by knowledgeable managers and others. Whatever their basis, some understanding of causal linkages among measures related to a model’s value drivers is required if the model is to be useful for management purposes. Over time, the proposed causal linkages included in value-creation models have become more prevalent, more comprehensive, and more empirically based. The model underlying the Deming Management Method, the earliest model reviewed here, entails hypothesized causal relationships, but they do not extend to financial outcomes. The Baldrige Model hypothesizes causal linkages that are better developed than those in the Deming Model but not as well developed as those in later models. The Skandia Intellectual Capital Model also entails the notion of cause and effect (i.e., human capital is seen as the driver of structural capital). The Service-Profit Chain and the Balanced Scorecard Strategy Map strongly emphasize causal linkages. The former involves linkages that connect operational and service measures to customer measures to financial measures, and the latter involves linkages that connect learning and growth measures to internal business process measures to customer measures to financial measures. Moreover, as demonstrated in the next section of the paper, many of the proposed linkages in the Service-Profit Chain and the Balanced Scorecard Strategy Map have been examined by empirical research. The Strategy Map is especially strong with respect to causal linkages. Recall the earlier observation that the Service-Profit Chain can be viewed as a detailed expansion of the customer dimension of the Balanced Scorecard (and customer-related elements of the internal business processes dimension). Kaplan and Norton (2001a) demonstrate that the Strategy Map can represent links in the Service-Profit Chain by recasting the Sears EmployeeCustomer-Profit Chain into the Strategy Map framework (see Fig. 16). In addition, the applicability of the Strategy Map to a functional area of an organization – specifically, the human resources function – is demonstrated by Datar and Epstein’s (2001) portrayal of the GTE ‘‘HR Linkage Model’’ as a Strategy Map (see Fig. 17). In this application, the operations and customer dimensions of the GTE model do not refer to the operations or customers of GTE, but to the operations and customers of GTE’s HR function. Nevertheless, this functional Strategy Map connects with the larger organization’s strategy in important ways, for example, by capturing the five key ‘‘strategic thrusts’’ that GTE identified as foundational for its future financial success. Shown at the foundation of the model in Fig. 17, these are talent, strategic competencies, a performance-based culture, organizational integration, and leadership (Becker, Huselid, & Ulrich, 2001).13

Value-Creation Models for Value-Based Management

29

Compelling Place to Invest

Revenue Growth Profit Contribution Sales per Square Foot Inventory Turnover

Customer Retention

Compelling Place to Shop

Customer Impression

Value/Price

Merchandise Selection

Returning Merchandise

Image

Turnover

Associates’ Behavior

• Customer Service • Recommend Products • Loyalty

Compelling Place to Work

Attitude about the Job

Attitude about the Company

Job Supervision Managers: • Business Knowledge • Customer Service Orientation • Job Context

Job Structure • Teamwork • Training

Job Context • Sears Being Fair/Ethical • Promotion Opportunities • Pay and Benefits

Fig. 16.

Strategy Map of the Sears Employee-Customer-Profit Chain. Source: Kaplan and Norton (2001a).

Static versus Dynamic View The second management-related issue is the extent to which value-creation models take a static versus a dynamic view of the process of value creation. At the risk of oversimplifying, a static view tends to regard causes as independent of each other, and as one-time events each of which produces an

30

FINANCIAL
Maximize Human Capital Corporate/Business Unit Business Partner (Strategic Support)

Contribute to Corporate Shareholder Value

Minimize HR Costs Employees

CUSTOMER

Organizational Health and Competitive Capability

Skills, Competencies, and Leadership

Low-Cost Provider

OPERATIONS
Align HR Planning with Business Strategy Provide Proactive Workforce Solutions Ensure a StrategyFocused Workforce Develop and Enhance WorldClass Programs Optimize Service Delivery through Streamlined Processes

STRATEGIC
Talent Capability (Build Strategic Competencies) Enable a Performance-Based Culture/Climate Organizational Integration • External trend data - HR best practices/ breakthroughs • Internal employee data - demographics • Organizational strategy • Industry trends • Integrated technology infrastructure (SAP) Leadership

• Grow the talent pool • Select, assimilate, and retain key talent • Organizational renewal • High potential development • Reduce turnover

• Culture that values: • Service delivery design - results • Organizational change skills - customer - staffing expectations - open communications - design interventions • Climate that exhibits: - provide reinforcement - flexibility • Relationship building - clarity • HR planning - high standards • Performance management • Workforce planning

• Invest in leadership growth • Leadership competencies • Structure rewards to foster leadership behavior

ROBERT H. ASHTON

Fig. 17.

GTE’s HR Linkage Model. Source: Datar and Epstein (2001).

Value-Creation Models for Value-Based Management

31

effect. A dynamic view, in contrast, tends to focus on patterns in which multiple causes produce multiple effects via sets of causal relationships in which effects feed back to influence causes and causes affect other causes in an ongoing process (Richmond, 1993, 2000). This is part of the system dynamics view of the world, where it is often maintained that ‘‘everything is connected to everything else’’ through a network of feedback loops and time delays (Sterman, 2000, p. 4). In essence, system dynamics takes a wholesystem view of the firm, its interconnected value-creating processes, and the complex network of economic relationships in which it is embedded (Allee, 2003; Dyer & Singh, 1998; Post, Preston, & Sachs, 2002). While the implications of the system dynamics view for modeling causal linkages among value drivers and outcomes have occasionally been recognized (e.g., Kaplan & Norton, 2001a, pp. 311–313), the value-creation models described in this paper generally do not explicitly reflect this view. However, several aspects of system dynamics suggest its relevance in the context of understanding long-run value creation (Klein, 1998, Sterman, 2000). For example, system dynamics clarifies that effects have multiple causes, causes produce multiple effects, and effects are often not proportional to causes. In addition, system dynamics emphasizes the importance of self-correcting and self-reinforcing feedback loops among value drivers.14 Finally, system dynamics focuses management’s attention on the sometimes counterintuitive outcomes that can result from networks of causal relationships, feedback loops, and the varying time delays that characterize valuecreation processes. Thus, system dynamics provides a powerful structure for modeling causal linkages, and for explicitly incorporating network effects into the understanding of value creation.

Weighting and Combining Measures The third management-related issue is whether the various non-financial measures that reflect a model’s key value drivers should be explicitly weighted and combined to form a ‘‘value index.’’ The advantages and disadvantages of using an index or composite measure versus a disaggregated set of measures to assess individual or organizational performance have been debated for decades (e.g., Ridgway, 1956, Schmidt & Kaplan, 1971). Weighting and combining measures reflects a ‘‘middle ground’’ between the measurement approach (disaggregate measures expressed in different units) and the valuation approach (a summary measure expressed in a monetary unit) to value creation. A value index could serve as a summary

32

ROBERT H. ASHTON

measure of current performance and possibly as a predictor of future performance. An index that has received attention recently is the Value Creation Index developed by Ernst & Young’s Center for Business Innovation (Low, 2000; Baum et al., 2000; Kalafut & Low, 2001; Funk, 2003). The index is based on data from a number of internal studies, particularly the ‘‘Measures that Matter’’ study (Ernst & Young LLP, 1997; Low & Siesfeld, 1998), as well as ‘‘industry literature and conversations with business and academic researchers’’ (Low, 2000, p. 254). The index summarizes nine categories of nonfinancial information: (1) innovation, (2) quality, (3) customer relations, (4) management capabilities, (5) alliances, (6) technology, (7) brand value, (8) employee relations, and (9) environmental and community issues. These categories, each involving several specific measures, were statistically combined to form the Value Creation Index. While such an index could be used as a management tool, Ernst and Young has emphasized the relationship between index scores and firms’ market values, finding an average correlation of 0.70. Of the nine valuedriver categories, innovation (as reflected in R&D expenditures and the number and importance of patents) has the highest association with market value, followed closely by management quality and employee relations. Alliances (both manufacturing and marketing, as well as joint ventures and other forms of partnership) are also positively associated with market value. Perhaps surprisingly, technology and customer satisfaction are unrelated to market value. It is possible, however, that a threshold level of technology and customer satisfaction are required just to ‘‘be in the game’’ but that they do not act as differentiators across companies. The Baldrige Model, which originally was developed for the purpose of awarding quality prizes to U.S. companies, also involves an explicit weighting scheme that produces composite scores (see Fig. 5). If the Baldrige Model is to be applied consistently in evaluating applicants for quality awards, a standard weighting scheme is needed. Of course, when the Baldrige Model is used as part of a strategic management system instead of in quality contests, it is not necessary to use the standard weighting scheme (or any weighting scheme) to compute composite scores. Weighting schemes could be applied, and indexes constructed, for the other models described here, but little interest has been shown in doing so. Some attention has been focused on an Intellectual Capital Index based on the Skandia Value Scheme, but this effort appears to be in a preliminary stage (Roos et al., 1998; Roos & Jacobsen, 1999). Weighting factors are sometimes applied to Balanced Scorecard measures, or at least to

Value-Creation Models for Value-Based Management

33

the four overall scorecard categories, but the extent of this practice is not clear. It is sometimes maintained, however, that Balanced Scorecard measures should be weighted and combined to produce a single summary measure of performance. Jensen (2001, p. 18) argues this point strongly, claiming that decision makers cannot make rational choices without some overall single dimensional objective to be maximized. Given a dozen or two dozen measures and no sense of the tradeoffs between them, the typical manager will be unable to behave purposefully, and the result will be confusion.

Without weighting the various measures to specify trade-offs among them, Jensen argues, there is no ‘‘balance’’ in the Balanced Scorecard and it does not yield a score that distinguishes winners from losers. Jensen maintains that a scorecard without weights and a composite score functions more like a dashboard or instrument panel, although he suggests it can still be useful for communicating a large amount of information within the firm. In practice, however, managers may find value both in the general framework of the Balanced Scorecard and in the hypothesized causal linkages (and perhaps in the process of constructing a scorecard). Moreover, explicitly weighting and combining Balanced Scorecard measures into a single metric – whether monetary as in Economic Value Added, or non-monetary as in the Intellectual Capital Index – would likely result in the kinds of criticisms leveled at traditional composite metrics, for example, earnings per share.

Additional Issues There are, of course, many additional issues concerning the mapping of value drivers to financial outcomes that are relevant to managers – and to management accounting researchers. One is the functional form of the relationship between value driver and outcome. Straightforward linear relationships are unlikely to adequately characterize all driver-outcome linkages. Ittner and Larcker (1998), for example, found that certain threshold levels of customer satisfaction must be achieved before either customer retention or revenue increases. Similarly, Anderson and Mittal (2000), in discussing the modeling of the Service-Profit Chain, point out that asymmetric and nonlinear relationships exist in many settings. Asymmetry implies that the impact of an increase in a causal variable differs in magnitude from that of an equivalent decrease, while non-linearity implies diminishing (or increasing) returns for adjacent changes of equivalent magnitude in a causal variable.

34

ROBERT H. ASHTON

Anderson and Mittal consider this issue in some detail for the driver of customer satisfaction, but it is likely to be relevant at other points along the Service-Profit Chain as well, and more generally for any causal model of value creation. Shields and Shields (2005) provide an insightful discussion of this and many other value driver/financial outcome relationships. These include the extent to which driver/outcome relationships are direct or indirect (i.e., operating through some intermediate value driver), the extent to which they are additive or interactive (i.e., conditional on the effect of another value driver), the timing of the relationship (contemporaneous or prospective), the duration of the value-enhancing effect, and the level of analysis that drives the effect (e.g., customer, product, organization and industry). While the relevance of such issues depends on the specific modeling application, it is essential to recognize that they apply to both statistical modeling and judgmental approaches to the linking of value drivers and outcomes.

RESEARCH ON VALUE-CREATION MODELS
The value-creation models described here are supported by a substantial amount of research that links measures of intangible value drivers to financial outcomes. Some of the research specifically addresses the management issues discussed above. Three broad streams of research are reviewed in this section. The first concerns the related categories of quality- and customer-oriented models, that is, research that addresses the Deming Management Method, the Baldrige Model, and the Service-Profit Chain. The second involves the large body of research relevant to the components of the Skandia Intellectual Capital Model. Finally, some of the rapidlygrowing body of research that investigates the adoption, use and performance effects of the Balanced Scorecard is reviewed.

Research on Quality and Customer Models Anderson et al.’s (1994) articulation of the theory underlying Deming’s work stimulated a few research studies on the Deming Management Method. For example, Anderson, Rungtusanatham, Schroeder, and Devaraj (1995) used performance data from the World-Class Manufacturing Project database (see Flynn, Schroeder, & Sakakibara, 1996; Flynn, Schroeder, Flynn,

Value-Creation Models for Value-Based Management

35

Sakakibara, & Bates, 1997) to test certain aspects of the theory. The data involved 41 manufacturing plants in the electronics, machinery, and transportation components industries. Within each industry were U.S. plants with world-class reputations, U.S. plants selected at large, and U.S. plants with Japanese ownership – to ensure variance in the application of quality management practices. Path analysis revealed that six of eight causal paths hypothesized by Anderson et al. (1994) were statistically significant, although the strength of the association was sometimes low. The study was replicated with Italian manufacturing companies with similar results (Rungtusanatham, Forza, Filippini, & Anderson, 1998). It was also replicated in a service context by Douglas and Fredendall (2004), using data from 193 U.S. hospitals studied earlier by Douglas and Judge (2001). Quality variables derived from the Deming Model were supplemented with additional variables derived from the Baldrige Health Care Criteria (Meyer & Collier, 2001), and the hospital’s financial performance was included as a performance dimension along with customer (patient) satisfaction. Douglas and Fredendall (2004) noted that the Baldrige Model is more comprehensive and includes more causal paths than the Deming Model as articulated by Anderson et al. (1994), a point made by other authors as well (e.g., Gale, 1994). While support was found for both the Deming and Baldrige perspectives, Douglas and Fredendall (2004) did find the Baldrige Model to be superior to the Deming Model in capturing the relationships among quality variables and performance. Other research has focused on the Baldrige Model more directly. For example, Flynn and Saladin (2001), using the World-Class Manufacturing Project database employed by Anderson et al. (1995), focused on 164 manufacturing plants in the electronics, machinery, and transportation components industries across five countries – U.S., U.K., Japan, Germany, and Italy. Path analysis revealed that 12 of 15 hypothesized causal paths were statistically significant. One intriguing result was that Leadership had a substantially stronger relationship with business results than any of the other Baldrige categories. Leadership has also emerged as the strongest variable in other empirical studies of the Baldrige Model (e.g., Wilson & Collier, 2000; Meyer & Collier, 2001; Pannirselvam & Ferguson, 2001), consistent with the decision of the model’s developers to place primary emphasis on this construct (Bell & Keys, 1998). Research with more of a customer focus has examined the various links in the Service-Profit Chain. These links, originally presented as ‘‘propositions’’ by Heskett et al. (1994), have been strongly supported by research. Zeithaml

36

ROBERT H. ASHTON

et al. (2001, p. 119) summarize a large body of research that connects improved service quality to increased firm profitability as follows:  service improvement efforts produce increased levels of customer satisfaction at the process or attribute level,  increased customer satisfaction at the process or attribute level leads to increased overall customer satisfaction,  higher overall service quality or customer satisfaction leads to increased behavioral intentions, such as greater repurchase intention,  increased behavioral intentions lead to behavioral impact, including repurchase or customer retention, positive word-of-mouth and increased usage, and  behavioral impact then leads to improved profitability and other financial outcomes. The studies on which these conclusions are based have examined a single link (or a small number of links) in the Service-Profit Chain in a piecemeal fashion.15 Loveman (1998), in contrast, appears to have studied all of the principal links in the model, but he examined them separately. Only recently have all of the posited relationships in the Service-Profit Chain been studied simultaneously in a single research setting (Kamakura, Mittal, de Rosa, & Mazzon, 2002). Kamakura et al. (2002) studied the Service-Profit Chain with data from more than 5,000 customers of more than 500 branches of a national bank in Brazil. A strategic analysis focused on the conceptual relationships within the Service-Profit Chain at the level of the bank, while an operational analysis focused on ways of implementing the model at particular branches. The goal was to operationalize the distinction between ‘‘strategic benchmarking’’ and ‘‘efficiency benchmarking’’ made earlier by Soteriou and Zenios (1999) – essentially, the distinction between identifying the key strategic links in the chain and exploiting the identified links operationally. Kamakura et al. supplemented variables specified in the Service-Profit Chain with some elements of the ‘‘Return on Quality’’ framework (Rust, Zahorik, & Keiningham, 1995), which emphasizes the financial cost of service-quality investments more explicitly than does the Service-Profit Chain. In the strategic analysis, a structural model with numerous hypothesized links was formulated, and the results were strongly supportive. Inferences were drawn concerning the relative amount of investment the bank should make in human versus technological resources for enhancing service quality. In the operational analysis, the results showed that particular branches must be successful at both operational efficiency and customer retention if they are to improve profitability.

Value-Creation Models for Value-Based Management

37

Research on Intangible (Intellectual Capital) Value Drivers A large number of studies have investigated the association between measures of intangible, or intellectual capital, value drivers and financial outcomes at both the firm and market levels. Ashton (2005) identifies almost 200 such studies conducted over the past 20 years from several disciplinary perspectives including accounting, marketing, operations, human resources, and information technology. Ashton organizes this body of research according to the four non-financial dimensions of the Skandia Business Navigator (see Fig. 9) – human, customer, process, and renewal and development. While none of the research was designed specifically to investigate the Skandia approach to intellectual capital, the results are clearly applicable to the Skandia Intellectual Capital Model. Moreover, because the Skandia Model is similar in many respects to other value-creation models discussed here, the results are also relevant to other models. This body of research examines multiple value drivers within each dimension of the Skandia Navigator. Research on human capital examines the performance effects of both individual manager characteristics (such as ability, experience, and certain personality traits) and systems of human resource practices (sometimes called ‘‘high-performance work systems’’) that involve rigorous selection procedures, management training and development activities, significant commitment to employee involvement, and performance-contingent incentive systems. Research on customer capital examines customer satisfaction (including customer retention and referrals) and drivers of satisfaction such as brand equity (including the establishment and extension of brand names and investments in advertising). Research on process capital examines the performance effects of quality-improvement initiatives (sometimes proxied by the winning of a major quality award) as well as the performance effects of investments in information technology. Finally, research on renewal and development capital examines R&D investments (both basic and applied) and patents (reflecting, e.g., new product development). The bottom line of the research reviewed by Ashton (2005) is that measures of intangible value drivers in all of these areas are positively associated with a wide array of firm- and market-level financial outcomes. Research on the Balanced Scorecard A considerable amount of research on the Balanced Scorecard has appeared. It includes case studies of single companies that have implemented scorecards, surveys of multiple companies in which comparative data are

38

ROBERT H. ASHTON

collected, analyses of archival data, and experimental studies (both lab and field) of scorecard adoption and use.16 Some of the research focuses on the incentive function of the scorecard, that is, incorporating non-financial measures into managers’ incentive compensation plans to motivate increased overall effort, or the reallocation of effort from other tasks, to the management of intangible value drivers. To the extent that non-financial measures are valid and contain information not captured in financial measures, the wisdom of including them in managers’ incentive plans is both intuitively apparent (e.g., Kerr, 1975) and supported by formal agencytheoretic research (e.g., Feltham & Xie, 1994). Most of the research, however, focuses only on the scorecard’s use as a strategic measurement system that provides information for decision-making beyond that provided by traditional financial measures. This section describes Balanced Scorecard research that addresses four important issues: (1) the extent to which the Balanced Scorecard has been adopted by companies, and its links with strategy and performance evaluation, (2) the use of both explicit and subjective weights for measures included in scorecards, (3) judgment biases that arise when a mixture of financial and non-financial measures is used in performance evaluation, and (4) whether non-financial scorecard measures are associated with financial outcomes. Balanced Scorecard Adoption While several surveys estimate that between 40 and 60% of large U.S. and U.K. companies have either adopted or experimented with the Balanced Scorecard for at least some of their business units (see Speckbacher, Bischof, & Pfeiffer, 2003), the adoption rate appears to be substantially lower (typically below 25%) in other European countries (Malmi, 2001; Gehrke & ´ Horvath, 2002; Speckbacher et al., 2003; Stemsrudhagen, 2004).17 Most studies find that the measures employed reflect the standard four perspectives (financial, customer, internal, and learning and growth), but applications involving as few as two or three perspectives (Speckbacher et al., 2003) and as many as seven perspectives (Kalagnanum, 2004) have been reported. Learning and growth measures appear much less frequently than measures in the other three perspectives, and are weighted less heavily when they do ´ appear (Gehrke & Horvath, 2002; Hoque & James, 2000; Ittner, Larcker, & Rajan, 1997; Malina & Selto, 2001; Maltz, Shenhar, & Reilly, 2003; Olson & Slater, 2002; Speckbacher et al., 2003; Stemsrudhagen, 2004). In addition to the common finding that scorecard adoption is more frequent in larger companies, adoption is mainly at the business-unit level (profit center, division, subsidiary) instead of the corporate or departmental

Value-Creation Models for Value-Based Management

39

level (Malmi, 2001; Speckbacher et al. 2003). Adoption has been found to be associated with the pursuit of differentiation as opposed to cost leadership strategies and with having won or been a finalist for a major quality award such as Baldrige (Ittner et al., 1997; Said, HassabElnaby, & Wier, 2003). Surveys also provide information about the linking of scorecard measures to each other and to strategy via causal models, and about the inclusion of non-financial scorecard measures in incentive compensation contracts. ´ Gehrke and Horvath (2002), Ittner, Larcker, and Meyer (2003a), Ittner, Larcker, and Randall (2003b), and Speckbacher et al. (2003) report that approximately 25% of their respondents claim to use models that causally link drivers to outcomes. Malmi (2001, p. 210), however, in a study based on semi-structured interviews, found that although most interviewees stated that scorecard measures were derived from strategy and were based on causeand-effect chains, further exploration revealed that ‘‘the claimed link between strategy and measures appeared weak for most companies.’’ Finally, ´ about half of the scorecard users in the Malmi (2001), Gehrke and Horvath (2002), and Speckbacher et al. (2003) surveys reported that scorecard measures were tied to compensation, often among other measures. Explicit versus Subjective Weights When scorecards are linked to performance evaluation, measures that represent the various value drivers must be weighted and combined, using either an explicit or subjective weighting scheme. Explicit weights have been the subject of research by Ittner et al. (1997) and Said et al. (2003), and subjective weights have been studied by Malina and Selto (2001), Ittner et al. (2003a), and Moers (2005). Ittner et al. (1997) studied CEO annual bonus plans involving financial and non-financial measures for 317 large companies in several industries. All of the plans had explicit weights for the measures employed. Thirty-six percent (114) of the companies used non-financial measures, and the mean weight placed on non-financial measures was 37.1% for the companies using non-financial measures (13.4% across all 317 companies). The mean numbers of financial and non-financial measures used were 1.7 and 2.3, respectively. In all, 13 types of non-financial measures were used, ranging from customer satisfaction (36.8% of the 114 companies) and product/service quality (21.0%) to leadership (5.2%) and innovation (2.6%). Significantly greater weight was placed on non-financial measures in companies that (1) were pursing differentiation as opposed to cost leadership strategies, (2) had won or been finalists for major quality awards such as Baldrige, (3) were more subject to regulation (utilities and telecommunications

40

ROBERT H. ASHTON

companies), and (4) had more noise in their financial performance measures (proxied by correlations between accounting returns and market returns and by time series variability in industry accounting returns). Said et al. (2003) followed up Ittner et al. (1997) by comparing companies that use both financial and non-financial measures in their annual bonus plans with companies that use only financial measures. The two samples were matched on size, industry, and past financial performance, and each included 1,441 firm-year observations over an 8-year period. Detailed information on the mean number of non-financial measures used and the mean weight placed on them was not reported. Consistent with the results of Ittner et al., however, the inclusion of explicitly weighted non-financial measures in annual bonus plans was positively associated with pursing differentiation instead of cost leadership strategies, winning or being a finalist for major quality awards, and regulation. Inconsistent with the earlier study, noise in financial performance measures was unrelated to the use of non-financial measures. Said et al. (2003) also examined the relation between use of nonfinancial measures and the length of both the product development cycle and the product life cycle (see Bushman, Indjejikian, & Smith, 1996), finding a positive relation with the former but not the latter. Malina and Selto (2001) described the development and use of a Balanced Scorecard by a single Fortune 500 company with annual sales of more than $6 billion. Subjective weights were employed to reflect management’s view of both the importance of the underlying drivers and the reliability of the numerical measures, and the weights were sometimes changed to mirror changes in management’s view. The scorecard involved 22 measures in the financial (3), customer (4), internal business processes (12), and learning and growth (3) categories.18 The subjective weights assigned to the measures within each category totaled 15% for financial, 40% for customer, 41% for internal business processes, and 4% for learning and growth. The percentage assigned to learning and growth measures (4%) had been reduced from 20% the previous year because management believed the measures were unreliable. Weights on a few measures in other categories were also reduced, with most of the weight reassigned to customer-related measures. Ittner et al. (2003a) studied the subjective weighting of Balanced Scorecard measures in a single financial services company – a U.S. retail bank. The company had previously replaced a bonus plan for branch managers that was based on a single financial measure (branch profitability) with a formula-based plan that also included operational and customer-based measures. This second plan was later replaced by a Balanced Scorecard containing six categories of measures. Three of the categories involved

Value-Creation Models for Value-Based Management

41

objective (quantitative) measures, and the other three involved subjective (qualitative) measures. Within each category, measures were weighed and combined to determine ‘‘category par scores,’’ and all six categories were weighted and combined subjectively (in contrast to the previous, formulabased plan) to determine an ‘‘overall par score’’ on which bonuses were based. The implicit weights placed on the various performance measures over a 15-quarter period were estimated by regression analysis; that is, the weights were those implied by the statistical relationships among branch performance on the scorecard measures, the par scores subjectively assigned to the branch managers, and the size of their quarterly bonuses. Four results are of particular interest. First, the overall par scores assigned to branch managers reflected greater weight on financial than on non-financial measures. Second, the category par scores often did not reflect the performance measures on which they presumably were based. Third, the entire set of scorecard measures explained only 50% of the variance in bonuses across branch managers. Finally, one scorecard measure that was common to all of the branches (overall customer satisfaction) was weighted very heavily, consistent with Lipe and Salterio’s (2000) finding (described later) that more weight is placed on common than on unique measures. Ittner et al. (2003a, p. 725) summarized their results:
We find that the subjectivity in the scorecard plan allowed superiors to reduce the ‘‘balance’’ in bonus awards by placing most of the weight on financial measures, to incorporate factors other than the scorecard measures in performance evaluations, to change evaluation criteria from quarter to quarter, to ignore measures that were predictive of future financial performance, and to weight measures that were not predictive of desired results.

Moers (2005) also studied the effects of subjectivity in the weighting of multiple performance measures – in the context of a single Dutch industrial firm in the maritime industry. This firm had replaced a fixed-wage compensation plan that was driven mostly by seniority with an incentive plan in which annual bonuses were based on a combination of objective and subjective measures. The objective measures related to quantitative dimensions of performance (e.g., whether a 5% reduction in absenteeism was achieved), while the subjective measures reflected qualitative assessments (e.g., whether a ‘‘good’’ use of resources or ‘‘adequate’’ planning was achieved). The link between individual performance and the bonus was subjectively determined in that direct superiors (1) allocated the size of the bonus between objective and subjective performance dimensions, (2) chose the number of measures within each dimension, and (3) provided ex post evaluations for each dimension. It was found that slightly more subjective than objective

42

ROBERT H. ASHTON

performance measures were used in determining bonuses (2.5 vs. 2.3 on average), a smaller average portion of the overall bonus awarded was based on subjective measures (46 vs. 54%), and the range in the number of performance measures used was 1–5 (1–6) for objective (subjective) measures. This latter result indicates that for some individuals only a single performance measure within a dimension was used. Judgment Biases Other Balanced Scorecard research has identified subtle but potentially important judgment biases that occur when a mixture of financial and nonfinancial measures is used for performance evaluation. For example, the Moers (2005) study, just described, found that the use of multiple objective measures (as opposed to a single objective measure) and the use of subjective measures were associated with two commonly observed biases in performance ratings – leniency and compression. Leniency (higher ratings) can result in increased compensation costs due to higher bonuses, while compression (less differentiation) can make promotion decisions more problematic. Both biases are traceable in part to the greater discretion in performance assessments that is allowed by multiple, and by subjective, measures. A series of experimental studies initiated by Lipe and Salterio (2000) has investigated a judgment bias – the ‘‘common measures bias’’ – that seems especially relevant in the Balanced Scorecard setting. Lipe and Salterio (2000) examined the effects of scorecard measures on evaluations of the performance of divisional managers. Some of the measures were common across divisions, while others were unique to a particular division, and the study focused on the relative weights placed by the evaluators on these two types of measures. One of the assumed strengths of the Balanced Scorecard when adopted at the business-unit level concerns the inclusion of measures that uniquely reflect the strategy and goals of the business unit. Although each business unit develops its own scorecard measures, some measures are likely to be common across all units and other measures are likely to be unique to a particular unit. If decision makers faced with both common and unique measures place more weight on common measures (as suggested by earlier research in the judgment/decision making literature), managers who evaluate multiple units may underuse or ignore the unique measures designed for each unit. Moreover, if unique measures do not affect superiors’ ex post performance evaluations of subordinate managers, then subordinate managers may be unlikely to use unique measures in ex ante decision making. Consequently,

Value-Creation Models for Value-Based Management

43

since unique measures are often leading and non-financial indicators while common measures are often lagging and financial indicators, managers may pay insufficient attention to leading and non-financial measures, possibly defeating a key purpose of implementing the Balanced Scorecard (i.e., expanding the set of measures that managers use). Participants in the study (MBA students) assumed the role of a senior executive who evaluated two divisions of a hypothetical retail firm – based on Kaplan and Norton’s (1996c) Kenyon Stores example – that used both common and unique measures. They were provided a 16-measure scorecard for each division, with each of the four scorecard categories including two common measures and two unique measures. As expected, the common measures had much stronger effects than the unique measures on the performance evaluation of the two divisional managers.19 Several studies have followed up the Lipe and Salterio (2000) results using modified versions of their task. Essentially, these studies have sought to ‘‘debias’’ the common measures bias by emphasizing causal linkages among scorecard measures (Banker, Chang, & Pizzini, 2004), highlighting accountability for one’s performance evaluations (Libby, Salterio, & Webb, 2004), providing an assurance report on the scorecard measures (Libby et al., 2004), disaggregating the performance evaluation task into rating and weighting activities followed by mechanical aggregation (Roberts, Albright, & Hibbets, 2004), and providing training in designing scorecards (Dilla & Steinbart, 2005). Banker et al. (2004) provided some of their participants (MBA students) with a brief narrative description of the two divisions’ strategies, while others received this same description plus a simple Strategy Map (Kaplan & Norton, 2004b) depicting linkages among the four scorecard categories. Both unique and common measures were included, and some measures of each type were explicitly linked with the division’s strategy while others were not. Participants receiving Strategy Maps relied more heavily on linked than on nonlinked measures, whether common or unique, whereas participants receiving only the narrative descriptions relied equally on linked and non-linked measures. Further, participants receiving Strategy Maps relied more on linked unique measures than on common non-linked measures, while there was no difference in reliance on unique and common measures in the absence of a Strategy Map. Thus, Banker et al. (2004) demonstrated that the common measures bias can be debiased when strategic linkages are made salient. Libby et al. (2004) also found that the common measures bias can be debiased. Relying on the notion that judgment biases typically have either effort-related or data-related sources (Kennedy, 1993, 1995), Libby et al.

44

ROBERT H. ASHTON

investigated one potential debiaser thought to be relevant to each source. If the tendency to focus on common measures and to ignore unique measures results from greater difficulty of evaluating unique measures (requiring greater cognitive effort), an accountability requirement that induces greater effort might be effective in mitigating the bias. On the other hand, if the common measures bias results from a perception that the measures unique to each division are of lower quality than common, companywide measures, an independent report that provides assurance on the relevance and reliability of the measures might mitigate the bias. Using Lipe and Saterio’s (2000) task and MBA students as participants, Libby et al. manipulated the requirement for participants to provide written justification for their performance evaluations and the provision of an assurance report on the scorecard measures. Both accountability and assurance mitigated the common measures bias. Roberts et al. (2004) modified the Lipe and Salterio (2000) task by having MBA-student participants rate the performance of each manager on each of the 16 scorecard items, multiply these ratings by a set of pre-determined weights, and sum to get a mechanically aggregated rating for each manager. Participants also provided a separate overall rating of each manager’s performance. (In making the overall ratings, participants were not strictly bound by their mechanically aggregated ratings, but the results showed that overall ratings and mechanically aggregated ratings correlated 0.74 for one division and 0.84 for the other.) While the pre-specified weights were allocated equally (25%) across the four scorecard categories, 64% of the total weight was assigned to the eight unique measures and only 36% to the eight common measures. Therefore, it is perhaps not surprising that the common measures bias was eliminated with the mechanical aggregation. Dilla and Steinbart (2005) investigated whether training and experience with scorecards could debias the common measures effect, also using a modified version of Lipe and Salterio’s (2000) task. They provided classroom instruction and individual and team-based practice in designing scorecards to undergraduate students who had no previous experience with scorecards. The principal finding was that both common and unique measures were used, although substantially greater emphasis was still placed on common measures. Association with Financial Outcomes A major issue in research on the Balanced Scorecard is whether scorecard adoption and use are associated with superior financial performance, either contemporaneously or prospectively, on key accounting- and market-based

Value-Creation Models for Value-Based Management

45

outcomes. A cross-sectional study by Said et al. (2003) and time series studies by Davis and Albright (2004) and Banker, Potter, and Srinivasan (2000) have investigated this issue. The Said et al. (2003) study, described earlier, which compared companies that use both financial and non-financial measures in their annual bonus plans with companies that use only financial measures, examined both accountingbased (ROA) and market-based (share return) performance measures. Both types of measures were calculated, on both a contemporary (same year) and prospective (1 year later) basis, resulting in four key comparisons between companies using only financial measures and those using both financial and non-financial measures. All four comparisons were directionally consistent with the hypothesis that companies using both types of measures would outperform those using only financial measures, although the effect based on ROA in the contemporaneous test was not statistically significant. Thus, 1-year-ahead accounting-based and market-based measures, as well as the contemporaneous market-based measure, indicated significantly better performance when non-financial variables were included in annual bonus plans. Davis and Albright (2004) studied a single banking organization in the U.S. Nine of the firm’s 14 branches were included in the study, four of which implemented a Balanced Scorecard during the time period of the study and five of which did not. Each branch used a set of nine key financial performance measures that pre-dated the introduction of the scorecard, and these nine measures were subjectively weighted and combined by management to form a single ‘‘composite key performance measure’’ for each branch. Performance on the nine measures determined each branch’s annual bonus level. The results indicated that, during the 2-year period following introduction of the scorecard, financial performance improved significantly for the scorecard branches but not for the non-scorecard branches, and that post-scorecard-introduction performance was significantly greater for the scorecard branches than for the non-scorecard branches. Banker et al. (2000) analyzed archival performance data for a 6-year period from 18 hotels managed by a national hotel chain. During this period corporate management introduced a new incentive plan for key managers at each hotel. While the previous plan had been based entirely on financial measures, the new plan – which was based on principles underlying the Service-Profit Chain – included both financial (operating profit) and nonfinancial (customer satisfaction) measures. The customer satisfaction measure had been reported prior to the introduction of the new incentive plan, but it had not been used as a basis for incentive compensation. Analysis of the time series of customer satisfaction and operating profit showed that

46

ROBERT H. ASHTON

both satisfaction and future profitability increased after adoption of the new incentive plan. Further analysis revealed that the profitability effect resulted from increased revenues rather than lower costs, and was driven by increased occupancy rather than increased room rates. Noting that the non-financial (customer satisfaction) measure was reported for information purposes before the new incentive plan was adopted, Banker et al. (2000, p. 89) considered the question of why the hotel managers had ignored it: ‘‘Our study raises some interesting questions. Hotel managers received a substantial annual bonus based on profit before the change in the incentive plan, and there was only a 6-month lag between customer satisfaction and profit. Why then did they not exert the appropriate effort to improve customer satisfaction, and why did the incorporation of the customer-satisfaction measures in compensation lead to improvements in both customer satisfaction and profit?’’ Referring to the complex interplay among the managers’ knowledge of customer satisfaction, the structure of their incentive plans, and hotel performance, Banker et al. (2000, pp. 89–90) reasoned that although hotel managers were aware of the strategic importance of customer satisfaction for financial performance, they did not know either the timing or the magnitude of this relation. Without such knowledge, managers did not recognize the true benefit of allocating more effort and resources to improve customer satisfaction, and did not do so until the change in the compensation plan that focused their attention on improving the customer-satisfaction measures.

They further suggested that when managers gain better knowledge of issues such as the timing and magnitude of the association between non-financial measures and financial results, the use of formal incentives based on nonfinancial measures may not be essential. In other words, the information role of non-financial measures may be sufficient to achieve improved financial results. Further Research While the amount of research that is relevant to the value-creation models discussed here is sizable, additional research on the use and effects of such models would be valuable. Concerning the Balanced Scorecard, substantial evidence exists about the frequency of scorecard adoption and the experiences of specific companies with respect to implementation and use,20 but much less is known about how companies view causal linkages among nonfinancial and financial measures or about the incorporation of non-financial measures in incentive compensation contracts – including the circumstances

Value-Creation Models for Value-Based Management

47

under which explicit or subjective weights are employed. Concerning the other value-creation models, they often have been studied in piecemeal fashion (e.g., specific links in the Service-Profit Chain) but much less often in a more comprehensive ‘‘structural equations’’ sense (e.g., the Baldrige and Deming models). And while much existing research can be interpreted within the framework of the Skandia Model, studies have not investigated this model directly – in contrast to all of the other models discussed. The relative strengths and weaknesses of the different models, and the specific managerial applications for which particular models are better or worse suited, have not been explored. Future research might benefit from combining the perspectives on value creation that are characteristic of the different models. For example, the Service-Profit Chain tends to emphasize customer-related drivers and service quality, while the Baldrige and Deming models tend to emphasize processrelated drivers and product quality. Research that combines the customer and quality perspectives of these models might result in a more penetrating analysis of value creation than research that derives from a single perspective. Similarly, research conducted from a Balanced Scorecard perspective might benefit from elaborating the customer dimension of the scorecard with insights from the Service-Profit Chain, as well as from elaborating the internal dimension of the scorecard with insights from the quality-oriented models. And the broad perspective provided by Skandia’s approach to intellectual capital, which is compatible with all of the other models, can likely enhance the understanding of value creation that results from application of the other models. Stated differently, since the Baldrige, Deming, Skandia, Service-Profit Chain, and Balanced Scorecard models reflect somewhat different (and incomplete) perspectives on value creation, gains will likely result from carefully combining key elements of multiple perspectives. Research could also focus on value-creation models that have been proposed more recently than those analyzed here. The Action-Profit Linkage Model (Epstein, Kumar, & Westbrook, 2000), a more elaborate version of the Service-Profit Chain which considers employee actions, customer actions, and economic impact, relates more to customer segments than to individual customers. The Value Dynamics Framework (Boulton, Libert, & Samek, 2000), which also emphasizes employees and customers as well as important aspects of the Skandia Model, is oriented toward ‘‘designing a business model [and] mastering risk’’ (p. 249). The Value Explorer Model (Andriessen & Tissen, 2000), while similar to the Skandia Model, emphasizes estimation of the financial contribution to firm value of the various core competences within a firm. The Value Chain Scorecard (Lev, 2001) focuses on three

48

ROBERT H. ASHTON

(largely sequential) stages of value creation – discovery/learning, implementation, and commercialization – with a strong emphasis on technology, research and development, and intellectual property. The Performance Prism (Neely, Adams, & Kennerley, 2002), which explicitly builds on more established models including Baldrige, Balanced Scorecard, and Skandia, considers what the firm wants from its stakeholders in addition to what the firm must do for them. To what extent do models such as these, which clearly incorporate key elements of earlier models, provide incremental insights into value creation? To what extent might they better guide value-based management activities in specific settings, for example, for certain types of firms or industries? Can selected elements or perspectives of ‘‘second generation’’ models be profitably blended with those of more established models? Other research could consider the potential usefulness of value-creation models for guiding external disclosures. Much interest has been shown in expanding the traditional external reporting model (e.g., American Institute of Certified Public Accountants (AICPA) (1994); FASB, 2001; Wallman, 1995, 1996). A recent analysis raised the question of whether firms should report non-traditional disclosures using an ‘‘integrated financial/non-financial framework’’ (Maines et al., 2002): It was observed that ‘‘an integrated framework could disclose specific non-financial performance measures and provide a description of the firm’s business model in the context of these measures and how these measures map into firm value’’ (Maines et al., 2002, p. 357, emphasis added). Blair and Wallman (2000) suggest that the lack of business models that describe the use to which such disclosures might be put is a significant hindrance to the development of expanded external disclosure practices. The value-creation models described in this paper are ‘‘integrated financial/non-financial frameworks’’ that suggest possible linkages among non-financial measures and financial outcomes, and they potentially could serve as frameworks for expanded external disclosure of non-financial information. Their value in this regard is an important area for research. Moreover, it seems apparent that the credibility of non-traditional disclosures such as those contemplated by value-creation models would be enhanced by some form of independent third-party association. In fact, it seems likely that third-party association concerning the reliability and/or relevance of such disclosures will be necessary if they are to be taken seriously by investors, thus creating new opportunities for attestation and assurance services and for related research.21 Further, new attestation and assurance applications are likely to require new skills on the part of assurors and new assurance standards for conducting and reporting the results of such engagements. Finally, new audit and review methodologies are likely to

Value-Creation Models for Value-Based Management

49

be needed, as current methodologies are strongly oriented toward examination of traditional financial statements and are unlikely to be adequate for examining long-run value-creation potential. Recently, one of the Big Four firms, KPMG, has actively developed a new methodology, the KPMG Business Measurement Process, that is specifically designed for understanding how firms create value through the transformation of tangible and intangible resources (Bell, Marrs, Solomon, & Thomas, 1997; Bell, Peecher, & Solomon, 2002, 2005). All of these areas – assuror skills, assurance standards and methodologies, and the expanded scope of attestation and assurance services – would likely benefit from research that reflects perspectives of the value-creation models discussed in this paper.

CONCLUSION
Value-creation models that have been proposed for supporting value-based management are discussed. Value-based management involves defining and implementing strategies for long-run shareholder value creation and aligning management systems and activities, as well as financial and non-financial performance measures, for value creation. Accordingly, management accounting research conducted within the value-based management paradigm focuses on the identification, measurement, management, and reporting of tangible and intangible value drivers, and generally involves some model of value creation that guides the identification and use of value drivers and related performance measures. Because each value-creation model involves a particular (and incomplete) perspective on the drivers and measures considered necessary for long-run value creation, it is essential to understand the differences among the principal models that have been proposed, and because substantial overlap exists among models it is important to understand their similarities as well. It is also useful to recognize that measures reflecting many of the intangible value drivers included in value-creation models are supported by significant streams of research. Thus, much of this paper is concerned with describing value-creation models, their similarities and differences, and the research evidence that supports them. Critical management considerations that value-creation models entail are also discussed, including linking measures of intangible value drivers to each other and to financial outcomes and whether such measures should be explicitly weighted and combined to form a ‘‘value index.’’ The discussion of these issues is necessarily preliminary

50

ROBERT H. ASHTON

given the current state of knowledge and experience in these areas, and is intended primarily to stimulate further discussion and research. The scope of research that could usefully address the role of valuecreation models in value-based management is extremely broad – both in terms of topics and methods – and includes research on the formulation, validation, implementation, and effectiveness of value-creation models using archival and field-based approaches among others. While a substantial amount of research already exists on management aspects of value-creation ` models, the research nevertheless is relatively limited vis-a-vis the range of intangible value drivers included in existing models and the complexity of the value-creation processes and relationships in which those models are embedded. The opportunities for further progress in understanding the role of value-creation models in value-based management are enormous, at both research and practical levels.

NOTES
1. Similarly, Crosby (1979) proposed a set of 14 Steps (which were oriented somewhat more toward top management than were Deming’s 14 Points), and Juran (1989) advocated the Juran Trilogy, the three managerial processes of quality planning, quality control, and quality improvement – analogous to the financial management processes of financial planning, financial control, and financial improvement. March (1986) and Ross (1999) provide comparative evaluations of the approaches of Deming, Juran, and Crosby. 2. All of the panel members had been actively involved in the implementation of Deming’s methods, and some had conducted research and written about quality management in general and the Deming Management Method in particular. 3. Of course, the weighting scheme in Fig. 5 is not necessarily relevant when the Baldrige Model is used as a value-creation framework. 4. Skandia’s efforts are described in several publications (e.g., Edvinsson, 1997; Edvinsson & Malone, 1997; Roos & Roos, 1997; Roos et al., 1998; Roos & Jacobsen, 1999, and Roos, Bainbridge, & Jacobsen, 2001), and a sizable literature concerning the approach has appeared (e.g., Bontis, Dragonetti, Jacobsen, & Roos, 1999; Choo & Bontis, 2002; Larsen, Bukh, & Mouritsen, 1999; Lynn, 1998a, 1998b; Mouritsen, 1998; Mouritsen, Bukh, Larsen, & Johansen, 2002; Mouritsen, Johansen, Larsen, & Bukh, 2001; Mouritsen, Larsen, & Bukh, 2001; OECD, 2000; Petty & Guthrie, 2000; Stewart, 1997, 2001; Sullivan, 1998, 2002; and The Conference Board, 1997). 5. These developments are described in more detail by Ashton (2005). 6. More recent expositions of the Balanced Scorecard refer to the internal business perspective as the ‘‘internal business processes’’ perspective or simply the ‘‘internal’’ perspective, and to the innovation and learning perspective as the ‘‘learning and growth’’ perspective (e.g., Kaplan & Norton, 1996c).

Value-Creation Models for Value-Based Management

51

7. The TQM activities that Hoque (2003) considers are (1) internal process improvement and manufacturing innovation, (2) benchmarking, (3) a zero-defects culture, (4) supplier relations, (5) customer relations, (6) employee training, (7) an open, less bureaucratic culture, and employee empowerment, and (8) executive commitment and management competence. 8. A measurement system based on Drucker’s ideas was incorporated into General Electric’s ‘‘Measurements Project’’ in the 1950s (Greenwood, 1974). 9. Bourguignon et al. (2004) describe four ways in which the Tableau de Bord differs from the Balanced Scorecard. First, the Balanced Scorecard, building on the framework of Porter (1980, 1985), works ‘‘from the outside in’’ (from customers to internal processes), while the Tableau de Bord works from both the outside in and the inside out, where the latter reflects a resources/core competences perspective (e.g., Barney, 1991; Prahalad & Hamel, 1990). Second, while the Balanced Scorecard assumes a generic cause-and-effect performance model (from Learning and Growth to Internal Processes to Customer to Financial perspectives), the Tableau de Bord assumes no systematic overall links among various perspectives. Third, the dominant approach of the Balanced Scorecard is top-down (cascading management’s objectives to lower organizational levels), while deployment of the Tableau de Bord involves more interaction and negotiation among different organizational levels. Finally, the Balanced Scorecard emphasizes incentives, rewards, and accountability, while the Tableau de Bord focuses more on information and learning. 10. Development of the Balanced Scorecard and its application in several companies is described by Kaplan and Norton in three books (Kaplan & Norton, 1996c, 2001a, 2004b) and a series of articles (Kaplan & Norton, 1992, 1993, 1996a, 1996b, 1996d, 2000, 2001b, 2001c, 2004a). 11. Oliva, Day, and DeSarbo (1987) had earlier proposed a ‘‘strategy map’’ for describing how performance measures, competitive tactics, and competitors’ performance are related. However, the application was at the industry level instead of the firm level, and the Oliva et al. strategy map was simply a graphical depiction in which individual firms were located relative to competitors on measures that were particularly relevant to their industry, for example, profitability, growth, and market share. 12. The distinction between valuation approaches and measurement approaches is consistent with Ijiri’s (1995) distinction between managing capital and managing resources. Ijiri maintains that because capital managers such as top executives and board members raise capital and allocate it among projects while resource managers such as division heads operate projects to achieve particular financial objectives, the two types of managers have different orientations and different information needs. Since capital is abstract, aggregate, and homogeneous, capital managers need aggregate information that allows them to assess investment returns and risks. Since resources are concrete, disaggregate, and heterogeneous, resource managers need disaggregate information that allows them to plan and execute the use of resources. Thus, in Ijiri’s framework valuation approaches are likely to better satisfy the information needs of capital managers, while measurement approaches are likely to better satisfy the information needs of resource managers. 13. This application of the Strategy Map also clarifies that HR functional managers are expected to focus on more than just efficiency by linking the five strategic thrusts to the broad themes of both ‘‘minimizing HR costs’’ and ‘‘maximizing human

52

ROBERT H. ASHTON

capital,’’ which are shown in the financial dimension of the scorecard as drivers of shareholder value. More generally, it is intended to incentivize HR managers to attend to a particular set of value drivers, to view HR as a strategic asset, and to be better able to demonstrate to others the contributions of HR to firm performance. 14. An early discussion of self-correcting and self-reinforcing feedback loops in the context of management accounting systems is provided by Ashton (1976). 15. Specific references are provided by Zeithaml (1988, 2000), Zahorik and Rust (1992), Rust, Zahorik, and Keiningham (1994), Oliver (1997), and Zeithaml et al. (2001). 16. Case studies include Ahn (2001), Braam and Nijssen (2004), Butler, Letza, and Neale (1997), Ittner et al. (2003a), Kalagnanum (2004), Malina and Selto (2001), Moers (2005), Mooraj, Oyon, and Hostettler (1999), Papalexandris, Ioannou, and Prastacos (2004), Schneiderman (2001), and Vaivio (1999). Surveys include Gehrke ´ and Horvath (2002), Hoque and James (2000), Ittner et al. (2003b), Malmi (2001), Maltz et al., 2003, Olson and Slater (2002), Speckbacher et al. (2003), and Stemsrudhagen (2004). Archival studies include Ittner et al. (1997) and Said et al. (2003). Laboratory studies include Banker et al. (2004), Dilla and Steinbart (2005), Libby et al. (2004), Lipe and Salterio (2000, 2002), and Roberts et al. (2004). Field studies include Banker et al. (2000) and Davis and Albright (2004). 17. Countries included in these surveys are Austria, Finland, France, Germany, Italy, Norway, and Switzerland. 18. The categories actually used by the company were not precisely these four, so Malina and Selto (2001) reclassified a few of the measures. 19. Lipe and Salterio (2002) adapted their original task to examine the effects on performance evaluation of different ways of organizing scorecard measures. Two specific types of organization were investigated – organization via the four standard scorecard categories and an ‘‘uncategorized list’’ of measures. It was expected that multiple measures that are organized into each of the four standard categories would have less impact on performance evaluations than the same measures distributed across the four categories (‘‘uncategorized’’) because categorized measures would be perceived by evaluators as somewhat redundant (not independent of each other). This expectation was supported. 20. Much of this evidence comes from Kaplan and Norton’s books (Kaplan & Norton, 1996c, 2001a, 2004b), in addition to studies listed in footnote 16. 21. Professional standards in the U.S. consider attestation services to include not only the traditional audit of historical financial statements, but also examinations of, and the issuance of reports on, the reliability of other types of assertions that are the responsibility of another party. In contrast, assurance services are independent professional services that improve the quality of information for decision makers, and they include relevance-assurance as well as reliability-assurance services (American Institute of Certified Public Accountants (AICPA), 1997, 2001).

ACKNOWLEDGMENTS
This research was supported by KPMG under its Business Measurement Research Program. I am indebted to KPMG for their support, and

Value-Creation Models for Value-Based Management

53

particularly to Timothy B. Bell, Director of Assurance Research at KPMG, for support and for many valuable conversations on value creation. The paper has also benefited from the substantial input of Alison Ashton (Duke University), Ram Menon (KPMG) and Ira Solomon (University of Illinois).

REFERENCES
Ahn, H. (2001). Applying the Balanced Scorecard concept: An experience report. Long Range Planning, 34(August), 441–461. Allee, V. (2003). The future of knowledge: Increasing prosperity through value networks. New York, NY: Butterworth-Heinemann. American Institute of Certified Public Accountants (AICPA) (1994). Improving business reporting: A customer focus: Meeting the information needs of investors and creditors. New York, NY: AICPA. American Institute of Certified Public Accountants (AICPA). (1997). Report of the special committee on assurance services. www.aicpa.org/assurance/scas/index.htm American Institute of Certified Public Accountants (AICPA) (2001). Attestation standards: Revision and recodification. New York, NY: AICPA. Anderson, E. W., & Mittal, V. (2000). Strengthening the satisfaction-profit chain. Journal of Service Research, 3(November), 107–120. Anderson, J. C., Rungtusanatham, M., & Schroeder, R. G. (1994). A theory of quality management underlying the Deming management method. Academy of Management Review, 19(July), 472–509. Anderson, J. C., Rungtusanatham, M., Schroeder, R. G., & Devaraj, S. (1995). A path analytic model of a theory of quality management underlying the Deming management method: Preliminary empirical findings. Decision Sciences, 26(September–October), 637–658. Andriessen, D., & Tissen, R. (2000). Weightless wealth. London: Prentice Hall. Ashton, R. H. (1976). Deviation-amplifying feedback and unintended consequences of management accounting systems. Accounting, Organizations and Society, 1(November), 289–300. Ashton, R. H. (2005). Intellectual capital and value creation: A review. Journal of Accounting Literature, 24, 53–134. Baldrige National Quality Program (2004). 2003 Criteria for performance excellence. Gaithersburg, MD: National Institute of Standards and Technology. Banker, R. D., Chang, H., & Pizzini, M. J. (2004). The Balanced Scorecard: Judgmental effects of performance measures linked to strategy. The Accounting Review, 79(January), 1–23. Banker, R. D., Potter, G., & Srinivasan, D. (2000). An empirical investigation of an incentive plan that includes nonfinancial performance measures. The Accounting Review, 75(January), 65–92. Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(March), 99–120. Barney, J. B. (1995). Looking inside for competitive advantage. Academy of Management Executive, 9(November), 49–61. Bartlett, C. A., & Mahmood, T. (1996). Skandia AFS: Developing intellectual capital globally. HBS Case 9-396-412. Boston, MA: Harvard Business School.

54

ROBERT H. ASHTON

Baum, G., Ittner, C., Larcker, D., Low, J., Siesfeld, T., & Malone, M. S. (2000). Introducing the new value creation index. Forbes ASAP, 165(April 3), 140–143. Becker, B. E., Huselid, M. A., & Ulrich, D. (2001). The HR scorecard: Linking people, strategy, and performance. Boston, MA: Harvard Business School. Beischel, M. E., & Smith, K. R. (1991). Linking the shop floor to the top floor. Management Accounting, 73(October), 25–29. Bell, R., & Keys, B. (1998). A conversation with Curt W. Reimann on the background and future of the Baldrige award. Organizational Dynamics, 26(Spring), 51–61. Bell, T., Marrs, F., Solomon, I., & Thomas, H. (1997). Auditing organizations through a strategic-systems lens: The KPMG business measurement process. Montvale, NJ: KPMG. Bell, T. B., Peecher, M. E., & Solomon, I. (2002). The strategic-systems approach to auditing. In: T. B. Bell & I. Solomon (Eds), Cases in strategic-systems auditing (pp. 1–34). Montvale, NJ: KPMG. Bell, T. B., Peecher, M. E., & Solomon, I. (2005). The 21st Century Public Company Audit: Conceptual elements of KPMG’s Global audit methodology. Montvale, NJ: KPMG. Bellis-Jones, R. (1989). Customer profitability analysis. Management Accounting, 67, 26–28. Bemowski, K., & Stratton, B. (1995). How do people use the Baldrige award criteria?. Quality Progress, 28(May), 43–47. Berliner, C., & Brimson, J. A. (Eds) (1988). Cost management for today’s advanced manufacturing. Boston, MA: Harvard Business School. Blair, M. M., & Wallman, S. M. H. (Eds) (2000). Unseen wealth: Report of the Brookings task force on understanding intangible sources of value. Washington, DC: The Brookings Institution. Blattberg, R. C., & Deighton, J. (1996). Manage marketing by the customer equity test. Harvard Business Review, 74(July–August), 136–144. Bontis, N., Dragonetti, N. C., Jacobsen, K., & Roos, G. (1999). The knowledge toolbox: A review of the tools available to measure and manage intangible resources. European Management Journal, 17(August), 391–402. Boulton, R. E. S., Libert, B. D., & Samek, S. M. (2000). Cracking the value code: How successful businesses are creating wealth in the new economy. New York, NY: Harper Business. Bourguignon, A., Malleret, V., & Norreklit, H. (2004). The American Balanced Scorecard versus the French tableau de Bord: The idealogical dimension. Management Accounting Research, 15(June), 107–134. Braam, G. J. M., & Nijssen, E. J. (2004). Performance effects of using the Balanced Scorecard: A note on the Dutch experience. Long Range Planning, 37(August), 335–349. Brown, M. G. (2003). Baldrige award winning quality: How to interpret the Baldrige criteria for performance excellence (12th ed.). Milwaukee, WI: ASQ Quality Press. Bushman, R. M., Indjejikian, R. J., & Smith, A. (1996). CEO compensation: The role of individual performance evaluation. Journal of Accounting and Economics, 21(April), 161–193. Butler, A., Letza, S. R., & Neale, B. (1997). Linking the Balanced Scorecard to strategy. Long Range Planning, 30(April), 242–253. Choo, C. W., & Bontis, N. (Eds) (2002). The strategic management of intellectual capital and organizational knowledge. New York, NY: Oxford University Press. Crosby, P. B. (1979). Quality is free: The art of making quality certain. New York, NY: Penguin.

Value-Creation Models for Value-Based Management

55

Cua, K. O., McKone, K. E., & Schroeder, R. G. (2001). Relationships between implementation of TQM, JIT, and TPM and manufacturing performance. Journal of Operations Management, 19(November), 675–694. Danish Ministry of Science and Technology. (2003). Intellectual Capital Statements – The New Guideline. Datar, S., & Epstein, M. J. (2001). Verizon Communications, Inc.: Implementing a human resources Balanced Scorecard. HBS Case 9-101-102. Boston, MA: Harvard Business School. Davis, S., & Albright, T. (2004). An investigation of the effect of Balanced Scorecard implementation on financial performance. Management Accounting Research, 15(June), 135–153. Deming, W. E. (1982). Quality, productivity, and competitive position. Cambridge, MA: MIT Center for Advanced Engineering Study. Deming, W. E. (1986). Out of the crisis. Cambridge, MA: MIT Center for Advanced Engineering Study. Dempsey, S. J., Gatti, J. F., Grinnell, D. J., & Cats-Baril, W. L. (1997). The use of strategic performance variables as leading indicators in financial analysts’ forecasts. Journal of Financial Statement Analysis, 2(Summer), 61–79. Dilla, W. N., & Steinbart, P. J. (2005). Relative weighting of common and unique Balanced Scorecard measures by knowledgeable decision makers. Behavioral Research in Accounting, 17, 43–53. Dixon, J. R., Nanni, A. J., & Vollmann, T. E. (1990). The new performance challenge: Measuring operations for world-class competition. Homewood, IL: Dow Jones-Irwin. Douglas, T. J., & Fredendall, L. D. (2004). Evaluating the Deming management model of total quality in services. Decision Sciences, 35(Summer), 393–422. Douglas, T. J., & Judge, W. Q. (2001). Total quality management implementation and competitive advantage: The role of structural control and exploration. Academy of Management Journal, 44(February), 158–169. Drucker, P. F. (1954). The Practice of Management. New York, NY: Harper and Brothers. Dyer, J. H., & Singh, H. (1998). The relational view: Cooperative strategy and sources of interorganizational competitive advantage. Academy of Management Review, 23(October), 660–679. Eccles, R. G. (1991). The performance measurement manifesto. Harvard Business Review, 69(January–February), 131–137. Eccles, R. G., & Pyburn, P. J. (1992). Creating a comprehensive system to measure performance. Management Accounting, 74(October), 41–44. Edvinsson, L. (1997). Developing intellectual capital at Skandia. Long Range Planning, 30(June), 366–373. Edvinsson, L., & Malone, M. S. (1997). Intellectual capital: Realizing your company’s true value by finding its hidden roots. New York, NY: HarperCollins. Ehrbar, A. (1998). EVA: The real key to creating wealth. New York, NY: Wiley. Epstein, M. J., Kumar, P., & Westbrook, R. (2000). The drivers of customer and corporate profitability: Modeling, measuring, and managing the causal relationships. Advances in Management Accounting, 9, 43–72. Epstein, M. J., & Manzoni, J. F. (1997). The Balanced Scorecard and tableau de Bord: Translating strategy into action. Management Accounting, 79(August), 28–36. Epstein, M. J., & Manzoni, J. F. (1998). Implementing corporate strategy: From tableau de Bord to Balanced Scorecard. European Management Journal, 2(April), 190–203.

56

ROBERT H. ASHTON

Epstein, M. J., & Manzoni, J. F. (Eds) (2002). Performance measurement and management control: A compendium of research. London: Elsevier. Epstein, M. J., & Manzoni, J. F. (Eds) (2004). Performance measurement and management control: Superior organizational performance. London: Elsevier. Epstein, M. J., & Westbrook, R. A. (2001). Linking actions to profits in strategic decision making. Sloan Management Review, 42(Spring), 39–49. Ernst & Young LLP. (1997). Measures that Matter. www.businessinnovation.ey.com Feltham, G. A., & Xie, J. Z. (1994). Performance measure congruity and diversity in multi-task principal-agent relations. The Accounting Review, 69(July), 429–453. Financial Accounting Standards Board (2001). Improving business reporting: Insights into enhancing voluntary disclosures. Norwalk, CT: FASB. Fisher, J. (1992). Use of nonfinancial performance measures. Journal of Cost Management, 6(Spring), 31–38. Flynn, B. B., & Saladin, B. (2001). Further evidence on the validity of the theoretical models underlying the Baldrige criteria. Journal of Operations Management, 19(November), 617–652. Flynn, B. B., Schroeder, R. G., Flynn, E. J., Sakakibara, S., & Bates, K. A. (1997). World-class manufacturing project: Overview and selected results. International Journal of Operations and Production Management, 17, 671–685. Flynn, B. B., Schroeder, R. G., & Sakakibara, S. (1996). The relationship between quality management practices and performance: Synthesis of findings from the world class manufacturing project. In: D. G. Fedor & S. Ghosh (Eds), Advances in the management of quality. Greenwich, CT: JAI Press. Foss, N. J. (Ed.) (1997). Resources, firms, and strategies: A reader in the resource-based perspective. New York, NY: Oxford University Press. Foster, G., & Gupta, M. (1994). Marketing, cost management and management accounting. Journal of Management Accounting Research, 6, 43–77. Foster, G., Gupta, M., & Sjoblom, L. (1996). Customer profitability analysis: Challenges and new directions. Journal of Cost Management, 10(Spring), 5–17. Funk, K. (2003). Sustainability and performance. Sloan Management Review, 44(Winter), 65–70. Gale, B. T. (1994). Managing customer value: Creating quality and service that customers can see. New York, NY: The Free Press. Garvin, D. A. (1987). Competing on the eight dimensions of quality. Harvard Business Review, 65(November–December), 101–109. Garvin, D. A. (1991). How the Baldrige Award really works. Harvard Business Review, 69(November–December), 80–93. ´ Gehrke, I., & Horvath, P. (2002). Implementation of performance measurement: A comparative study of French and German organizations. In: M. J. Epstein & J. F. Manzoni (Eds), Performance measurement and management control: A compendium of research (pp. 159–180). London: Elsevier. Goldman Sachs Global Equity Research. (2000). Skandia Group. London: Goldman Sachs (October 24). Greenwood, R. G. (1974). Managerial decentralization: A study of the general electric philosophy. Lexington, MA: D. C. Heath. Guilding, C. E., & McManus, L. (2002). The incidence, perceived merit and antecedents of customer accounting: An exploratory note. Accounting, Organizations and Society, 27(January–March), 45–59.

Value-Creation Models for Value-Based Management

57

Heskett, J. L., Jones, T. O., Loveman, G. W., Sasser, W. E., & Schlesinger, L. A. (1994). Putting the service-profit chain to work. Harvard Business Review, 72(March–April), 164–174. Heskett, J. L., Sasser, W. E., & Schlesinger, L. A. (1997). The service profit chain: How leading companies link profit and growth to loyalty, satisfaction, and value. New York, NY: The Free Press. Hoque, Z. (2003). Total quality management and the Balanced Scorecard approach: A critical analysis of their potential relationships and directions for research. Critical Perspectives in Accounting, 14(July), 553–566. Hoque, Z., & James, W. (2000). Linking Balanced Scorecard measures to size and market factors: Impact on organizational performance. Journal of Management Accounting Research, 12, 1–17. Ijiri, Y. (1995). Segment statements and informativeness measures: Managing capital vs. managing resources. Accounting Horizons, 9(September), 55–67. Ittner, C. D., & Larcker, D. F. (1998). Are nonfinancial measures leading indicators of financial performance? An analysis of customer satisfaction. Journal of Accounting Research, 36(Supplement), 1–35. Ittner, C. D., & Larcker, D. F. (2001). Assessing empirical research in managerial accounting: A value-based management perspective. Journal of Accounting and Economics, 32(December), 349–410. Ittner, C. D., Larcker, D. F., & Meyer, M. W. (2003a). Subjectivity and the weighting of performance measures: Evidence from a Balanced Scorecard. The Accounting Review, 78(July), 725–758. Ittner, C. D., Larcker, D. F., & Rajan, M. (1997). The choice of performance measures in annual bonus contracts. The Accounting Review, 72(April), 231–255. Ittner, C. D., Larcker, D. F., & Randall, T. (2003b). Performance implications of strategic performance measurement in financial services firms. Accounting, Organizations and Society, 28(October–November), 715–741. Jensen, M. C. (2001). Value maximization, stakeholder theory, and the corporate objective function. Journal of Applied Corporate Finance, 14(Fall), 8–21. Johnson, H. T., & Kaplan, R. S. (1987). Relevance lost: The rise and fall of management accounting. Boston, MA: Harvard Business School. Juran, J. M. (1989). Juran on leadership for quality: An executive handbook. New York, NY: The Free Press. Kalafut, P. C., & Low, J. (2001). The value creation index: Quantifying intangible value. Strategy & Leadership, 29(September–October), 9–15. Kalagnanum, S. S. (2004). The adoption of the Balanced Scorecard in government-owned corporations. In: M. J. Epstein & J. F. Manzoni (Eds), Performance measurement and management control: Superior organizational performance (pp. 407–425). London: Elsevier. Kamakura, W. A., Mittal, V., de Rosa, F., & Mazzon, J. A. (2002). Assessing the service-profit chain. Marketing Science, 21(Summer), 294–317. Kaplan, R. S. (1998). Innovation action research: Creating new management theory and practice. Journal of Management Accounting Research, 10, 89–118. Kaplan, R. S., & Narayanan, V. G. (2001). Measuring and managing customer profitability. Journal of Cost Management, 15(September–October), 5–15. Kaplan, R. S., & Norton, D. P. (1992). The Balanced Scorecard: Measures that drive performance. Harvard Business Review, 70(January–February), 71–79.

58

ROBERT H. ASHTON

Kaplan, R. S., & Norton, D. P. (1993). Putting the Balanced Scorecard to work. Harvard Business Review, 71(September–October), 134–147. Kaplan, R. S., & Norton, D. P. (1996a). Linking the Balanced Scorecard to strategy. California Management Review, 39(Fall), 53–79. Kaplan, R. S., & Norton, D. P. (1996b). Strategic learning & the Balanced Scorecard. Strategy & Leadership, 24(September–October), 18–24. Kaplan, R. S., & Norton, D. P. (1996c). The Balanced Scorecard: Translating strategy into action. Boston, MA: Harvard Business School. Kaplan, R. S., & Norton, D. P. (1996d). Using the Balanced Scorecard as a strategic management system. Harvard Business Review, 74(January–February), 75–85. Kaplan, R. S., & Norton, D. P. (2000). Having trouble with your strategy? Then map it. Harvard Business Review, 78(September–October), 167–176. Kaplan, R. S., & Norton, D. P. (2001a). The strategy-focused organization: How Balanced Scorecard companies thrive in the new business environment. Boston, MA: Harvard Business School. Kaplan, R. S., & Norton, D. P. (2001b). Transforming the balanced scorecard from performance measurement to strategic management: Part I. Accounting Horizons, 15(March), 87–104. Kaplan, R. S., & Norton, D. P. (2001c). Transforming the balanced scorecard from performance measurement to strategic management: Part II. Accounting Horizons, 15(June), 147–160. Kaplan, R. S., & Norton, D. P. (2004a). Measuring the strategic readiness of intangible assets. Harvard Business Review, 82(February), 52–63. Kaplan, R. S., & Norton, D. P. (2004b). Strategy maps: Converting intangible assets into tangible outcomes. Boston, MA: Harvard Business School. Kennedy, J. (1993). Debiasing audit judgment with accountability: A framework and experimental results. Journal of Accounting Research, 31(Autumn), 231–245. Kennedy, J. (1995). Debiasing the curse of knowledge in audit judgment. The Accounting Review, 70(April), 249–273. Kerr, S. (1975). On the folly of rewarding A, while hoping for B. Academy of Management Journal, 18(December), 769–783. Klein, G. (1998). Sources of power: How people make decisions. Cambridge, MA: MIT Press. Larsen, H. T., Bukh, P. N. D., & Mouritsen, J. (1999). Intellectual capital statements and knowledge management: ‘‘Measuring’’, ‘‘reporting’’, ‘‘acting’’. Australian Accounting Review, 9, 15–26. Lebas, M. (1994). Managerial accounting in France: Overview of past tradition and current practice. European Accounting Review, 3(September), 471–487. Lemon, K. N., Rust, R. T., & Zeithaml, V. A. (2001). What drives customer equity. Marketing Management, 10(Spring), 20–25. Lev, B. (2001). Intangibles: Management, measurement, and reporting. Washington, DC: The Brookings Institution. Lev, B., & Mintz, S. L. (1999). Seeing is believing: A better approach to estimating knowledge capital. CFO, 15(February), 29–37. Libby, T., Salterio, S. E., & Webb, A. (2004). The balanced scorecard: The effects of assurance and process accountability on managerial judgment. The Accounting Review, 79(October), 1075–1094. Lipe, M. G., & Salterio, S. E. (2000). The balanced scorecard: Judgmental effects of common and unique performance measures. The Accounting Review, 75(July), 283–298.

Value-Creation Models for Value-Based Management

59

Lipe, M. G., & Salterio, S. E. (2002). A note on the judgmental effects of the balanced scorecard’s information organization. Accounting, Organizations and Society, 27(August), 31–40. Loveman, G. W. (1998). Employee satisfaction, customer loyalty, and financial performance: An empirical examination of the service profit chain in retail banking. Journal of Service Research, 1(August), 18–31. Low, J. (2000). The value creation index. Journal of Intellectual Capital, 1(September), 252–262. Low, J., & Siesfeld, T. (1998). Measures that matter: Wall Street considers non-financial performance more than you think. Strategy & Leadership, 26(March–April), 24–30. Lynn, B. (1998a). Intellectual capital. CMA: The Management Accounting Magazine, 72(February), 10–15. Lynn, B. (1998b). The Management of Intellectual Capital: The Issues and the Practice. Management Accounting Issues Paper 16. Hamilton, ON: Society of Management Accountants of Canada. Maines, L. A., Bartov, E., Fairfield, P. M., Hirst, D. E., Iannaconi, T. E., Mallett, R., Shrand, C. M., Skinner, D. J., & Vincent, L. (2002). Recommendations on disclosure of nonfinancial performance measures. Accounting Horizons, 16(December), 353–362. Maines, L. A., Bartov, E., Fairfield, P. M., Hirst, D. E., Iannaconi, T. E., Mallett, R., Shrand, C. M., Skinner, D. J., & Vincent, L. (2003). Implications of accounting research for the FASB’s initatives on disclosure of information about intangible assets. Accounting Horizons, 17(June), 175–185. Maisel, L. S. (1992). Performance measurement: The balanced scorecard approach. Journal of Cost Management, 6(Summer), 47–52. Malina, M. A., & Selto, F. H. (2001). Communicating and controlling strategy: An empirical study of the effectiveness of the balanced scorecard. Journal of Management Accounting Research, 13, 47–90. Malmi, T. (2001). Balanced scorecard in Finnish companies: A research note. Management Accounting Research, 12(June), 207–220. Maltz, A. C., Shenhar, A. J., & Reilly, R. R. (2003). Beyond the balanced scorecard: Refining the search for organizational success measures. Long Range Planning, 36(April), 187–204. March, A. (1986). A note on quality: The views of Deming, Juran, and Crosby. HBS Note 9-687011. Boston, MA: Harvard Business School. Meyer, S. M., & Collier, D. A. (2001). An empirical test of the causal relationships in the Baldrige health care pilot criteria. Journal of Operations Management, 19(July), 403–425. Mintz, S. L. (2000). A knowing glance: The second annual knowledge capital scorecard. CFO, 16(February), 52–62. Moers, F. (2005). Discretion and bias in performance evaluation: The impact of diversity and subjectivity. Accounting, Organizations and Society, 30(January), 67–80. Mooraj, S., Oyon, D., & Hostettler, D. (1999). The balanced scorecard: A necessary good or an unnecessary evil?. European Management Journal, 17(October), 481–491. Mouritsen, J. (1998). Driving growth: Economic value added versus intellectual capital. Management Accounting Research, 9(December), 461–482. Mouritsen, J., Bukh, P. N. D., Larsen, H. T., & Johansen, M. R. (2002). Developing and managing knowledge through intellectual capital statements. Journal of Intellectual Capital, 3(February), 10–29. Mouritsen, J., Johansen, M. R., Larsen, T., & Bukh, P. N. (2001). Reading an intellectual capital statement: Describing and prescribing knowledge management strategies. Journal of Intellectual Capital, 2(October), 359–383.

60

ROBERT H. ASHTON

Mouritsen, J., Larsen, H. T., & Bukh, P. N. D. (2001). Intellectual capital and the ‘‘capable firm’’: Narrating, visualizing and numbering for managing knowledge. Accounting, Organizations and Society, 26(October–November), 735–762. Neely, A., Adams, C., & Kennerley, M. (2002). The performance prism: The Scorecard for measuring and managing business success. London: Prentice Hall. Oliva, T. A., Day, D. L., & DeSarbo, W. S. (1987). Selecting competitive tactics: Try a strategy map. Sloan Management Review, 28(Spring), 5–15. Oliver, D. (1996). Skandia Assurance and Financial Services: Measuring and visualizing intellectual capital. IMD Case 396-116-1. Lausanne, Switzerland: International Institute for Management Development. Oliver, R. L. (1997). Satisfaction: A behavioral perspective on the consumer. Boston, MA: Irwin/ McGraw-Hill. Olson, E. M., & Slater, S. F. (2002). The balanced scorecard, competitive strategy, and performance. Business Horizons, 45(May–June), 11–16. Organization for Economic Cooperation and Development (OECD). (2000). Final Report: Measuring and reporting intellectual capital: Experience, issues, and prospects. Paris: OECD. Pannirselvam, G. P., & Ferguson, L. A. (2001). A study of the relationships between the Baldrige categories. International Journal of Quality and Reliability Management, 18, 14–34. Papalexandris, A., Ioannou, G., & Prastacos, G. P. (2004). Implementing the balanced scorecard in Greece: A software firm’s experience. Long Range Planning, 37(August), 351–366. Petty, R., & Guthrie, J. (2000). Intellectual capital literature review: Measurement, reporting and management. Journal of Intellectual Capital, 1(September), 155–176. Porter, M. E. (1980). Competitive strategy. New York, NY: The Free Press. Porter, M. E. (1985). Competitive advantage. New York, NY: The Free Press. Post, J. E., Preston, L. E., & Sachs, S. (2002). Redefining the corporation: Stakeholder management and organizational wealth. Stanford, CA: Stanford University Press. Powell, T. C. (1995). Total quality management as competitive advantage: A review and empirical study. Strategic Management Journal, 16(January), 15–37. Prahalad, C. K., & Hamel, G. (1990). The core competence of the corporation. Harvard Business Review, 68(May–June), 79–91. Previts, G. J., Bricker, R. J., Robinson, T. R., & Young, S. J. (1994). A content analysis of sellside financial analyst company reports. Accounting Horizons, 8(June), 55–70. Rappaport, A. (1986). Creating shareholder value: The new standard for business performance. New York, NY: The Free Press. Rappaport, A. (1998). Creating shareholder value: A guide for managers and investors. New York, NY: The Free Press. Rejc, A. (2005). Performance measurement in Central and Eastern Europe, Western Europe, and North America: A comparison of research and company practices. Working Paper, University of Ljubljana, Slovenia. Richmond, B. (1993). Systems thinking: Critical thinking skills for the 1990s and beyond. System Dynamics Review, 9(Summer), 113–133. Richmond, B. (2000). The ‘‘thinking’’ in systems thinking: Seven essential skills. The Toolbox Reprint Series. Waltham, MA: Pegasus Communications. www.pegasuscom.com Ridgway, V. F. (1956). Dysfunctional consequences of performance measurements. Administrative Science Quarterly, 1(September), 240–247. Roberts, M. L., Albright, T. A., & Hibbets, A. R. (2004). Debiasing balanced scorecard evaluations. Behavioral Research in Accounting, 16, 75–88.

Value-Creation Models for Value-Based Management

61

Roos, G., Bainbridge, A., & Jacobsen, K. (2001). Intellectual capital analysis as a strategic tool. Strategy & Leadership, 29(July–August), 21–26. Roos, G., & Jacobsen, K. (1999). Management in a complex stakeholder organization. Monash Mt. Eliza Business Review, (July), 83–93. Roos, G., & Roos, J. (1997). Measuring your company’s intellectual performance. Long Range Planning, 30(June), 413–426. Roos, J., Roos, G., Edvinsson, L., & Dragonetti, N. C. (1998). Intellectual capital: Navigating in the new business landscape. New York, NY: New York University Press. Ross, J. E. (1999). Total quality management: Text, cases and readings. New York, NY: St. Lucie Press. Rucci, A. J., Kirn, S. P., & Quinn, R. T. (1998). The employee-customer-profit chain at Sears. Harvard Business Review, 76(January–February), 82–97. Rungtusanatham, M., Forza, C., Filippini, R., & Anderson, J. C. (1998). A replication study of a theory of quality management underlying the Deming management method: Insights from an Italian context. Journal of Operations Management, 17(December), 77–95. Rust, R. T., Lemon, K. N., & Zeithaml, V. A. (2000). Driving customer equity: How customer lifetime value is reshaping corporate strategy. New York, NY: The Free Press. Rust, R. T., Zahorik, A. J., & Keiningham, T. L. (1994). Return on quality: Measuring the financial impact of your company’s quest for quality. Homewood, IL: Irwin. Rust, R. T., Zahorik, A. J., & Keiningham, T. L. (1995). Return on quality (ROQ): Making service quality financially accountable. Journal of Marketing, 59(April), 58–70. Said, A. A., HassabElnaby, H. R., & Wier, B. (2003). An empirical investigation of the performance consequences of nonfinancial measures. Journal of Management Accounting Research, 15, 193–223. Schmidt, F. L., & Kaplan, L. B. (1971). Composite vs. multiple criteria: A review and resolution of the controversy. Personnel Psychology, 24(Autumn), 419–434. Schneiderman, A. M. (2001). The first balanced scorecard: Analog Devices, 1986–1988. Journal of Cost Management, 15(September–October), 16–26. Shapiro, B. P., Rangan, V. K., Moriarty, R. T., & Ross, E. B. (1987). Manage customers for profits (not just sales). Harvard Business Review, 65(September–October), 2–9. Shields, J. F., & Shields, M. D. (2005). Revenue drivers: Reviewing and extending the accounting literature. Advances in Management Accounting, 14, 33–60. Skandia. (1994). Visualizing intellectual capital in Skandia. Supplement to Skandia’s 1994 Annual Report. Stockholm: Skandia. Skandia. (1995a). Renewal and development. Supplement to Skandia’s 1995 Interim Report. Stockholm: Skandia. Skandia. (1995b). Value-creating processes. Supplement to Skandia’s 1995 Annual Report. Stockholm: Skandia. Skandia. (1996a). Power of innovation. Supplement to Skandia’s 1996 Interim Report. Stockholm: Skandia. Skandia. (1996b). Customer value. Supplement to Skandia’s 1996 Annual Report. Stockholm: Skandia. Skandia. (1997). Intelligent enterprising. Supplement to Skandia’s 1997 Interim Report. Stockholm: Skandia. Skandia. (1998). Human capital in transformation. Supplement to Skandia’s 1998 Annual Report. Stockholm: Skandia.

62

ROBERT H. ASHTON

Soteriou, A., & Zenios, S. A. (1999). Operations, quality, and profitability in the provision of banking services. Management Science, 45(September), 1221–1238. Speckbacher, G., Bischof, J., & Pfeiffer, T. (2003). A descriptive analysis on the implementation of balanced scorecards in German-speaking countries. Management Accounting Research, 14(December), 361–387. Stemsrudhagen, J. I. (2004). The structures of balanced scorecards: Empirical evidence from Norwegian manufacturing industry. In: M. J. Epstein & J. F. Manzoni (Eds), Performance measurement and management control: Superior organizational performance (pp. 303–321). London: Elsevier. Sterman, J. D. (2000). Business dynamics: Systems thinking and modeling for a complex world. Boston, MA: McGraw-Hill. Stewart, G. B. (1991). The quest for value. New York, NY: Harper Business. Stewart, T. A. (1997). Intellectual capital: The new wealth of organizations. New York, NY: Doubleday. Stewart, T. A. (2001). The wealth of knowledge: Intellectual capital and the twenty-first century organization. New York, NY: Currency Doubleday. Sullivan, P. H. (1998). Profiting from intellectual capital: Extracting value from innovation. New York, NY: Wiley. Sullivan, P. H. (2002). Value-driven intellectual capital: How to convert intangible corporate assets into market value. New York, NY: Wiley. The Conference Board. (1997). Leveraging intellectual capital. HR Executive Review, Vol. 5, No. 3. New York: The Conference Board. Vaivio, J. (1999). Exploring a ‘‘non-financial’’ management accounting change. Management Accounting Research, 10(December), 409–437. Waldman, D. A. (1994). The contributions of total quality management to a theory of work performance. Academy of Management Review, 19(July), 510–536. Wallman, S. M. H. (1995). The future of accounting and disclosure in an evolving world: The need for dramatic change. Accounting Horizons, 9(September), 81–91. Wallman, S. M. H. (1996). The future of accounting and financial reporting. Part II: The colorized approach. Accounting Horizons, 10(June), 138–148. Walton, M. (1986). The Deming Management Method. New York, NY: Putnam. Webber, A. M. (2000). New math for a new economy. Fast Company, 31(January–February), 214–224. Wernerfelt, B. (1984). A resource-based view of the firm. Strategic Management Journal, 5(April–June), 171–180. Wilson, D. D., & Collier, D. A. (2000). An empirical investigation of the Malcolm Baldrige National Quality Award causal model. Decision Sciences, 31(Spring), 361–390. Zahorik, A. J., & Rust, R. T. (1992). Modeling the impact of service quality on profitability: A review. In: T. Schwartz (Ed.), Advances in services marketing and management (Vol. 1, pp. 247–276). Greenwich, CT: JAI Press. Zeithaml, V. A. (1988). Consumer perceptions of price, quality and value: A means-end model and synthesis of evidence. Journal of Marketing, 52(July), 2–22. Zeithaml, V. A. (2000). Service quality, profitability, and the economic worth of customers: What we know and what we need to learn. Journal of the Academy of Marketing Science, 28(Winter), 67–85. Zeithaml, V. A., Rust, R. T., & Lemon, K. N. (2001). The customer pyramid: Creating and serving profitable customers. California Management Review, 43(Summer), 118–142.

PERFORMANCE STANDARDS AND MANAGERS’ ADOPTION OF RISKY PROJECTS$
Chee W. Chow, James M. Kohlmeyer, III and Anne Wu
ABSTRACT
Innovation is the key to competitive advantage, and attaining innovation often requires taking on higher-than-usual levels of risk. Yet, while managers commonly profess support for efforts in innovation, they often emphasize safe, short-term results over more risky, long-term outcomes. As a result, a major challenge to firms is increasing employees’ willingness to adopt risky yet more profitable alternatives. This study uses an experiment to test how the level of performance standard, per se, affect employees’ propensity to take on (more) risky projects. Using participants from the U.S. and Taiwan to represent higher versus lower individualism national cultures, it also examines the effects of national culture on employee actions. The findings are consistent with expectations from combining goal and prospect theories that a specific high standard motivates greater risk taking than a low standard. We find only limited difference between the U.S. and Taiwanese samples’
$

Alphabetically ordering is followed in listing the authors. All authors contributed equally to this project.

Advances in Management Accounting, Volume 16, 63–105 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16002-0

63

64

CHEE W. CHOW ET AL.

individualism/collectivism scores, which may help to explain the lack of significant differences between their reactions to the performance standard treatment.

1. INTRODUCTION
This study explores how performance standards affect managers’ willingness to pursue risky projects. This topic is important because in the current era of globalized competition, innovation has been identified as a key to firms’ continued survival and success (Bouchikhi & Kimberly, 2001; Jacobson, 1992; Page, 1993). In turn, attempts at innovation often require venturing into uncharted waters, thus exposure to higher-than-usual levels of risk (Chatterjee, Wiseman, Fiegenbaum, & Devers, 2003; Mullins, Forlani, & Walker, 1999; Wan, Ong, & Lee, 2005). Yet, despite professing to support innovation, more often than not managers emphasize safe, short-term results at the expense of more risky long-term opportunities (Kuczmarski, 1996; Kunz & Pfaff, 2002; Stewart, Watson, & Carland, 1999; Wiseman & Gomez-Mejia, 1998). Managers’ tendency to shun risk can be understood when one compares the effects of risky projects from their perspective versus that of the firm. A firm typically is comprised of many projects across numerous managers. Having a large portfolio, in turn, helps to diversify away much of the project-specific risks. In contrast, individual managers generally have authority over, and are held accountable for, a far more limited set of projects. Consequently, they also have far less ability to diversify away the risk of individual projects (Barney & Hesterly, 1996; Fama, 1980; Jensen & Murphy, 1990). Given this divergence between them and their managers, firms are faced with the challenge of increasing the latter’s willingness to pursue risky projects (Atkinson et al., 1997; Baysinger, Kosnik, & Turk, 1991; Garen, 1994; Hoskisson, Hitt, & Hill, 1991; Jensen & Meckling, 1976; Lambert, 2001; Tosi & Gomez-Mejia, 1989). An important part of firms’ tool kit for motivating managers is the performance evaluation and reward system, and the design and effects of performance-based compensation schemes have long been a major area of management accounting research (Sprinkle, 2003). However, this literature has mostly focused on effects other than risk taking (e.g., effort, performance). While a few studies have explored how performance evaluation and compensation schemes affect employees’ risk taking tendency, they have only considered a limited set of features or special circumstances.

Performance Standards and Managers’ Adoption of Risky Projects

65

The current study extends investigation to the level of the performance standard. This focus is predicated on a substantial prior literature having demonstrated powerful effects of performance standards on employee motivation (Bonner & Sprinkle, 2002; Merchant & Manzoni, 1989; Otley, 1987). A secondary focus of this study is potential cross-cultural differences in performance standards’ motivational effects. This aspect of the study is motivated by the increasing globalization of economic activities. Most theories of behavior used to derive management implications have made the assumption that people would act opportunistically for private gain at the expense of the collective (e.g., the firm) (Baiman, 1990; Koford & Penno, 1992). While this assumption may accurately capture a major behavioral tendency of, say, Anglo-Americans, it may only partially, or even inaccurately, represent the leanings of individuals from other national cultures (e.g., Asia; see Chow, Kato, & Merchant, 1996; Hofstede, 1984; Kachelmeier & Shehata, 1997). As such, it is important to evaluate the global applicability of extant theories and findings. In their review of crosscultural management accounting research, Harrison and McKinnon (1999) specifically identified as worthwhile future research on the interplay between national culture and issues such as risk-taking and innovative propensities. In the current study, national culture is operationalized by comparing subjects from the U.S. and Taiwan. The remainder of this article is structured as follows. Section 2 provides an overview of related prior literatures as the basis for developing two hypotheses. Then, the data collection method is explained in Section 3. Section 4 presents the results, and Section 5 provides a summary and discussion.

2. LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT
This section is divided into two subsections. The first subsection discusses prior studies on the motivational effects of standard-based compensation systems. This review provides the basis for our first hypothesis. In the second subsection, prior studies on the effects of national culture are used for developing our second hypothesis. 2.1. The Motivational Effects of Standard-based Compensation Systems Organizations often try to motivate employees via explicit performance standards, such as ones relating to profit, return on investment, sales

66

CHEE W. CHOW ET AL.

revenue, and production cost (Merchant, 1998). Numerous studies have examined the effects of such standard-based compensation systems on a wide range of outcomes, including performance (Bonner, Hastie, Sprinkle, & Young, 2000; Fatseas & Hirst, 1992; Nouri & Parker, 1998; Sprinkle, 2000; Young & Lewis, 1995), effort (Awasthi & Pratt, 1990; Berg, Daley, Gigler, & Kanodia, 1990; Bonner & Sprinkle, 2002), satisfaction (Awasthi, Chow, & Wu, 2001; Birnberg, Turopolec, & Young, 1983; Hopwood, 1972; Parker & Kohlmeyer, 2005), and budgetary slack (Chow, Cooper, & Waller, 1988; Fisher, Frederickson, & Peffer, 2002; Waller, 1988; Young, 1985; Young, Fisher, & Lindquist, 1993). However, few studies have examined which incentive schemes or dimensions thereof affect managers’ risk taking behavior (Sprinkle, 2003). Among the small number of accounting studies in this area, Chow and Haddad (1991), later supplemented by Frederickson (1992), reported experimental results supportive of Holmstrom’s (1982) proposition that by filtering out the effects of common uncertainty, relative performance evaluation increases individuals’ propensity to take risk. Sayre, Rankin, and Fargher (1998) investigated compensation systems of the ‘‘winner-take-all’’ type (e.g., one manager is selected out of several for promotion). In their experiment, subjects tended to make investments of extremely high or low risk depending on how their performance compared with that of the leader. Along the same vein, Ruchala (1999) focused on performance relative to the budget goal rather than the performance of others. She found that subjects assigned to a not-achieving-budget-goals condition made riskier decisions. The highest risks were taken by those who were both not meeting budget goals and being compensated with bonus-based compensation. However, while this study demonstrates that budget goals can affect risk-taking behavior, it does not examine the effect of different goal levels. A key issue in designing a standard-based compensation system is selecting the level of the performance standard. An extensive literature on goal setting has long maintained that specific and challenging goals (performance standards) lead to higher performance than easy goals, ‘‘do your best’’ goals, or no goals at all (Brown & Latham, 2000; Latham, 2004; Locke & Latham, 1990). Studies in the accounting literature (e.g., Chong & Chong, 2002; Chow, 1983; Fatseas & Hirst, 1992; Hopwood, 1974; Otley, 1987) have reported findings consistent with this view. However, a pre-condition for the (difficult) standard to have positive motivating effects is that it has to be accepted by employees. If a performance standard is perceived as being unattainable, then it can lead to poorer performance because employees would become discouraged and give up trying for its attainment (Fatseas &

Performance Standards and Managers’ Adoption of Risky Projects

67

Hirst, 1992). This latter concern has led some authors to question the superiority of difficult performance standards. A field study by Merchant and Manzoni (1989), for example, has reported that profit center managers consider it most advantageous to the firm to have performance targets that are achievable more than 80% of the time. For purposes of understanding how performance standards affect employee’s risk-taking, a limitation of prior studies is that they have largely focused on tasks with well-defined input/output relationships. With such tasks, superiors can infer subordinates’ effort or action choices from their outputs. In contrast, risk taking is an ex ante concept, where a ‘‘good’’ outcome can come from a ‘‘bad’’ decision, and a ‘‘good’’ decision can give rise to a ‘‘bad’’ outcome. To the extent that superiors cannot infer subordinates’ decisions based on their outcomes, whether these prior studies’ findings can be generalized to risk-taking tasks is an open question. Both psychology-based and economics-based theories (e.g., expectancy and agency theories) suggest that when compensation is linked to performance relative to the standard, employees can be motivated to attain higher performance standards by increasing the rewards for standard attainment (e.g., Demski & Feltham, 1978; Holmstrom, 1979; Isaac, Zerbe, & Pitt, 2001; Prendergast, 1999; Vroom, 1964). This inference is predicated on the assumption that employees are risk averse. Yet people’s choices among risky prospects often exhibit patterns that are inconsistent with risk aversion (Fiegenbaum, 1990; Piron & Smith, 1995; Wiseman & Bromiley, 1996; Wiseman & Gomez-Mejia, 1998). For example, sometimes they may opt for a more risky alternative even if its expected value, or certainty equivalent in terms of economic theory, is lower than that of the less risky option (e.g., purchase of a lottery ticket). Prospect theory was proposed by Kahneman and Tversky (1979) to explain these choices, as well as risky choices in general. This theory suggests that people may be either risk averse or risk taking, depending on how the uncertain outcomes compare with a reference point. The function representing how people value different alternatives is envisioned as an S-shaped curve that passes through the reference point, being concave for gains and convex for losses, and also steeper for losses than for gains (Kahneman & Tversky, 1979). This shape implies that people are more concerned with avoiding loss to wealth than attracting additional wealth (i.e., they are loss avoiders rather than wealth maximizers) (Gomez-Mejia, Welbourne, & Wiseman, 2000; Wiseman & Gomez-Mejia, 1998). Stated another way, the aggravation that one experiences in losing a certain sum of money is greater than the pleasure associated with gaining the same amount (Kahneman & Tversky, 1979).

68

CHEE W. CHOW ET AL.

In general, prospect theory implies that in choosing among risky alternatives, decision makers will act conservatively when facing potential gains, but take greater risks when facing potential losses (Sitkin & Pablo, 1992; Thaler & Johnson, 1990; Tversky & Kahneman, 1986, 1991). In the accounting literature, Rutledge and Harrell (1993, 1994) and Sharp and Salter (1997) have provided support for prospect theory in their study of escalation-ofcommitment, while Luft (1994) has demonstrated its applicability to employees’ choice of bonus schemes. Overall, prospect theory, in conjunction with goal theory, implies that the level of performance standard, per se, can affect employees’ risk taking behavior. This inference is consistent with the argument of Payne, Laughhunn, and Crun (1980) that the performance target (standard) influences the location of the reference point in people’s value functions. While the existence of any explicit performance standard will demarcate outcomes into gains and losses, a high performance standard will cause more outcomes to be framed as loss-making situations as compared with a low performance standard. And given that people are more averse to a loss than they value a gain of the same magnitude, they should be more inclined toward risk taking to avoid failing to meet a high performance standard than to exceed a low performance standard. Testing this prediction is the primary objective of our study. In focusing on the motivating effects of performance standard level, we are quick to acknowledge the implication of expectancy and agency theories that employees can be motivated to take greater risks if they are compensated for doing so, such as by paying more for attaining a higher performance standard. Tests of this implication also are worthwhile. But if it is found that the level of performance standards, independent of standard-based compensation, affects employees’ risk taking behavior, this additional knowledge can further enrich managers’ repertoires in designing the overall control system. H1. The level of performance standard, per se, affects employees’ willingness to take risk. A higher performance standard leads to greater risk taking as compared with a lower performance standard. 2.2. The Nature and Effects of National Culture National culture influences a person’s actions either by supplying the values toward which the actions are oriented, or by shaping a repertoire of action strategies in which certain patterns of action are facilitated while others are discouraged (Erez & Earley, 1993; Hodgetts & Luthans, 1997; Triandis, 1989). Across multiple taxonomies that have been proposed for

Performance Standards and Managers’ Adoption of Risky Projects

69

operationalizing the national culture construct (e.g., Hofstede, 1980, 1991; Ronen & Shenkar, 1985; Schwartz, 1994; Smith, Dugan, & Trompenaars, 1996; Trompenaars, 1994), individualism/collectivism has consistently been identified as being a basic, or core, value that distinguishes members of different cultural groups from one another (Harrison, 1993; Hofstede, 1980, 1991; Lachman, Nedd, & Hinings, 1994; Smith, Peterson, & Schwartz, 2002; Triandis, 1995; Triandis, Bontempo, Villareal, Asai, & Lucca, 1988; Sondergaard, 1994; see Earley & Gibson, 1998 for a comprehensive review). Furthermore, across a large number of studies, individualism/collectivism has been identified as the most important dimension of national culture in comparing East and West (Triandis, 1995; Smith et al., 1996). Individualism and its opposite, collectivism, relate to the relative emphasis that members of a society place on their self-interests versus those of the group. People from a collectivist culture tend to define themselves in terms of their relationships with others, and they are more inclined to give up their individual needs when there is a conflict between them and group needs (Earley, 1993; Triandis, 1995). In contrast, members of an individualist culture tend to define themselves as autonomous entities independent of groups (Hofstede, 1991; Markus & Kitayama, 1991), and are more likely to emphasize their individual needs over group needs (Redding, 1993; Wagner, 1995). We operationalize individualism/collectivism by comparing U.S. nationals to Chinese nationals residing in Taiwan. Based on Hofstede’s (1980) measurement scale, the former had a score of 91 on individualism/ collectivism, while the latter were at the low end of the scale with a score of 45. More generally, much research has isolated the self-interest motive as being a cornerstone of American theories and practices (Bellah, Madsen, Sullivan, Swidler, & Tipton, 1987; Earley, 1993; Harris & Moran, 1987; Triandis et al., 1988). In contrast, Chinese nationals have repeatedly been cited for an emphasis on subjugating one’s own interests to those of the collective (Bond, Leung, & Wan, 1982; Bond & Hwang, 1986; Leung & Bond, 1984). Based on the preceding characterization of individualism/ collectivism, when risk taking poses a conflict between self and collective interests, members of a more collectivist culture (i.e., Chinese nationals) are expected to take on more risk as compared with members of a more individualistic culture (i.e., U.S. nationals), thereby placing the interests of the collective over their own interests. Thus, we hypothesize: H2. Holding constant the level of performance standard, when it is in the firm’s best interests to take on more risky projects, Chinese nationals in Taiwan will engage in greater risk taking than will their U.S. counterparts.

70

CHEE W. CHOW ET AL.

3. METHOD
3.1. Design An experiment was used to control for the effects of extraneous factors and for its replicability. There were two between-subjects treatments: national culture (U.S., Chinese) and performance standard (high, low). As explained further below, the dependent variable was the riskiness of projects selected by the participants. 3.2. Task and Procedure An experimenter and an assistant were present at all times to answer questions and to preclude interaction among participants during the experiment. Subjects were asked to assume the role of a product manager for XYZ Company. Their task was to select products for the company to produce and market. The appendix provides the experimental materials for the high performance standard case. The materials for all the treatments were pre-tested with upper-level accounting major students in both the U.S. and Taiwan. All reported that the ‘‘story’’ of the task motivated them to take the decision making seriously. The experimental session lasted $75 min and comprised the following three stages: 3.2.1. Stage One Participants read through a description of XYZ Company and the product manager’s role. Then they worked through several practice exercises which focused on the nature of the probability distributions reflecting the possible profit outcomes from each product, and how product profit outcomes affected the company’s profit and the participant’s pay. Correct answers were provided after participants were satisfied with their own answers. The XYZ Company was described as being in the business of manufacturing and marketing fad items that have a 1-year market life. At the beginning of each year, the Company allotted each of its many product managers $1,000,000 to produce and market one product (all products cost $1,000,000 to manufacture and market).1 The product manager’s task was to select one out of the seven new products that his/her staff proposed each year. Each product proposal included the staff’s estimate of the probability that specific profit rates (return on the $1,000,000 invested) will be achieved over the product’s 1 year life.

Performance Standards and Managers’ Adoption of Risky Projects

71

For simplicity, the same seven probability distributions were used in each year. All of these distributions were symmetrical about a mean of 15%, with a range of 1–29%.2 The three distributions at the low end of riskiness were bell-shaped, with increasingly dispersed distributions about the mean. The fourth distribution was uniform in shape (i.e., a rectangle, with all outcomes being equally likely). Then at the high end of riskiness were three bimodal distributions, with the two modes leaning increasingly toward the lowest and highest possible outcomes as one moved toward the high end of the risk continuum. In terms of outcome variance (our measure of riskiness), the values for the seven distributions were 21.96, 37.3, 50.48, 70.15 (uniform distribution), 100.54, 114.81, and 129.05.3 Participants were randomly assigned to a high performance standard (a 25% profit rate) or a low performance standard (a 5% profit rate). To maintain our focus on the level of performance standard, per se, pay was entirely based on actual performance and not affected by performance relative to the standard. This treatment was operationalized as follows. Participants with a high performance standard were paid a base salary of $125,000 ( ¼ 50% of profit at the 25% performance standard profit rate), plus an adjustment (+ or À) equal to 50% of the deviation between actual and standard performance. Participants with a low performance standard (5% profit rate) were paid a base salary of $25,000 ( ¼ 50% of profit at the 5% profit rate), plus an adjustment (+ or À) equal to 50% of the deviation between actual and standard performance. Use of the parameter ‘‘50%’’ to derive all subjects’ base pay and as the bonus/penalty factor assured that under both performance standard levels, pay was exactly the same for any level of realized profit.4 Lastly, participants were told that XYZ Company has been in existence for a number of years and on average, its products had produced profits equal to 15% of their manufacturing and marketing costs. But because the industry was intensely competitive and getting more so, the Company believed that if it was to survive and prosper, it must come out with a number of breakthrough products that can produce significantly higher levels of profits than the historical level. Participants were explicitly told that because the Company had a large number of product managers, it was not concerned with individual managers taking on highly risky projects.5 3.2.2. Stage Two Participants completed three experimental periods. Each period began with each participant receiving the profit probability distributions for seven proposed products. These products were labeled At, Bt, y , Gt, where the

72

CHEE W. CHOW ET AL.

subscript ‘‘t’’ stood for the experimental period (thus, t ¼ 1, 2, 3). After a participant had written down (on a form provided for this purpose) his/her choice among these products, he/she brought the form to the experimenter at the front of the room. The experimenter had an array of seven brown paper bags, labeled At, Bt, y , Gt just like the seven proposed products. The participant was asked to draw out a folded piece of paper from the bag corresponding to his/her product choice, and to record the outcome (realized profit rate) on his/her form. After returning the folded piece of paper to the bag, the participant returned to his/her seat to compute the Company’s profit and his/her total pay for the period, then started the next experimental period. Participants were told that the first two experimental periods were for practice, and that they would only be paid based on their profit performance in the third period. Since participants’ product choices may be affected by prior outcomes, they were randomly assigned to two outcome sequences for the two practice periods (by rigging the numbers written on the folded sheets in each brown bag). The ‘‘good-bad’’ sequence was a 22% realized profit rate for the first period, and 7% for the second period. These outcomes were reversed in the ‘‘bad-good’’ sequence. For the ‘‘real’’ (third) period, participants drew from bags that accurately reflected their selected products’ probability distributions. 3.2.3. Stage Three After completing the third (real) experimental period, participants completed an exit questionnaire. Then they were debriefed, paid, and dismissed.6 In addition to several demographic questions, the exit questionnaire asked each participant how much he/she had made his/her decisions in the experiment as if facing a real world situation. There also were questions on the degree to which their decision making had been influenced by each of the experimental treatments: performance standard, relation between actual performance and pay, and the company’s desire for breakthrough products.7 All used a 9-point response scale, with 1 ¼ ‘‘not at all’’ and 9 ¼ ‘‘completely.’’ Participants who answered above ‘‘1’’ were asked to elaborate on their numerical answers. Finally, subjects completed three scales from Hofstede (1980), Earley (1993), and Chinese Cultural Connection (1987), respectively, on individualism/collectivism. To avoid biases or errors due to language proficiency, the Taiwanese portion of the experiment was conducted in Chinese. The English version of the instrument was first translated into Chinese by a person not affiliated with the study. Then two bilingual members of the research team evaluated

Performance Standards and Managers’ Adoption of Risky Projects

73

the translation against the English version for content equivalence. Only minor changes were deemed necessary; there were made through a consultative process.

3.3. Sample A total of 40 (46) U.S. (Chinese) nationals voluntarily participated in the experiment. All were full time Masters level business students. The U.S. participants attended a large public university in the southeast, while the Taiwanese subjects were enrolled in a large public university in Taipei. Three (9) participants were dropped from the U.S. (Taiwan) sample due to answering 4 or below about the extent to which they had made decisions in the experiment as if facing a situation in real life. Thus, the sample used for hypothesis testing comprised 37 U.S. nationals and 37 Chinese nationals residing in Taiwan. Both national samples were split 17/20 between the low and high performance standard treatments. On average, the U.S. participants were 26 years old (range: 21–42 years) and had 3.16 years of full time equivalent working experience (range: 0–13 years). Slightly over half (51%) were male. The average Taiwanese subject was slightly younger (23.86 years; range: 22–27 years) and tended not to have had significant full time working experience (mean ¼ 0.38 years; range: 0–3.5 years). Slightly over half (54%) were male. Within each national sample, there was no statistically significant demographic difference across performance standard treatments. Independent samples tests indicated that the U.S.–Taiwan differences in work experience and age were statistically significant. However, neither factor had a significant effect when included in the hypothesis tests.

4. RESULTS
4.1. Validation of Cultural Differences Table 1 presents the two national samples’ scores on each item of the three culture scales. Panel A contains six questions from Hofstede (1980, p. 220) which he identified as being significantly related to the individualism/ collectivism dimension. Of the six items, individualists are expected to assign greater importance (i.e., higher scores) to the first three, while collectivists are expected to ascribe greater importance to the latter three. Panel A shows

74

CHEE W. CHOW ET AL.

Table 1. Comparison of U.S. and Taiwan Samples’ Culture Scores.
Panel A: Individualism/Collectivism Scale from Hofstede (1980)a U.S. Part I: Mean scores on individual items Have sufficient time for personal or family life Have considerable freedom to adopt own approach to the job Challenging work Fully use skills and abilities Good physical working conditions Have training opportunities Part II: Sums of subsets of items SUM1–3 (first 3 items) SUM4–6 (last 3 items) Independent Samples Test: U.S. versus Taiwan Mean Difference SUM1–3 – SUM4–6 SUM1–3/SUM4–6 0.35 0.02 t 0.72 0.48 df 72 72 Sig. (2-tailed) 0.47 0.63 Taiwan

4.13 3.35 3.97 3.73 3.84 3.86 11.46 11.43

3.97 3.73 3.84 3.97 3.84 4.05 11.54 11.86

Panel B: Individualism/Collectivism Scale from Earley (1993)b U.S. Part I: Mean scores on individual items Employees like to work in a group rather than by themselves If a group is slowing me down, it is better to leave and work alone To be superior, a man must stand alone One does better work working alone than in a group I would rather struggle through a personal problem by myself than discuss it with my friends An employee should accept the group’s decision even when personally he/she has a different opinion Problem solving by groups gives better results than problem solving by individuals The needs of people close to me should take priority over my personal needs Part II: Summed score over all 8 items Taiwan

2.73 2.75 3.70 3.25 3.75 2.47 3.56 3.09 25.30

3.78 2.86 2.86 3.49 4.03 3.70 3.41 3.73 27.86

Performance Standards and Managers’ Adoption of Risky Projects

75

Table 1.

(Continued)

Independent Samples Test: U.S. versus Taiwan Mean Difference Summed score over all 8 items À2.57 t À3.33 df 72 Sig. (2-tailed) 0.00

Panel C: Integration (Individualism) Scale from Chinese Cultural Connection (1987)c U.S. Part I: Mean scores on individual items a. Tolerance of others b. Harmony with others c. Solidarity with others d. Non-competitiveness e. Trustworthiness f. Contentedness with one’s position in life g. Being conservative h. A close, intimate friend i. Filial piety j. Patriotism k. Chastity in women Part II: Scores on Integration culture dimension Integration Independent Samples Test: U.S. versus Taiwan Mean Difference Integration À2.03 t À1.24 df 72 Sig. (2-tailed) 0.22 Taiwan

6.89 6.62 5.78 3.92 8.35 7.65 4.50 7.54 2.81 3.57 5.70 63.34

5.86 7.78 7.58 7.55 7.20 5.65 4.11 7.68 2.14 5.19 4.62 65.36

a Response scale for individual items: 5 ¼ of utmost importance and 1 ¼ of little or no importance. b Response scales for individual items: For items 1, 6, 7, and 8, 5 ¼ strongly agree and 1 ¼ strongly disagree; for items 2, 3, 4, and 5, 5 ¼ strongly disagree and 1 ¼ strongly agree. c Response scale for individual items: 9 ¼ ‘‘Of supreme importance’’ and 1 ¼ ‘‘Of no importance at all.’’ For items i, j, and k, the raw ratings were reversed in calculating the Integration scores.

that the expected directional difference between the Taiwan and U.S. samples is found for four out of the six items. Following Chow, Deng, and Ho (2000), two indices were constructed to test the significance of the Taiwan–U.S. difference on Hofstede’s scale. One was the difference between the summed scores for the first and second sets of

76

CHEE W. CHOW ET AL.

three questions. The other was the ratio between these two summed scores. Consistent with the U.S. sample being more individualistic, it had a higher numerical value than the Taiwanese sample for both indices (0.35 and 0.02). However, neither difference was statistically significant (t ¼ 0.72, p ¼ 0.47; t ¼ 0.48, p ¼ 0.63). Panel B of Table 1 presents the 8-item individualism/ collectivism scale adopted from Earley (1993). The responses are coded such that a higher score indicates higher collectivism. This panel indicates that the Taiwanese sample is higher on six out of the eight items, and the sum of the eight items also is higher for the Taiwanese than the U.S. sample (27.86 vs. 25.30). Although the difference between the Taiwanese and U.S. samples is numerically quite small, it nevertheless is statistically significant (t ¼ 3.33, p ¼ 0.00), and is consistent with the former being more collectivistic. Finally, Panel C of Table 1 contains 11 items from the Chinese Cultural Connection’s (1987) ‘‘Integration’’ cultural dimension, which they found to be positively related to Hofstede’s individualism measure. Of these 11 items, the first eight are positively, and the last three are negatively, loaded on the Integration measure. While the sum of the 11 Integration items’ scores (with the last three being reverse scaled) has a higher mean for the Taiwanese than for the U.S. sample (65.36 vs. 63.34), this difference is not statistically significant (t ¼ 1.24, p ¼ 0.22). Hence, across the three measurement scales for individualism/collectivism, there is only limited indication that the U.S. nationals in our sample were more individualistic than our sample of Chinese nationals from Taiwan. 4.2. Hypothesis Tests Table 2 presents the means of variable Choice3, which is based on the subjects’ product choices in the third (real) experimental period. The numeric value of this variable was based on assigning to each of the seven possible choices the values 1–7, progressing from the product with the lowest variance (product A) to that with the highest variance (product G). Thus, a higher mean indicates that subjects in that treatment had chosen proportionally Table 2. The Means of Choice3 within and between National Origins.
U.S. Low standard High standard 3.33 4.84 Taiwan 3.18 4.50

Note: Choice3 is the riskiness of subjects’ product choices in the experimental period.

Performance Standards and Managers’ Adoption of Risky Projects

77

more of the higher variance products. Exhibit 1 presents these results visually by plotting the mean values of Choice3 across the subjects’ national origins and performance standard treatments. H1 states that a higher performance standard would lead to greater risk taking as compared with a lower performance standard. Both Table 2 and Exhibit 1 strongly suggest existence of such an effect. More formally, H1 is supported by the ANOVA results in Table 3, which indicate a significant main effect for performance standard. Within this table, the results labeled ‘‘A’’ and ‘‘B’ differ in that the latter also includes the sequence of outcomes
5 U.S., 4.84 The means of choice 3 4.5 Taiwan, 4 U.S., 3.33 3.5 3 Low Standard High Standard Performance Standard Taiwan, 3.18 4.5

Exhibit 1.

Plot of Choice3 against Performance Standard, U.S. versus Taiwan.

Table 3.

ANOVA Results for Hypothesis Testing (Dependent Variable: Choice3).
Sum of Squares df F-Value Sig.

A. National origin Perf. Std. National origin  Perf. Std. Error B. National origin Perf. Std. Good/bad National origin  Perf. Std. Perf. Std.  Good/bad Error

1.15 36.97 0.16 224.00 0.62 35.61 0.51 0.35 2.64 221.00

1 1 1 70 1 1 1 1 1 68

0.36 11.55 0.05

0.55 0.00 0.82

0.19 10.96 0.16 0.11 0.81

0.66 0.00 0.69 0.74 0.37

78

CHEE W. CHOW ET AL.

(good/bad vs. bad/good) in the two trial periods. Both approaches indicate that only performance standard has a statistically significant main effect (F ¼ 11.55, 10.96; both p values areo0.001). H2 states that when it is in the firm’s best interests to take on more risky projects, Chinese nationals in Taiwan will engage in greater risk-taking than their U.S. counterparts. The ANOVA results in Table 3 indicate no significant main or interaction effects for national origin. Thus, H2 is not supported. 4.2.1. Additional Analyses Recall that the exit questionnaire had asked each subject the extent that his/ her product choice had been affected by the performance standard, the linkage of actual performance to pay, and the company’s expressed desire for managers to pursue breakthrough products. Table 4 reports the mean responses (on a 1–9 scale) to these questions by each national sample. Table 5 presents the results of separate ANOVAs on these three metrics, with national origin, performance standard, and their interaction as factors. Performance standard level has a significant main effect on the influence of the performance standard (F ¼ 7.92, p ¼ 0.00), while the interaction between national origin and performance standard level is significant for the linkage to pay (F ¼ 5.38; p ¼ 0.02). To gain further insights into how the performance standard level had affected the subjects’ product choices, we examined their open-ended responses explaining how they had been affected by this factor. Most subjects merely described their decision rules (e.g., ‘‘If the performance standard is higher, I would take more risks to select the products with high profit’’). Nevertheless, there was a discernable shared concern with the probability Table 4. Mean Ratings of the Influence of the Performance Standard, Link between Performance and Pay, and Company’s Expressed Desire for Breakthrough Products.
U.S. Perf. Std. Pay Link Co. Breakthrough Desire 3.59 2.76 1–9 Perf. Std. Taiwan Pay Link Co. Breakthrough Desire 3.92 2.70 1–9

Mean Std. Dev. Range

6.65 2.37 1–9

6.46 2.42 1–9

6.16 1.99 1–9

6.92 1.61 3–9

Performance Standards and Managers’ Adoption of Risky Projects

79

Table 5. ANOVA Results for Influences of the Performance Standard, Link between Performance and Pay, and Company’s Expressed Desire for Breakthrough Products.
Sum of Squares Dependent variable: Perf. Std. National origin Treatment National origin  treatment Error Dependent variable: Pay link National origin Treatment National origin  treatment Error df F-Value Sig.

4.95 35.08 0.18 310.18 4.52 11.94 20.85 271.06

1 1 1 70 1 1 1 70 1 1 1 70

1.12 7.92 0.04

0.29 0.00 0.84

1.17 3.08 5.38

0.28 0.08 0.02

Dependent variable: Co. breakthrough desire National origin 1.68 Treatment 12.74 National origin  treatment 7.09 Error 524.94

0.22 1.70 9.45

0.64 0.20 0.99

that the performance standard would not be met. Among the 17 Taiwanese responses from the low standard treatment, 13 explicitly noted this concern (‘‘I would avoid selecting the products whose profit rate is below 5% with high probability;’’ ‘‘The level of performance standard would affect the probabilities of profit rates that are higher than performance standard. Therefore, performance standard would affect my product selection’’). Seventeen of their 20 countrymen in the high standard case noted the same concern (‘‘Because of the existence of a performance standard, I would think about the risk (probability) when I made decisions’’). Out of 17 U.S. subjects facing a low performance standard, 10 out of the 13 who had answered the open-ended question reported focusing on the risk of failing to meet the performance standard (‘‘I looked to see what was the probability of getting at least the performance standard profit rate’’; ‘‘Based on likelihood of greater than 5% profit and moderate risk of achieving 15% or more’’). Of the 20 U.S. subjects in the high performance standard case, 12 also named this as their major concern (‘‘I wanted to evaluate the products to see which had the greatest probability of a 25% profit’’). Among the Taiwanese and U.S. subjects who did not explicitly mention risk, the common concern was achieving the performance standard (‘‘The higher the

80

CHEE W. CHOW ET AL.

performance standard, the more pressure there is to make a good decision,’’ ‘‘You have to select a product that exceeds or meets your performance standard’’).

5. SUMMARY AND DISCUSSION
This experimental study has tested the effect of performance standard level, per se, on employees’ choice among risky projects. Based on goal setting theory and prospect theory, it hypothesizes that a high performance standard would induce greater risk taking by employees as compared with a low performance standard. It further hypothesizes that due to their purported higher collectivism (or, conversely, lower individualism), Chinese as compared with U.S. nationals would voluntarily take more risk to benefit the company at the expense of their personal interests. Our findings strongly supported the predicted relative effects of high and low performance standards. Across both national samples, subjects under a high performance standard selected significantly riskier projects than their counterparts in the low standard treatment. Responses to open-ended questions revealed that subjects facing both low and high performance standards were very much focused on the risk of failing to attain the standard. This was despite their pay being totally determined by actual outcome, rather than outcome relative to the standard. We interpret this finding as evidence that the performance standard, per se, influences subjects’ risk taking. We did not find significant differences in the risky choices of the two national samples. While our data are insufficient for determining the reason for this finding, we believe that it can be attributed, at least in part, to the interaction of two factors. First is the weakness of the treatment creating a firm–individual interest conflict, as reflected in the subjects’ lack of consideration for the firm’s desire for breakthrough products (thus the desirability of risk taking). Compounding this is the relative lack of difference in the two national samples’ degrees of individualism. This lack of difference reinforce calls for cross-national research to measure the participants’ cultural values (e.g., Gernon & Wallace, 1995; Harrison & McKinnon, 1999), instead of simply assuming that purported differences exist. In addition to the potential for a sample to be non-representative of the larger population, the increasing globalization of economic and cultural exchanges could be bringing about a convergence of work-related values. In the case of Taiwan, its close economic ties to the U.S. over the past several decades could have induced an acculturation to the latter’s work-related values, thereby blunting the

Performance Standards and Managers’ Adoption of Risky Projects

81

effects of cultural differences in management settings (HassabElnaby & Mosebach, 2005; Venezia, 2005). Further insights into the acculturation process and more broadly, the nature and magnitude of cultural effects will require engaging subjects from more divergent national cultures. Overall, the most important implication of our study is that performance standards, per se, can affect employees’ risk taking behavior at work. Given the importance of innovation for success and survival in an increasingly competitive and global marketplace, this finding can inform the design of systems for supporting innovation initiatives. At the same time, it is important to recognize the exploratory nature of the current study, hence the need for efforts to test the robustness of its findings. Like all laboratory studies, this study’s findings are a function of the experimental design, including nature of the task, the subjects, the salience ascribed by subjects to the experimental treatments, and the values of parameters (e.g., relation between pay and performance). There is much room for expanding and refining the experimental design. For example, the lack of salience for the company’s desire for breakthrough products may be due to the fictitious nature of the company, thus a lack of identification with its objectives. (In contrast, the performance standards may have been more salient because they directly affected the subjects’ personal performance evaluation.) There also is room for examining the nature of effects in a multi-period setting. When individuals make choices across multiple periods, they may view the outcomes across time as forming a portfolio for risk diversification. Also, performance standards often are used in tandem with performance-based compensation systems, wherein total compensation is tied to performance ` vis-a-vis the standard. It is desirable to study how these elements jointly affect risk taking by employees, and both expectancy and agency theories can provide guidance to such undertakings. And in light of the increasing use of teamwork in today’s organizations, it is worthwhile to examine if the effects of performance standards differ between individual and team settings. Even more broadly, management systems generally have many elements that work together as a whole, with some elements complementing, and others substituting, for one another (Chow, Kato, & Shields, 1994). By expanding the scope of analysis to include more facets of management systems, one can increase assurance that the findings are not subject to biases from important omitted variables. Finally, there is room to triangulate the investigation by using multiple methods (e.g., field experiments, in-depth case studies, surveys, and analysis of archival records). For example, while laboratory experiments are advantageous for determining causal relationships, they only allow a limited

82

CHEE W. CHOW ET AL.

number of variables to be examined simultaneously. In comparison, surveys permit many more variables to be included, but they tend to preclude detailed investigation of phenomena. Case studies are particularly suited to in-depth exploration of processes as well as the ‘‘how’’ and ‘‘why’’ of phenomena, but their generalizability tends to be limited by small feasible sample sizes. Given that each research method has its relative strengths and weaknesses, the use of multiple methods will help to increase the reliability and richness of findings (Birnberg, Shields, & Young, 1990). Such efforts are warranted in view of the topic’s importance to practice.

NOTES
1. The Taiwanese experimental materials were stated in New Taiwanese dollars (NT$). Taiwanese masters graduates’ starting salaries are about one third those of our U.S. graduates. Using the exchange rate of US$1 ¼ NT$30 approximately, the magnitudes of numbers in NT$ were ten times those in the U.S. instrument. 2. Because of the discrete probability plots used in our examples and experimental materials (for ease of understanding by the subjects), we were not able to make all seven distributions’ expected outcomes exactly 15%. The actual range of expected values was 14.85–15.08%, with product B having the highest expected value. This may have created a slight bias in favor of this low risk product, but as will be reported later, we still found a significant main effect due to performance standard level. 3. Our use of variance as a proxy for project risk is based on the voluminous literature in finance dealing with decision making under risk and uncertainty. It also is analogous to the approach of Chow and Haddad (1991) and Sayre et al. (1998), where outcome variability via operating leverage was used as the measure of project risk. Our original intent was to create a set of probability distributions with equally spaced variances, centered on the variance of the uniform distribution. Unfortunately, we were unable to achieve this result even after exhaustive efforts. As a result, we treat the riskiness of the subjects’ product choices as an ordinal, rather than a ratio scale. The hypothesis testing results were not qualitatively affected by whether we used the ordinal or ratio representations of project risk. 4. The subjects were informed that their experimental earnings would be converted into cash pay using a ratio of $5,000 in experimental earnings to $1. 5. A reviewer has noted that with all the products having the same expected value, a risk neutral firm would be indifferent among them. We acknowledge that it would have been more realistic to ascribe higher expected values to products with higher risk and as such, holding the expected value constant could bias our results against finding significant treatment effects. In this regard, our design was aimed at reducing the effects of omitted factors. If we had conferred higher expected values to higher risk projects, then the relation that we created between risk and return could have driven the results depending on how much ‘‘compensation’’ we had built in for higher risk. With the probability distributions we provided to the subjects, there still was a conflict of interests between individual managers and the firm. The managers were not compensated for adopting higher-risk projects with higher expected returns,

Performance Standards and Managers’ Adoption of Risky Projects

83

yet the firm could only have increased probabilities of very high outcomes by taking on more risk. As reported in the next section, despite facing equal expected values, our subjects still adopted more risky products when they were faced with higher performance standards. 6. The U.S. subjects earned an average of $15.125 while the Taiwanese subjects’ average earning was US$6.18. This proportional relationship approximates the 1:3 ratio between graduates’ beginning salaries in the two countries. 7. We did not ask about product risk because all subjects faced the same set of probability distributions. As reported in the next section, the subjects’ open-ended responses indicated that subjects in both performance standard treatments were highly conscious that risk was present in their product choices.

ACKNOWLEDGMENTS
An earlier version of this paper was presented at the American Accounting Association’s Management Accounting Section 2006 midyear conference. The authors are indebted to the discussant, David Otley, participants at the conference, and May Zhang for their helpful comments. They also thank the reviewers and editor for their guidance in improving the paper.

REFERENCES
Atkinson, A. A., Balakrishnan, R., Booth, P., Cote, J. M., Groot, T., Malmi, T., Roberts, H., Uliana, E., & Wu, A. (1997). New directions in management accounting research. Journal of Management Accounting Research, 9, 79–108. Awasthi, V., & Pratt, J. (1990). The effects of financial incentives on effort and decision performance: The role of cognitive characteristics. The Accounting Review, 65, 797–811. Awasthi, V. N., Chow, C. W., & Wu, A. (2001). Cross-cultural differences in the behavioral consequences of imposing performance evaluation and reward systems: An experimental investigation. International Journal of Accounting, 36(3), 83–109. Baiman, S. (1990). Agency research in managerial accounting: A second look. Accounting, Organizations and Society, 15, 341–371. Barney, J. B., & Hesterly, W. (1996). Organizational economics: Understanding the relationship between organizations and economic analysis. In: S. R. Clegg, C. Hardy & W. R. Nord (Eds), Handbook of organization studies (pp. 115–147). London: Sage. Baysinger, B., Kosnik, R. D., & Turk, T. A. (1991). Effects of board and ownership structure on corporate R&D strategy. Academy of Management Journal, 34, 205–214. Bellah, R., Madsen, R., Sullivan, W., Swidler, A., & Tipton, S. (Eds) (1987). Individualism and commitment in American life. New York: Harper and Row. Berg, J. E., Daley, L. A., Gigler, F., & Kanodia, C. (1990). The value of communication in agency contracts: Theory and experimental evidence. Vancouver, Canada: Canadian Certified General Accountants Research Foundation. Birnberg, J., Shields, M., & Young, M. (1990). The case for multiple methods in empirical management accounting research (With an illustration from budget setting). Journal of Management Accounting Research, 2(1), 33–66.

84

CHEE W. CHOW ET AL.

Birnberg, J., Turopolec, L., & Young, M. (1983). The organizational context of accounting. Accounting, Organizations and Society, 8(2), 111–129. Bond, M., & Hwang, K. (1986). The social psychology of Chinese people. In: M. Bond (Ed.), The psychology of the Chinese people (pp. 213–266). Hong Kong: Oxford University Press. Bond, M., Leung, K., & Wan, K. (1982). How does cultural collectivism operate? The impact of task and maintenance contributions on reward distribution. Journal of Cross-Cultural Psychology, 13(2), 186–200. Bonner, S. E., Hastie, R., Sprinkle, G. B., & Young, S. M. (2000). A review of the effects of financial incentives on performance in laboratory tasks: Implications for management accounting. Journal of Management Accounting Research, 12, 19–64. Bonner, S. E., & Sprinkle, G. B. (2002). The effects of monetary incentives on effort and task performance: Theories, evidence, and a framework for research. Accounting, Organizations and Society, 27, 303–345. Bouchikhi, H., & Kimberly, J. R. (2001). ‘It’s difficult to innovate’: The death of the tenured professor and the birth of the knowledge entrepreneur. Human Relations, 54(January), 77–84. Brown, T. C., & Latham, G. P. (2000). The effects of goal setting and self-instruction training on the performance of unionized employees. Relations Industrielles, 55(1), 80–93. Chatterjee, S., Wiseman, R. M., Fiegenbaum, A., & Devers, C. E. (2003). Integrating behavioral and economic concepts of risk into strategic management: The twain shall meet. Long Range Planning, 36(1), 61–79. Chinese Cultural Connection (1987). Chinese values and the search for culture-free dimensions of culture. Journal of Cross-Cultural Psychology, 18(2), 143–164. Chong, V. K., & Chong, K. M. (2002). Budget goal commitment and informational effects of budget participation on performance: A structural equations modeling approach. Behavioral Research in Accounting, 14, 65–86. Chow, C. W. (1983). The effects of job standard difficulty and compensation schemes on performance: An exploration of linkages. The Accounting Review, 58(4), 667–685. Chow, C. W., Cooper, J., & Waller, W. (1988). Participative budgeting: Effects of a truthinducing pay scheme and information asymmetry on slack and performance. The Accounting Review, 63(1), 111–122. Chow, C. W., Deng, J., & Ho, J. (2000). The openness of knowledge sharing within organizations: A comparative study of the United States and People’s Republic of China. Journal of Management Accounting Research, 12, 65–95. Chow, C. W., & Haddad, K. M. (1991). Relative performance evaluation and risk taking in delegated investment decision. Decision Sciences, 22(3), 583–593. Chow, C. W., Kato, Y., & Merchant, K. A. (1996). The use of organizational controls and their effects on data manipulation and management myopia: A Japan vs. U.S. comparison. Accounting, Organizations and Society, 21, 175–192. Chow, C., Kato, Y., & Shields, M. (1994). National culture and the preference for management controls: An exploratory study of the firm-labor market interface. Accounting, Organizations and Society, 19, 381–400. Demski, J. S., & Feltham, G. A. (1978). Economic incentives in budgetary control systems. The Accounting Review, 53(2), 336–359. Earley, C. (1993). East meets West meets Mideast: Further explorations of collectivistic and individualistic work groups. Academy of Management Journal, 36(2), 319–348.

Performance Standards and Managers’ Adoption of Risky Projects

85

Earley, C., & Gibson, C. B. (1998). Taking stock in our progress on individualism-collectivism: 100 years of solidarity and community. Journal of Management, 24(3), 265–304. Erez, M., & Earley, C. (1993). Culture, self-identity, and work. New York: Oxford University Press. Fama, E. F. (1980). Agency problems and the theory of the firm. Journal of Political Economy, 88, 288–307. Fatseas, V. A., & Hirst, M. K. (1992). Incentive effects of assigned goals and compensation schemes on budgetary performance. Accounting and Business Research, 22(88), 347–355. Fiegenbaum, A. (1990). Prospect theory and the risk-return association: An empirical examination in 85 industries. Journal of Economic Behavior and Organization, 14, 187–204. Fisher, J., Frederickson, J. R., & Peffer, S. A. (2002). The effect of information asymmetry on negotiated budgets: An empirical investigation. Accounting, Organizations and Society, 27, 27–43. Frederickson, J. R. (1992). Relative performance information: The effects of common uncertainty and contract type on agent effort. The Accounting Review, 67(4), 647–669. Garen, J. E. (1994). Executive compensation and principal-agent theory. Journal of Political Economy, 102, 1175–1199. Gernon, H., & Wallace, O. (1995). International accounting research: A review of its ecology, contending theories and methodologies. Journal of Accounting Literature, 14, 54–106. Gomez-Mejia, L. R., Welbourne, T. M., & Wiseman, R. M. (2000). The role of risk sharing and risk taking under gain sharing. Academy of Management Review, 25(3), 492–507. Harris, P., & Moran, R. (1987). Managing cultural difference. Houston, TX: Gulf Publishing. Harrison, G. L. (1993). Reliance on accounting performance measures in superior evaluative style: The influence of national culture and personality. Accounting, Organizations and Society, 18, 319–339. Harrison, G. L., & McKinnon, J. (1999). Cross-cultural research in management control systems design: A review of the current state. Accounting, Organizations and Society, 24, 483–506. HassabElnaby, H., & Mosebach, M. (2005). Culture’s consequences in controlling agency costs: Egyptian evidence. Journal of International Accounting, Auditing and Taxation, 14, 19–32. Hodgetts, R., & Luthans, F. (1997). International management. New York: McGraw-Hill. Hofstede, G. H. (1980). Culture’s consequences: International differences in work-related values. Beverly Hills, CA: Sage Publications. Hofstede, G. H. (1984). Cultural dimensions in management and planning. Asia Pacific Journal of Management, 1(2), 81–99. Hofstede, G. H. (1991). Cultures and organizations: Software of the mind. Berkshire, UK: McGraw-Hill. Holmstrom, B. (1979). Moral hazard and observability. Bell Journal of Economics, 10, 74–91. Holmstrom, B. (1982). Moral hazard in teams. Bell Journal of Economics, 13, 334–340. Hopwood, A. (1974). Accounting and human behavior. Englewood Cliffs, NJ: Prentice-Hall. Hopwood, A. G. (1972). An empirical study of the role of accounting data in performance evaluation. Journal of Accounting Research, 10(3), 156–182. Hoskisson, R. E., Hitt, M. A., & Hill, C. W. L. (1991). Managerial risk taking in diversified firms: An evolutionary perspective. Organization Science, 2(3), 296–314. Isaac, R. G., Zerbe, W. J., & Pitt, D. C. (2001). Leadership and motivation: The effective application of expectancy theory. Journal of Managerial Issues, 13(2), 212–226.

86

CHEE W. CHOW ET AL.

Jacobson, R. (1992). The ‘‘Austrian’’ school of strategy. Academy of Management Journal, 17, 782–807. Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305–360. Jensen, M. C., & Murphy, K. J. (1990). Performance pay and top-management incentives. Journal of Political Economy, 98, 225–264. Kachelmeier, S. J., & Shehata, M. (1997). Internal auditing and voluntary cooperation in firms: A cross-culture experiment. The Accounting Review, 72, 407–432. Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263–291. Koford, K., & Penno, M. (1992). Accounting, principal-agent theory, and self-interested behavior. In: N. E. Bowie & R. E. Freeman (Eds), Ethics and agency theory: An introduction (pp. 127–142). New York, NY: Cambridge University Press. Kuczmarski, T. D. (1996). Creating an innovative mind-set. Management Review, 85, 47–50. Kunz, A. H., & Pfaff, D. (2002). Agency theory, performance evaluation, and the hypothetical construct of intrinsic motivation. Accounting, Organizations and Society, 27, 275–295. Lachman, R., Nedd, A., & Hinings, B. (1994). Analyzing cross-national management and organizations: A theoretical framework. Management Science, 40(1), 40–55. Lambert, R. A. (2001). Contracting theory and accounting. Journal of Accounting and Economics, 32, 3–87. Latham, G. P. (2004). The motivational benefits of goal-setting. Academy of Management Executive, 18(4), 126–129. Leung, K., & Bond, M. (1984). The impact of cultural collectivism on reward allocation. Journal of Personality and Social Psychology, 47(4), 793–804. Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task performance. Englewood Cliffs, NJ: Prentice-Hall. Luft, J. (1994). Bonus and penalty incentives: Contract choice by employees. Journal of Accounting and Economics, 18(2), 181–206. Markus, H., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224–253. Merchant, K. (1998). Modern management control systems: Text and cases. Upper Saddle River, NJ: Prentice-Hall. Merchant, K. A., & Manzoni, J. F. (1989). The achievability of budget targets in profit centers: A field study. The Accounting Review, 64(3), 539–558. Mullins, J. W., Forlani, D., & Walker, O. C., Jr. (1999). Effects of organizational and decisionmaker factors on new product risk taking. The Journal of Product Innovation Management, 16, 282–294. Nouri, H., & Parker, R. J. (1998). The effect of organizational commitment on the relation between budgetary participation and budgetary slack. Behavioral Research in Accounting, 8, 74–90. Otley, D. T. (1987). Accounting control and organizational behavior. UK: William Heinemann Ltd. Page, A. L. (1993). Assessing new product development practices and performance: Establishing new norms. The Journal of Product Innovation Management, 10, 273–290. Parker, R., & Kohlmeyer, J. (2005). Organizational justice and turnover in public accounting firms: A research note. Accounting, Organizations and Society, 30, 357–369. Payne, J. W., Laughhunn, D. J., & Crun, R. (1980). Translation of gambles and aspiration level effects in risky choice behavior. Management Science, 26, 1039–1060. Piron, R., & Smith, L. R. (1995). Testing risk love in an experimental racetrack. Journal of Economic Behavior and Organization, 27, 465–474.

Performance Standards and Managers’ Adoption of Risky Projects

87

Prendergast, C. (1999). The provision of incentives in firms. Journal of Economic Literature, 37, 7–63. Redding, S. G. (1993). The spirit of Chinese capitalism. New York: Walter de Gruyter. Ronen, S., & Shenkar, O. (1985). Clustering countries on attitudinal dimensions: A review and synthesis. Academy of Management Review, 10(3), 435–454. Ruchala, L. (1999). The influence of budget goal attainment on risk attitudes and escalation. Behavioral Research in Accounting, 11, 16–46. Rutledge, R., & Harrell, A. (1993). Responsibility and framing of accounting information. International Journal of Management, 10, 300–313. Rutledge, R., & Harrell, A. (1994). The impact of responsibility and framing of budgetary information on group-shifts. Behavioral Research in Accounting, 6, 92–109. Sayre, T. L., Rankin, F. W., & Fargher, N. L. (1998). The effects of promotion incentives on delegated investment decisions: A note. Journal of Management Accounting Research, 10, 313–324. Schwartz, S. (1994). Cultural dimensions of values: Toward an understanding of national differences. In: U. Kim, H. Triandis, C. Kagitcibasi, S. Choi & G. Yoon (Eds), Individualism and collectivism: Theory, method, and application (pp. 85–119). Thousand Oaks, CA: Sage. Sharp, D., & Salter, S. (1997). Project escalation and sunk costs: A test of the international generalizability of agency and prospect theories. Journal of International Business Studies, 28, 101–121. Sitkin, S. B., & Pablo, A. L. (1992). Reconceptualizing the determinants of risk behavior. Academy of Management Review, 17, 9–38. Smith, P., Dugan, S., & Trompenaars, F. (1996). National culture and the values of organizational employees: A dimensional analysis across 43 nations. Journal of Cross-Cultural Psychology, 27(2), 231–264. Smith, P. B., Peterson, M. F., & Schwartz, S. H. (2002). Cultural values, sources of guidance, and their relevance to managerial behavior. Journal of Cross-Cultural Psychology, 33(2), 188–208. Sondergaard, M. (1994). Research note-Hofstede’s consequences: A study of reviews, citations and replications. Organization Studies, 15(3), 447–456. Sprinkle, G. B. (2000). The effect of incentive contracts on learning and performance. The Accounting Review, 75, 299–326. Sprinkle, G. B. (2003). Perspectives on experimental research in managerial accounting. Accounting, Organizations and Society, 28, 287–318. Stewart, W., Watson, W., & Carland, J. (1999). A proclivity for entrepreneurship: A comparison of entrepreneurs, small business owners and corporate managers. Journal of Business Venturing, 14, 189–214. Thaler, R. H., & Johnson, E. J. (1990). Gambling with the house money and trying to break even: The effects of prior outcomes on risky choice. Management Science, 36, 643–660. Tosi, H. L., & Gomez-Mejia, L. (1989). The decoupling of CEO pay and performance: An agency theory perspective. Administrative Science Quarterly, 34, 169–190. Triandis, H. (1989). The self and social behavior in differing culture contexts. Psychological Review, 96(3), 506–520. Triandis, H. (1995). Individualism and collectivism. Boulder, CO: Westview Press. Triandis, H., Bontempo, R., Villareal, M., Asai, M., & Lucca, N. (1988). Individualism and collectivism: Cross-cultural perspective on self-ingroup relationships, Journal of Personality and Social Psychology, (February), 323–338.

88

CHEE W. CHOW ET AL.

Trompenaars, F. (1994). Riding the waves of culture. New York, NY: Irwin. Tversky, A., & Kahneman, D. (1986). Rational choice and the framing of decisions. Journal of Business, 59(4), S251–S278. Tversky, A., & Kahneman, D. (1991). Loss aversion in riskless choice: A reference dependent model. Quarterly Journal of Economics, 107, 1039–1061. Venezia, G. (2005). Impact of globalization of public administration practices on Hofstede’s cultural indices. Journal of American Academy of Business, 6(2), 344–349. Vroom, V. (1964). Work and motivation. New York, NY: Wiley. Wagner, J. A., III. (1995). Studies of individualism-collectivism: Effects on cooperation. Academy of Management Journal, 38, 152–167. Waller, W. S. (1988). Slack in participative budgeting: The joint effect of a truth-inducing pay scheme and risk preferences. Accounting, Organizations and Society, 13, 87–98. Wan, D., Ong, C., & Lee, F. (2005). Determinants of firm innovation in Singapore. Technovation, 25, 261–268. Wiseman, R. M., & Bromiley, P. (1996). Toward a model of risk in declining organizations: An empirical examination of risk, performance and decline. Organization Science, 7, 524–543. Wiseman, R. M., & Gomez-Mejia, L. R. (1998). A behavioral agency model of managerial risk taking. Academy of Management Review, 23(1), 133–153. Young, S. M. (1985). Participative budgeting: The effects of risk aversion and asymmetric information on budgetary slack. Journal of Accounting Research, 23(2), 829–842. Young, S. M., Fisher, J. G., & Lindquist, T. M. (1993). The effects of intergroup competition and intragroup cooperation on slack and output in a manufacturing setting. The Accounting Review, 68, 466–481. Young, S. M., & Lewis, B. L. (1995). Experimental incentive-contracting research in managerial accounting. In: R. H. Ashton & A. H. Ashton (Eds), Judgment and decision making research in accounting and auditing (pp. 55–75). New York, NY: Cambridge University Press.

APPENDIX: EXPERIMENTAL MATERIALS (HIGH PERFORMANCE STANDARD VERSION)
An Overview Of Overall Procedures In this business simulation, you will be asked to make decisions in the role of a product manager working for XYZ Company. You will be paid based on the outcomes of your decisions in the simulation. The entire simulation will have three major stages: Stage One: You will be asked to read through a description of the XYZ Company, and the nature of a product manager’s job in the company. Then, you will work through practice exercises to solidify your understanding of this information.

Performance Standards and Managers’ Adoption of Risky Projects

89

Stage Two:

The simulation will proceed in the following manner: You will go through a total of three experimental periods. The first two are practice periods to help you understand the procedures and how your decisions will produce outcomes. You will not be paid for these two periods. The third period is the real experiment and you will earn cash pay based on the outcome of your product selection in this period.

(1) At the beginning of the first period, you will be given some information and then will be asked to select a product for manufacturing and marketing. (2) After you have written down your product selection on a form provided in your packet, you will come to front of the class and select a piece of paper from a bag. This piece of paper will indicate the actual return (profit rate) on the project you selected from the information in step (1). (3) You will write down on the product selection form the actual return (profit rate) on the product you had selected. (This would be the number written on the piece of paper you had pulled from the bag.) (4) You will return to your desk and use that profit rate to calculate your performance (outcome to XYZ Company) and amount of pay. (5) The second practice period will begin. You will repeat steps (1)–(4) again for this second trial period. (6) After completing the second practice period, you will begin the real period by repeating steps (1)–(4). Your actual pay ($) will be based on your decision in this period. Stage Three: After completing the real period, you will be asked to complete a questionnaire. When you hand in your materials to the researcher, he will verify how much you have earned in the simulation based on your decision in the real period. Keep the participant ID number slip and give it to the researcher when you have finished the questionnaire. Then the researcher will pay you.

Please note that your materials for the simulation will have a participant ID. The only purpose of this ID is to keep track of materials in the simulation, so that the correct amount of cash pay can be distributed to the right individuals. You are not asked to write down your name anywhere, and no attempt will be made to link the simulation outcomes with specific individuals.

90

CHEE W. CHOW ET AL.

Participant ID:_____ Background Information This business simulation will ask you to make decisions as a product manager working for XYZ Company. To prepare for the simulation, please carefully read through the following description of the company and the job of their product managers. XYZ Company is in the business of manufacturing and marketing novelty items. Because of the fad nature of its products, the entire life span of the company’s products, from selection for manufacturing/marketing to expiration of all market appeal, is 1 year. The task of selecting products to manufacture and market is delegated to a number of product managers. Each product manager oversees his/her own staff of product developers, whose job is to design or seek out potential new products. At the beginning of each year, the product development staff for each product manager will propose to their manager seven new products. The proposal for each product will include the staff’s estimate of the probability that specific rates of return (profit rate) can be achieved by the product over its 1 year life. At the beginning of each year, each product manager is allocated $1,000,000 for that year. He/she then uses this allocated amount to manufacture and market one product out of the seven that his/her staff proposes (all products cost $1,000,000 to manufacture and market). Because of the transience of tastes, proposed products that are rejected in 1 year typically would have no market potential in a later year. Thus, each product manager’s staff has to propose seven new products each year. As a product manager for the XYZ Company, you are paid an annual base salary of $125,000 plus a profit-based ‘‘adjustment’’ in each funding cycle. This adjustment can be positive or negative, and is equal to 50% of the amount by which a product manager’s realized profit exceeds or falls short of his performance standard. To illustrate with some simple numbers, suppose that product manager A had been allocated $1,000 in a given yearly cycle, and that he had used these funds to manufacture and market a product. Assume that his performance

Performance Standards and Managers’ Adoption of Risky Projects

91

standard is set at a profit rate of 10% on the amount of his annual allocation. Further, suppose that the actual profit rate on the project is 15%. The total amount of profit from this product would be $150 ($1,000 Â 15%). After subtracting 10% of $1,000, or $100, the excess profit generated by product manager A is $50. Product manager A would be paid, on top of his base salary, an additional $25 ( ¼ 50% of the $50 overage). Alternately, suppose that the actual profit rate achieved on the product was 5%, such that the total amount of profit generated was $50. In this case, the total profit would be $50 short of the required 10% return on the $1,000, and manager A would be paid his base salary minus $25 ( ¼ 50% of the $50 shortfall). The Company does not impose an upper limit to the additional pay that a product manager can earn by generating profit in excess of his/her performance standard. It also does not limit the amount of negative profitbased adjustment that a product manager may face. The XYZ Company has been in existence for a number of years and on average, its products have produced profits equal to 15% of what it costs to manufacture and market them. But because the industry is intensely competitive and getting more so, the Company believes that if it is to survive and prosper, it must come out with a number of breakthrough products that can produce significantly higher levels of profits than this historical level. Having a number of products that stand out in the market will help to increase the company’s name recognition among customers, increase the willingness of retailers to stock and prominently display its products, and also strengthen the Company’s financial ability to support development and improvement. The Company recognizes that trying to achieve breakthrough profit performance likely will require taking risks. But because the Company can diversify away much of the risk of individual products due to having a number of product managers, it is not concerned with individual managers selecting highly risky products. Ã Ã Ã Ã Ã Ã Ã Ã Ã Ã Illustration of Products’ Profit Rate Distributions As a project manager, each year you will select one of seven products proposed by your product development staff. The examples below (Products X, Y, and Z) illustrate the three general forms that a product’s possible

92

CHEE W. CHOW ET AL.

profit rates may be distributed and each rate’s probability of being realized (in other words, a probability distribution). Product X’s probability distribution is often referred to as a bell-shaped curve, with one possible outcome (in the middle) being the most likely to be realized, and outcomes on both the higher and lower sides of it being progressively less likely. Product Y’s probability distribution is called a uniform distribution. All possible outcomes are equally likely. Finally, Product Z’s distribution of possible profit rates is called a bi-modal distribution. The values in the middle are the least likely to be realized, while outcomes toward the higher and lower extremes are progressively more likely to be realized. The profit rate distributions of Products X and Y are on the next page. The distribution for Product Z is on the page following the next page. As you probably already know, the total of all the possible outcomes’ probabilities is 100%. In the graph for each of the products, the horizontal axis (the line at the bottom) will be clearly labeled with each of the possible outcomes in terms of profit rates, only one of which will be realized. On the vertical axis (the vertical line on the left hand side) will be the probability of each specific profit rate being realized.

Product X 7% 6% Probability 5% 4% 3% 2% 1% 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29% Profit Rates

Performance Standards and Managers’ Adoption of Risky Projects

93

Product Y 6.0% 5.0% Probability 4.0% 3.0% 2.0% 1.0% 0.0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29% Profit Rates Product Z 7% 6% 5% Probability 4% 3% 2% 1% 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29% Profit Rates

To further illustrate, suppose that you wanted to ascertain the probability of a particular outcome (say, a 10% profit rate) being realized by a given product. If you looked at the profit rate probability distribution for Product

94

CHEE W. CHOW ET AL.

X, the answer would be 4%. For Product Y, the answer would be approximately 3.5%, and for Product Z, the answer would be 3%. If what you wanted to know is the probability that the realized profit rate would be at least some level (i.e., that level and above), you would add up the individual probabilities of each profit rate from that profit rate and up. Again, let us assume that you wanted to know the probability of the realized profit rate being at least 10%. For Product X, the answer would be 74%. (This is the sum of the probability of the 10% profit rate, plus that of the 11% profit rate, y , plus the probability of the 29% profit rate.) For Product Y, the answer would be approximately 68.5%, and for Product Z, the answer would be 60.5%. On the other hand, if what you are interested in is the probability that the realized outcome would be below some level, you would add up the individual probabilities of each profit rate below (to the left of) this outcome level. Practice Exercises For the results of this simulation to be meaningful, it is crucial that you fully understand the role of product managers in the XYZ Company, how their decisions affect the company, and how they are evaluated and paid. Please work through the following two practice examples when you feel confident that you fully understand all of the information provided above. Please feel free to re-read any of the preceding materials and to take all the time that you need. You also may refer back to the preceding materials while working on the practice examples. First, to make sure that you fully understand the information that your product development staff will provide you each period, please work through the follow practice exercises for the three types of probability distributions. Please write down your answers in the spaces provided. The Probability Distribution for Product X: (a) What is the probability that this product will produce an actual profit rate equal to 10%?__________ (b) What is the probability that this product will produce an actual profit rate equal to or above 20%?__________ (c) What is the probability that this product will produce an actual profit rate below 5%?__________

Performance Standards and Managers’ Adoption of Risky Projects

95

The Probability Distribution for Product Y: (a) What is the probability that this product will produce an actual profit rate equal to 10%?__________ (b) What is the probability that this product will produce an actual profit rate equal to or above 20%?__________ (c) What is the probability that this product will produce an actual profit rate below 5%?__________ The Probability Distribution for Product Z: (a) What is the probability that this product will produce an actual profit rate equal to 10%?__________ (b) What is the probability that this product will produce an actual profit rate equal to or above 20%?__________ (c) What is the probability that this product will produce an actual profit rate below 5%?__________ When you are satisfied that you have the correct answers to these questions, please open the next page (which is taped shut) to check the answers. If some of your answers above differ from the correct answers, please go back to examine why this had occurred. Correct Answers to the Practice Exercises Product X: (a) 4% (This is the probability indicated for the profit rate of 10%.) (b) 30% (This is the sum of the probabilities for the profit rate of 20% as well as all the profit rates above it, i.e., to its right.) (c) 10% (This is the sum of the probabilities of all the profit rates below 5%.) Product Y: (a) 3.5% (This is the probability indicated for the profit rate of 10%.) (b) 35% (This is the sum of the probabilities for the profit rate of 20% as well as all the profit rates above it, i.e., to its right.) (c) 14% (This is the sum of the probabilities of all the profit rates below 5%.) Product Z: (a) 3% (This is the probability indicated for the profit rate of 10%.) (b) 42.5% (This is the sum of the probabilities for the profit rate of 20% as well as all the profit rates above it, i.e., to its right.) (c) 21.5% (This is the sum of the probabilities of all the profit rates below 5%.)

96

CHEE W. CHOW ET AL.

The next two practice examples focus on how your product selection decisions will translate into realized profits for the Company, as well as the total amount that you will be paid by the Company for this performance. For both practice examples below, please assume that as a product manager, you are paid an annual base salary of $100, plus an adjustment (+ or À) equal to 50% of the amount by which your realized profit exceeds or falls short of your performance standard. Your performance standard is 10% of the amount of funds you have been allocated for the year. The magnitudes of the numbers in these examples are not the ones that will be used in the experiment. They have been chosen to simplify the necessary calculations. Practice Example One. Suppose that you had been allocated $1,000 for a given year. The actual profit rate you had achieved from the product you selected was 5%. Based on this information, what would be the total profit you had generated for the Company for this year? $_________ How much total pay would you receive from the Company for this year? $__________________. Practice Example Two. Suppose that you had been allocated $1,000 for a given year. The actual profit rate you had achieved from the product you selected was 15%. Based on this information, what would be the total profit you had generated for the Company for this year? $_________ How much total pay would you receive from the Company for this year? $__________. When you are satisfied that you have the correct answers to these questions, please open the next page (which is taped shut) to check the answers. If some of your answers above differ from the correct answers, please go back to examine why this had occurred.

Performance Standards and Managers’ Adoption of Risky Projects

97

Correct Answers to the Practice Examples Practice Example One. Suppose that you had been allocated $1,000 for a given year. The actual profit rate you had achieved from the product you selected was 5%. Based on this information, what would be the total profit you had generated for the Company for this year? ANSWER: $50 ( ¼ 5% times $1,000). How much total pay would you receive from the Company for this year? ANSWER: $75. Explanation: You would be paid your base salary of $100, minus an adjustment equal to 50% of your profit shortfall. The amount of the shortfall is $50, which is the difference between the actual profit realized of $50 (5% times $1,000) and your performance standard of 10% of $1,000, or $100. Thus, in total, your pay would be $100 – (50% of $50 shortfall), or $75. Practice Example Two. Suppose that you had been allocated $1,000 for a given year. The actual profit rate you had achieved from the product you selected was 15%. Based on this information, what would be the total profit you had generated for the Company for this year? $150 ( ¼ 15% times $1,000). How much total pay would you receive from the Company for this year? ANSWER: $125. Explanation: You would be paid your base salary of $100, plus an adjustment equal to 50% of your profit overage. The amount of the overage is $50, which is the difference between the actual profit realized of $150 (15% times $1,000 and your performance standard of 10% of $1,000, or $100. Thus, in total, your pay would be $100+(50% of the $50 overage), or $125. Ã Ã Ã Ã Ã Ã Ã Ã Ã Ã Now we are almost ready to start the simulation. Before we do that, please go back and carefully re-read all the background information about the Company, your role as a product manager, how your product choices affect the Company as well as your own pay. We will not start the simulation until you have done so. Ã Ã Ã Ã Ã Ã Ã Ã Ã Ã

98

CHEE W. CHOW ET AL.

Participant ID:______ Business Simulation Please assume the role of a product manager of the XYZ Company. In each of three periods, you will be asked to make decisions in this role. There will be two practice periods and a ‘‘real’’ period. You will be paid cash based on the outcome of your decision in the ‘‘real’’ period. All of the preceding information, including product managers’ responsibilities, the company’s situation and desires, and how product managers are evaluated and paid, will apply. The only difference is that your base pay in each yearly period is $125,000 rather than the $100 used in the practice examples. The 50% adjustments to pay for excess or deficient net income are the same as in those examples. At the end of the third period (‘‘real’’ period), your total pay earned in the third period will be converted into cash pay at a rate of $5,000 in experimental pay equals $1. (Thus, experimental pay of $50,000 will convert into cash pay of $10.) Please permit us to emphasize that for the results of this simulation to be meaningful, it is absolutely necessary that you act in this simulation as if you are facing a situation in real life. Please do your best to do so. Thank you. Information for Period One The company has allocated $1,000,000 to you for product manufacturing and marketing in this period. Your base pay is $125,000 and your performance standard is a 25% profit rate. Your product development staff has proposed seven new products for this period. The probability distributions of profit rates for these products are in Products A1–G1. These probability distributions are stapled in a packet named ‘‘Probability Distributions of Profit Rates for Products in Period One.’’

Performance Standards and Managers’ Adoption of Risky Projects 99

Probability Distributions of Profit Rates for Products in Period One
Product A 0.14 0.12 0.10 Probability 0.08 0.06 0.04 0.02 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29% Profit Rates Product B 0.12 0.10 Probability 0.06 0.04 0.02 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29% Profit Rates 0.08

Product C 0.07 0.06 0.05 Probability 0.04 0.03 0.02 0.01 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29% Profit Rates

100

Probability

Probability
Probability 0.01 0.02 0.03 0.04 0.05 0.06 -

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

-

1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29%

1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29%

1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29%

Profit Rates

Profit Rates

Product E

Product D

Product F

Profit Rates

CHEE W. CHOW ET AL.

Performance Standards and Managers’ Adoption of Risky Projects
Product G 0.12 0.10 Probability 0.08 0.06 0.04 0.02 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29%

101

Profit Rates

Participant ID:______ Decision and Recording Form for Period One (Same for Periods Two and Three) I have decided as follows (Please check the box representing the product of your choice): Product I will manufacture and market:
A1 B1 C1 D1 E1 F1 G1

Now please walk to the front of the class and pull a piece of folded paper from the bag of your project choice. The bag for each product contains 100 pieces of folded paper, with proportions corresponding to the probabilities provided by your product development staff. (Thus, a given profit rate that has a 9% probability of occurring has nine folded pieces of paper containing this particular profit rate.) Please open the piece of folded paper and show it to the researcher. Write down in the space below the realized profit rate that is written on this piece of paper. Realized profit rate: __________%

102

CHEE W. CHOW ET AL.

Then, please return to your seat and complete the rest of this form using your performance standard and how your pay is determined by the Company.

The amount of actual profit my selected product has generated for the Company: $_____________________________________ The total amount I will be paid by the Company: $_____________________________________

After you have completed your calculations for this period, please proceed to the next period. Participant ID:______ Exit Questionnaire DIRECTIONS: Please complete the questionnaire below, answering each question as accurately and honestly as you can. Remember, all answers are anonymous and cannot be traced back to you in any way. 1. 2. 3. 4. Two digit identification number (on front of brown packet) _________ Gender (circle one): M/F Age ________ Years of full time equivalent working experience _______ years

For each of the following statements, please circle the number that best describes your actions and beliefs in the experiment. (1) In the ‘‘real’’ period of this business simulation, when you made decisions in the role of a product manager in XYZ Company, to what extent did you act as if you were facing a situation in real life? (Please circle one number.)

1 Not at all

2

3

4

5

6

7

8

9 Completely

Performance Standards and Managers’ Adoption of Risky Projects

103

(2)

To what extent did your performance standard affect your product selection for the real period?
1 Not at all 2 3 4 5 6 7 8 9 Completely

(3)

If your numerical answer to (2) was a number greater than one, please briefly describe how the performance standard had affected your product selection decision: _______________________________________________________ _______________________________________________________ To what extent was your product selection for the real period influenced by the way that your pay depended on the actual outcome from your selected product?
1 Not at all 2 3 4 5 6 7 8 9 Completely

(4)

(5)

If your numerical answer to (4) was a number greater than one, please briefly describe how your product selection decision was affected by how the company determined your pay based on the actual outcome from your selected product: _______________________________________________________ _______________________________________________________ To what extent did your company’s desire for break through/stand out products affect your product selection for the real period?
1 Not at all 2 3 4 5 6 7 8 9 Completely

(6)

(7)

If your numerical answer to (6) was a number greater than one, please briefly describe how the company’s desire for break through/stand out products had affected your product selection decision: _______________________________________________________ _______________________________________________________

104

CHEE W. CHOW ET AL.

(8)

Below are six items related to the nature of one’s job. Please indicate how important each of them is to you personally. (Please check a box for each item.)
Of little or no Importance Of moderate Importance Of little Importance Of utmost Importance Of Supreme Importance Very Important

Have a job that leaves you sufficient time for your personal or family life. Have considerable freedom to adopt your own approach to the job. Have challenging work to do-work for which you can get a personal sense of accomplishment. Fully use your skills and abilities on the job. Have good physical working conditions (good ventilation and lighting, adequate workspace, etc.) Have training opportunities (to improve your skills or to learn new skills).

(9)

Please indicate how important each of the following is to you personally. (Please check a box for each item.)
Of no Importance at all Of Moderate Importance

(1) Tolerance of others Harmony with others Solidarity with others Non-competitiveness Trustworthiness Contentedness with one's position in life Being conservative A close, intimate friend Filial piety (Obedience to parents, respect for parents, honoring of ancestors, financial support of parents) Patriotism Chastity in women

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

Performance Standards and Managers’ Adoption of Risky Projects

105

(10) Please indicate your agreement or disagreement with each of the following statements. (Please check a box for each item.)
Strongly Disagree Disagree Strongly Agree Neutral Agree

Employees like to work in a group rather than by themselves. If a group is slowing me down, it is better to leave it and work alone. To be superior, a man must stand alone. One does better work working alone than in a group. I would rather struggle through a personal problem by myself than discuss it with my friends. An employee should accept the group's decision even when personally he or she has a different opinion. Problem solving by groups gives better results than problem solving by individuals. The needs of people close to me should take priority over my personal needs.

This is the end of the questionnaire. Thank you.

THE EFFECTS OF ORGANIZATIONAL CULTURE ON BUDGETARY CONFLICT: INTEGRATIVE VERSUS DISTRIBUTIVE CONFLICT RESOLUTION
Nabil Elias and William W. Notz
ABSTRACT
Like conflict in general, budgetary conflict is perceived by conflicting parties as a zero-sum game or distributive: one party’s gain is the other party’s loss. We identify an organizational culture that promotes this view as ‘‘traditional.’’ We propose that changing certain elements of organizational culture is sufficient to produce more integrative, nonzero-sum outcomes. We call this changed organizational culture ‘‘empowering.’’ We propose and test the effects of an empowering organizational culture (EOC) in contrast to the traditional organizational culture (TOC). We hypothesize that an EOC would produce more integrative conflict resolution than the typical TOC. Based on our review of the literature, we identify two elements of the EOC that are essential in producing more

Advances in Management Accounting, Volume 16, 107–140 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16003-2

107

108

NABIL ELIAS AND WILLIAM W. NOTZ

integrative solutions to budgetary conflict. The two elements that we simultaneously manipulate are the superior’s empowering style (or lack thereof) as reflected in encouragement to freely negotiate, and the superior’s intervention process in failed negotiations (a process that encourages the search for integrative solutions and avoids imposed compromises that dampen the desire to negotiate). Using a laboratory experiment, 84 subjects forming 42 dyads negotiated the allocation of discretionary budgets face-to-face. The results of the experiment confirm our hypotheses that the EOC produces more integrative budget negotiation outcomes, greater convergence, and greater satisfaction with the outcome than TOC.

1. BUDGETARY CONFLICT AND THE PRODUCTION OF INTEGRATIVE RESOLUTION
Budgetary conflict is inherent in many organizations and has been studied from different perspectives including budgetary slack in hierarchical relationships (e.g., Schiff & Lewin, 1970; Chow, Cooper, & Waller, 1988), the setting of managerial goals (Etherington & Tjosvold, 1992), the effects of information asymmetry on the budgeting negotiation process and budget slack (Fischer, Fredrickson, & Peffer, 2002a), and the effects of using budgets for performance evaluation (Fischer, Maines, Peffer, & Sprinkle, 2002b). Negotiation research in accounting is relatively recent (Dejong, Forsythe, Kim, & Uecker, 1989; Elias, 1990; Chalos & Haka, 1990; Anctil & Dutta, 1999; Kachelmeier & Towry, 2002; Ghosh, 2000) and negotiation research in budgeting is fairly limited (Fischer et al., 2002a, 2002b). The cost of conflict in organizations can be considerable (Slaikeu & Hasson, 1998), and most of that cost is not explicit. Like conflict in general, budgetary conflict is perceived by conflicting parties as a zero-sum game or distributive: one party’s gain is the other party’s loss. We identify an organizational culture (OC) that promotes this view as ‘‘traditional.’’ We propose that changing certain elements of OC is sufficient to produce more integrative, nonzero-sum outcomes. We call this changed OC ‘‘empowering.’’ Managing conflict to increase integrative resolution in organizations has not received much attention in the literature – much less managing budgetary conflict. Conflict does not have to be costly; on the contrary, it could stimulate the search for integrative agreements (Baron, 1990; Deutsch, 1990). An integrative agreement occurs when a resolution is found that produces high

The Effects of Organizational Culture on Budgetary Conflict

109

joint utility for the parties in conflict. We propose and test the effects of an empowering organizational culture (EOC) in contrast to the traditional organizational culture (TOC). We hypothesize that an EOC would produce more integrative conflict resolution than the TOC. Based on our review of the literature, we identify two major elements of the EOC that are essential in producing more integrative solutions to budgetary conflict. Compared to the TOC, the two major elements that we simultaneously manipulate are the superior’s empowering style, as reflected in encouragement to freely negotiate, and the superior’s intervention process in failed negotiations (a process that encourages the search for integrative solutions and avoids imposed compromises that dampen the desire to negotiate). Integrative resolution of budgetary conflict can decrease explicit and implicit costs of conflict and yield positive outcomes. This paper focuses on how changes in these two major elements of the TOC may produce such integrative budgetary behavior. Organizational scholars assert that conflict in organizations is both inevitable and prevalent (Thomas, 1992; Bazerman and Neale, 1983). Axelrod (1984) suggests that coordination between conflicting organizational subunits can be achieved by allowing them to bargain directly with each other. Accordingly, we propose that a managerial style that empowers subunits to negotiate their conflict and expects them to resolve their conflict on their own is likely to reach more integrative solutions as in the EOC treatment. On the other hand, in a TOC when negotiators fail to reach an agreement, they expect intervention by superiors who usually impose a compromise solution. The expectation of a compromise solution naturally discourages the search for integrative solutions and dampens the motivation to negotiate in the first place. We simultaneously manipulate these two elements of OC in our experiment: superior’s managerial style (empowering vs. traditional) and superior’s intervention process in failed negotiations, and treat them as one composite variable. We examine the effect of two OCs: the ‘‘empowering culture’’ and the ‘‘traditional culture’’ on negotiation outcomes of allocating discretionary budget amounts in a laboratory experiment setting. The remainder of this paper is organized into five sections. Section 2 reviews the development of the budgetary OC variable and provides a discussion of two types of negotiation outcomes: distributive and integrative. Section 3 develops the hypotheses. Section 4 describes the experimental design. Section 5 provides the results of the experiment. Finally Section 6 provides a discussion of the results and conclusions, including limitations of the study.

110

NABIL ELIAS AND WILLIAM W. NOTZ

2. THE CONFLICT ENVIRONMENT AND NEGOTIATION OUTCOME
2.1. Organizational Culture The conflict environment is inextricably linked to OC. Even though there is considerable debate about the meaning of OC (Martin, 1991; Ott, 1989; Morrill, 1995), Hofstede (2003) and Hofstede, Neuijen, Ohayv, and Sanders (1990) have demonstrated empirically that OC can be used to meaningfully describe differences between organizations. The OC surrounding conflict and leadership are inseparable; how leaders view the organization largely determines the OC surrounding conflict (Schein, 1992). We focus on aspects of OC surrounding conflict related to leadership style and intervention process and hypothesize that different OCs are likely to affect the way organizations deal with budget conflict. Kolb and Sheppard (1985) suggest that organizations dominated by hierarchical authority would likely discourage bargaining activity between subunits. If allowed to negotiate, parties in a traditional OC would likely feel less empowered to resolve the conflict. On the other hand, an empowering OC would generally be characterized by norms of teamwork, decentralized hierarchical structures, consultation, and negotiation. In such organizations, negotiation and bargaining are encouraged and desirable, reflecting senior management’s belief in the strength of negotiated solutions (Schein, 1992). The empowerment would allow conflicting parties to deal directly with their conflict rather than rely on an expected compromise by their superior. We treat the superior’s managerial style and the superior’s intervention process in failed negotiations as one composite OC variable, because we anticipate that the presence of one is unlikely in the absence of the other. For example, it would be inconsistent that an OC that empowers subunit negotiation would at the same time use an intervention process that introduces a predictable riskless compromise if negotiations fail.1 Our treatment of OC as ‘‘empowering’’ or ‘‘traditional’’ is consistent with the literature on OC, in particular Hofstede et al. (1990), who describe an organization’s communication climate as open or closed, and its internal (hierarchical) structure as loose or tight. 2.2. Negotiation Outcomes Distributive bargaining within organizations occurs when the conflicting subunits perceive the outcome to be negatively correlated; one party’s gain is

The Effects of Organizational Culture on Budgetary Conflict

111

the other’s loss. It implies bargaining over finite resources such as an organization’s discretionary budget, and a zero-sum game where an increase in benefit to one subunit necessarily means a decrease in that benefit to another. The tendency for subunits to bargain distributively will vary with the extent to which their interests are seen as finite and with the nature and amount of costs they expect if the conflict cannot be bargained to resolution. As opposed to the zero-sum nature of distributive bargaining, integrative bargaining includes positive-sum elements; the divergent needs of the subunits in conflict are ‘‘integrated’’ (Follett, 1940; Walton & McKersie, 1965) and the agreements are tantamount to achieving the greatest good for the bargainers (Pruitt & Lewis, 1975). It requires recognition of a potential common ground and the possibility of alternatives that are superior to a simple distributive (compromise) resolution of the conflict (Butler, 1994). The possibility that novel solutions will be produced makes conflict potentially valuable rather than damaging to the subunits and the organization. In other words, integrative bargaining is a behavior that unlocks the constructive potential of conflict. Walton and McKersie (1965) postulate that one of the conditions of integrative bargaining is that several issues be considered simultaneously such that trade-offs can be made between bargainers. Several researchers tested this trade-off, called ‘‘logrolling,’’ including Froman and Cohen (1970), Pruitt and Lewis (1975), and Thompson and Hastie (1990). For example, in a classic study, Pruitt and Lewis (1975) used a logrolling methodology that included negotiating more than one factor having different values to negotiators, thus allowing integrative resolution. While these studies addressed integrative behavior, none of them dealt with intraorganizational conflict or budgeting contexts. Previous research indicates that the quality of outcomes depends on negotiators’ perceptions of the conflict as a ‘‘fixed pie’’ problem (Lax & Sebenius, 1986; Thompson, 1990; Thompson & Lowenstein, 1992). The ‘‘fixed pie’’ perception leads to inadequate search for information about opponent preferences and error in the interpretation of available information (Bazerman & Neale, 1992; Carnevale & Lawler, 1986; Carroll, Bazerman, & Maury, 1988; Neale & Bazerman, 1983; Pinkley, Griffith, & Northcraft, 1995; Thompson, 1990; Thompson & Hastie, 1990). As Pinkley et al. (1995) observe, alleviating these biases does not necessarily produce integrative agreements unless the ‘‘fixed pie’’ expectation of individual negotiators is altered. In this study, we examine dyadic budgetary conflict in the context of different OCs.

112

NABIL ELIAS AND WILLIAM W. NOTZ

One aspect that affects negotiation outcomes is third party intervention, or threat of intervention, in the event of failed negotiations. Different third party intervention processes, or threat of intervention, in the context of labor negotiations produced different effects on settlements (Notz & Starke, 1978). Elias and Ezzamel (1990) suggest that information search by negotiators subject to final offer arbitration intervention, where an arbitrator selects one or the other of the final offers, may be greater than information search by negotiators subject to conventional arbitration intervention, where an arbitrator imposes a compromise settlement usually between the two positions. These studies, however, were confined to one-factor negotiation (e.g., wage increase) that did not allow for logrolling. As a component of the OC, we introduce the application of different types of intervention in the context of intra-organizational budget conflict. Conventional intervention usually splits the difference and shields negotiators from failed negotiations; indeed it leads conflicting parties to take extreme positions to protect their final outcome. Final offer intervention forces negotiators to explore, probe, and reach an agreement to avoid the risk of their opponent’s final offer being selected. In summary, we use multiple budget items of different value to each negotiator to allow for logrolling and integrative solutions. We simulate the OC by simultaneously varying (a) the managerial style empowering or not empowering subunit negotiation, and (b) the managerial intervention process used in the case of failed negotiations. Based on our discussion of OC and negotiation outcome, we develop our hypotheses.

3. HYPOTHESES
A traditional OC reflects a hierarchical environment that attempts to control outcomes. If negotiation between subunits takes place, failed negotiations are virtually riskless if negotiating subunits take extreme positions, or in the words of Argyris (1973) ‘‘not owning up.’’ On the other hand, an empowering OC encourages subunit cooperation and genuine negotiation. If negotiation takes place, the risk of failed negotiations could be substantial to the negotiating subunits, which leads them, in the words of Argyris (1973), to ‘‘owning up.’’ This contrast between the two negotiating environments is summarized in Fig. 1. We hypothesize that a TOC would produce a bargaining process characterized by a distributive zero-sum orientation, constrained information exchange, low levels of trust, and a bargaining outcome that is relatively

The Effects of Organizational Culture on Budgetary Conflict

113

Traditional Organizational Culture (TOC) Hierarchical order; controlling Riskless failed negotiation Not owning up Distributive negotiation outcome

Empowering Organizational Culture (EOC) Delegation; freedom to negotiate Substantial burden & risk of failed negotiation Owning up Integrative negotiation outcome

Fig. 1.

Contrast of Two Types of Organizational Culture.

distributive in nature. In contrast, an EOC would produce a bargaining process characterized by an integrative positive-sum orientation, information sharing, trust, and a bargaining outcome that is relatively integrative in nature. The specific hypotheses are listed below. 3.1. Budget Negotiation Outcomes H1. The EOC treatment will produce more integrative budget negotiation outcomes than the TOC treatment.2 We test this hypothesis from two angles as follows: H1A. The EOC (TOC) treatment will result in more (fewer) integrative agreements and fewer (more) distributive agreements3 and arbitrated settlements. H1B. The EOC (TOC) treatment will result in higher (lower) benefit to negotiators. 3.2. Budget Negotiation Expectations H2. The EOC (TOC) treatment will exhibit more (less) realistic expectations and greater (lesser) dyadic budget convergence at different stages of the negotiation process. We test this hypothesis from two angles as follows: H2A. The EOC (TOC) treatment will exhibit more (less) realistic expectations of budget settlement at different stages of the negotiation process. H2B. The EOC (TOC) treatment will exhibit greater (lesser) dyadic convergence at different stages of the budget negotiation process.

114

NABIL ELIAS AND WILLIAM W. NOTZ

3.3. Budget Negotiation Perceptual and Cognitive Variables H3. The EOC (TOC) treatment will result in relatively more (less) positive perceptions related to trust, cooperation, and information sharing.

4. EXPERIMENTAL DESIGN
We test our hypotheses with experimental data from a simulated organization with two different OCs in a laboratory setting.

4.1. Subjects Eighty-four undergraduates from a School of Business at a major public university took part in this experiment. All subjects were recruited volunteers who received payment for their participation in the study. The 84 subjects formed 42 negotiating dyads.

4.2. Task All subjects participating in the experiment received a description of a university that had received a substantial cut in its annual funding grant along with instructions from the funding authority, the University Grants Commission (UGC), to employ a new approach of internal budget reallocation. The UGC instructed the university that it must seek a synergistic reallocation of the budget cut by initially making a cut to mega units, consisting of two faculties (schools) that seemed to have some significant potential for synergistic reallocation of resources between them. Subjects were told that the university’s response to the instructions from the UGC was first to create mega units. The mega unit of interest consisted of the schools of Agriculture and Engineering. Each subject was assigned the role of a dean of one of these two schools by a toss of a coin. Each subject playing the role of one of the deans received instructions from his/her superior, the university vice president. Aside from forming a mega unit, the university’s response to the instructions from the UGC varied as part of the manipulation of the organization’s corporate culture. Each subject received a voucher to receive $55 and information explaining that the amount the subject would eventually receive as pay for the

The Effects of Organizational Culture on Budgetary Conflict

115

experiment depended on bargaining performance with the other subject. Each subject received the information that the mega unit in question had to absorb an aggregate cut of two million dollars. Each subject received an identical schedule (Table 1) that showed the relationship between the size of the budget cut for the subject’s school and the amount deducted from the subject’s voucher. We provided identical schedules to create distributive Table 1. How Budget Settlements Affect Subjects’ Pay Schedule of Budget Cuts and Personal Consequences for You.
Budget Cut for Your Faculty Column 1 $2,000,000 1,900,000 1,800,000 1,700,000 1,600,000 1,500,000 1,400,000 1,300,000 1,200,000 1,100,000 1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0 Example Amount of Budget Cut Dollars Deducted from Your Voucher More than À$55 À$55 À$21 Value of Your Voucher Your Pay Dollars Deducted from Your Voucher Column 2 (Confidential) À$110 À110 À107 À104 À100 À95 À89 À82 À74 À65 À55 À45 À36 À28 À21 À15 À10 À6 À3 À1 0

More than $1,000,000 $1,000,000 $600,000

$55 $55 $55

You get nothing You get nothing You get $34

116

NABIL ELIAS AND WILLIAM W. NOTZ

tension between bargainers; a gain by one subject was a loss to the other. Each subject received various items of information about the subject’s school, such as the consequences of a substantial budget cut on that school, and about the dean’s personal situation such as age and health. The instruction the subjects received was to reallocate the two million dollar budget cut across the two schools. All subjects received information prior to negotiating that: a. Bargaining would start with one round of negotiations, with a possibility of a second round. b. Their budget consisted of four separate budget items of differing value to each of the faculty deans but that they would receive more information on the relative value of these four budget items if they received approval from the vice president to continue negotiations in a second round.4 c. If allowed to continue to negotiate, they would then be able to negotiate each budget item separately. d. Both members of a negotiating dyad would have to request to continue to negotiate.5 If only one made such a request, the vice president would invoke the intervention process (arbitrate). e. They would have fifteen minutes to bargain. If at the end of fifteen minutes the subjects reached agreement, they signed the appropriate forms and filled out a post-negotiation questionnaire. If they did not reach agreement, they could either request the vice president’s intervention (using the applicable intervention process) or request another round of negotiations with additional information.6 In the latter case, they had to submit an impasse offer the vice president can use for intervention purposes if permission to continue to negotiate was not forthcoming.7 Agreements at this stage (end of Round 1) were, by definition, distributive, since no logrolling was possible and one’s gain was another’s loss. Agreements at this stage were not profitable to reach (see Table 1); any resolution at this stage could not be synergistic or integrative and required very little search. If dyads requested a second round of negotiations, their task continued to find an agreeable reallocation of the two million dollar budget cut between the two schools. Each subject entering the second round of negotiations received information about the relative value of each of the four components of their budget to the subject. The relative value of each of the four components was different to the two schools as represented by a different corresponding voucher deduction amount for each subject in a dyad at each level of budget cut (Table 2a and 2b). This allowed for logrolling and

The Effects of Organizational Culture on Budgetary Conflict

Table 2a. How Individual Item Budget Settlements Affect Subjects’ Pay. Schedule of Budget Cuts and Personal Consequences for You. Report from the Vice President (Session 2) – Agriculture.
Budget Cut for Item A (Geotechnical Engineers) Column 1 Dollars Deducted from Your Voucher Budget Cut for Item B (Design Engineers) Dollars Deducted from Your Voucher Budget Cut for Item C (Programmers) Dollars Deducted from Your Voucher Budget Cut for Item D (Technical Staff) Dollars Deducted from Your Voucher

Column 2 (Confidential) À$187 À187 À180 À174 À166 À156 À144 À132 À120 À108 À98 À94 À86 À74 À59 À46 À42 À32 À16 À5 0

Column 3

Column 4 (Confidential) À$101 À101 À98 À94 À90 À84 À78 À70 À60 À52 À46 À45 À42 À37 À31 À25 À22 À16 À8 À2 0

Column 5

Column 6 (Confidential) À$48 À48 À48 À47 À46 À45 À43 À40 À37 À32 À24 À24 À23 À21 À18 À13 À11 À8 À4 À1 0

Column 7

Column 8 (Confidential) À$104 À104 À102 À101 À98 À95 À91 À86 À79 À68 À52 À51 À49 À46 À40 À28 À25 À18 À8 À2 0

$2,000,000 1,900,000 1,800,000 1,700,000 1,600,000 1,500,000 1,400,000 1,300,000 1,200,000 1,100,000 1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0

$2,000,000 1,900,000 1,800,000 1,700,000 1,600,000 1,500,000 1,400,000 1,300,000 1,200,000 1,100,000 1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0

$2,000,000 1,900,000 1,800,000 1,700,000 1,600,000 1,500,000 1,400,000 1,300,000 1,200,000 1,100,000 1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0

$2,000,000 1,900,000 1,800,000 1,700,000 1,600,000 1,500,000 1,400,000 1,300,000 1,200,000 1,100,000 1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0

Note: There is a minimum budget cut for the MEGA UNIT of $300,000 in each of items A, B, C, and D; the total cut for the MEGA UNIT must be a minimum of $2,000,000.

117

118

Table 2b. How Individual Item Budget Settlements Affect Subjects’ Pay. Schedule of Budget Cuts and Personal Consequences for You. Report from the Vice President (Session 2) – Engineering.
Budget Cut for Item A (Geotechnical Engineers) Column 1 Dollars Deducted from Your Voucher Budget Cut for Item B (Design Engineers) Dollars Deducted from Your Voucher Budget Cut for Item C (Programmers) Dollars Deducted from Your Voucher Budget Cut for Item D (Technical Staff) Dollars Deducted from Your Voucher

Column 2 (Confidential) À$49 À49 À48 À47 À46 À44 À42 À40 À37 À32 À24 À24 À23 À21 À19 À14 À13 À10 À5 À1 0

Column 3

Column 4 (Confidential) À$111 À111 À108 À107 À104 À100 À96 À92 À85 À76 À56 À54 À52 À48 À42 À30 À27 À19 À9 À2 0

Column 5

Column 6 (Confidential) À$190 À190 À181 À177 À168 À158 À146 À130 À115 À99 À96 À93 À87 À77 À60 À45 À41 À30 À15 À5 0

Column 7

Column 8 (Confidential) À$90 À90 À87 À85 À82 À78 À72 À66 À59 À53 À41 À41 À38 À32 À27 À21 À19 À15 À7 À2 0

$2,000,000 1,900,000 1,800,000 1,700,000 1,600,000 1,500,000 1,400,000 1,300,000 1,200,000 1,100,000 1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0

$2,000,000 1,900,000 1,800,000 1,700,000 1,600,000 1,500,000 1,400,000 1,300,000 1,200,000 1,100,000 1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0

$2,000,000 1,900,000 1,800,000 1,700,000 1,600,000 1,500,000 1,400,000 1,300,000 1,200,000 1,100,000 1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0

$2,000,000 1,900,000 1,800,000 1,700,000 1,600,000 1,500,000 1,400,000 1,300,000 1,200,000 1,100,000 1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0

NABIL ELIAS AND WILLIAM W. NOTZ

Note: There is a minimum budget cut for the MEGA UNIT of $300,000 in each of items A, B, C, and D; the total cut for the MEGA UNIT must be a minimum of $2,000,000.

The Effects of Organizational Culture on Budgetary Conflict

119

discovery of integrative solutions following Thompson and Hastie (1990) and Pruitt and Lewis (1975). Achieving an integrative solution required subjects to use negotiations to find which is more beneficial. Table 2 is generally consistent with Table 1; an evenly distributed budget cut across the four budget items in Table 2 would produce the same amount of voucher deduction as the total budget cut would in Table 1.8 It is important to note that the two rounds of negotiations were used to simulate one aspect of search. Further search for information resulted in the negotiation process in the second round. Real world negotiations are typically not concluded in one round unless negotiations are either easy (no conflict), or bargainers give up without searching for information or solutions. The two rounds of negotiations were available to all dyads who were informed that the budget consisted of four separate items of differing value to each subject, that they would be able to negotiate each budget item separately but that they would receive more information on the relative value of these four items if they chose and were approved to have a second round. 4.3. Experimental Design and Procedures The basic design is a 1 Â 2 factorial design with one level of conflict and two levels of OCs, traditional and empowering. The remainder of this subsection describes the experimental procedures and manipulations. As subjects arrived at the experimental setting, they were told that they would be participating in a simulated budget negotiation in a university. After screening pairs of subjects to ensure they are not known to each other, dyads were randomly assigned to either the TOC or the EOC treatment, and each subject within a dyad was randomly assigned the role of either the dean of engineering or the dean of agriculture. Each was directed to a separate room, given general information about the university and their particular school. Subjects were then presented with information that differed across the two treatments. In the TOC treatment, subjects received the following statement from the university vice president:
The University Grants Commission has required that we allow Deans to negotiate directly, and that is why you are negotiating. If you ask me, I think this is really a waste of time. I can reach a superior decision on my own as I have done in the past. I think the hierarchy of decision making in the University has to be maintained and that I have the best idea of how the budget cuts should be made. I hope that your negotiations will lead to a reallocation decision that is as good as the one I reached on my own. If it’s not, I will ask you to either continue to negotiate or to provide me with your impasse offer.

120

NABIL ELIAS AND WILLIAM W. NOTZ

Subjects in the TOC treatment also received information explaining that if an impasse in negotiations occurred, the vice president’s intervention process for conflict resolution would be to examine the impasse offers of both subjects (deans) and decide on a settlement that was fair to both – a decision that would generally fall somewhere between their respective demands and impose a compromise. Subjects also received a hypothetical illustration of the vice president’s impasse intervention process and answered questions to indicate they fully understood its application. In the EOC treatment, subjects received the following statement from the vice president:
I seriously believe that the two of you can negotiate a much better agreement from the University’s point-of-view than any solution that I can reach on my own without your involvement or commitment. I have made these decisions in the past without your involvement, but these solutions did not always prove to work well. On the other hand, I have already made a tentative decision on your reallocation and I will not accept any solution that is less advantageous from the University’s point-of-view than my own decision. If you reach an agreement that is less beneficial from the University’s pointof-view, you will be asked either to continue to negotiate or to provide me with your impasse offer.

Subjects in the EOC treatment received information that the university hoped that its new decentralized structure would produce more creative, better informed, more equitable, and substantially more acceptable budget reallocations. The university further realized that this would demand a very significant effort from the deans to understand each other’s position, each other’s dilemmas, opportunities, constraints, fears, etc. In order to encourage and reward these behaviors, the vice president explained the supporting intervention process for dealing with impasses between deans. The vice president’s intervention process in the EOC treatment consisted of two components. Subjects in this treatment read the following description of the first component:
The Vice President y will first examine the impasse offers of both Deans for all issues. If the impasse offers are judged to be reasonable, even though short of an agreement, y then the Vice President will select either one or the other of the two impasse offers in its entirety, the one which is the most fair to both deans. No compromise is possible. The one offer the Vice-President selects will then determine the outcomes for both you and the other dean.

Subjects also read a hypothetical example that further illustrated the operation of this part of the intervention process. They also received information about the second component of the EOC treatment intervention process; the vice president will examine the deans’ impasse offers and if the

The Effects of Organizational Culture on Budgetary Conflict

121

examination revealed excessive self-interest in the offers of either one or both deans, then role reversal (in budget cut allocation) would be imposed to resolve the impasse and correct the self-interest problem. This would apply by switching (or reversing) the voucher deduction amount. Subjects received an illustration of the budget cut assignment of role reversal with a modification of the previous example. They also answered questions to indicate they fully understood both aspects of this intervention process (final offer selection and role reversal). Subjects in both treatments (EOC and TOC) were given the opportunity to ask questions and in one or two cases were referred back to the material they already had for clarification. They then completed a pre-negotiation instrument (described later). Manipulation checks were performed through questions that examined each subject’s understanding of the vice president’s instructions, the intervention process, and the negotiating situation. If any misunderstanding of these items occurred (and this happened only rarely), subjects were referred back to the relevant material and were asked to answer the questions again. All questions were answered correctly. Subjects came together in another room and bargained for fifteen minutes. They faced each other across a table with a low partition between them and, with their knowledge, were tape-recorded.9 At the conclusion of the bargaining session, subjects returned to their separate rooms. Two outcomes were possible for each dyad: either they reached an agreement or they did not. Subjects in a dyad who reached agreement then completed a post-negotiation questionnaire, and received information about their pay. They were also prompted to provide an offer the vice president could use to settle the dispute, should the agreement not receive his/her acceptance. Dyads who did not reach agreement could either request immediate intervention by the vice president, or permission to continue bargaining in a second round of negotiations. All subjects were required to submit an impasse offer to the vice president, whether or not they had requested permission to continue bargaining.10 Subjects who requested immediate intervention were informed of the vice president’s decision in accordance with the applicable intervention process. These subjects then completed a post-negotiation questionnaire and received information about their pay. Subjects who requested continued bargaining were also required to submit their impasse offer for use by the vice president if continued bargaining was denied.11 Thus at the end of the first stage of negotiations in the experiment, all subjects had either reached an agreement or made an impasse offer, thereby providing a useful measure of their bargaining behavior. See Fig. 2 for the experimental sequence.

122

Random Assignment of Subjects

Role Faculty of: 1. Engineering 2. Agriculture

Experimental Treatment (Conflict Environment) 1. TOC 2. EOC

No Agreement Or

Round 1 Negotiation

Agreement* (Distributive)

Request for Continued Bargaining and Accompanying Impasse Offer

Request for Intervention and Submission of Impasse Offer

Vice President’s Intervention (Decision) Using Applicable Intervention Procedure

NABIL ELIAS AND WILLIAM W. NOTZ

No Agreement

Round 2 Negotiation

Agreement* (Integrative) Request for Intervention and Submission of Impasse Offer

*In all cases of agreement, subjects also filled out an offer to be used by the arbitrator in case the agreement was not approved.

Vice President’s Intervention (Decision) Using Applicable Intervention Procedure

Fig. 2.

Experimental Sequence.

The Effects of Organizational Culture on Budgetary Conflict

123

All dyads requesting continued negotiations received authorization to have another round of negotiation. Each subject received a ‘‘Report from the Vice President’’ which contained the applicable decomposed budget cut levels and corresponding voucher deduction (Table 2a or 2b). Subjects had ten minutes to prepare for the second round of negotiation. Then, they completed another questionnaire (similar to the pre-negotiation instrument), and were brought together for the second round of face-to-face negotiations. Dyads had twenty minutes to negotiate at the end of which they returned to their separate rooms.12 If they had reached agreement, they received instructions similar to those who had agreed in the previous session. Subjects in dyads who had not reached agreement completed their impasse offers for resolution by the vice president based on the applicable intervention process. The remaining procedures for these subjects were identical to the previous description. 4.4. Dependent Variable Measures In this study, the dependent variable revolves around budget conflict bargaining behavior in terms of its distributive or integrative nature. Consistent with the hypotheses, the dependent variable measures include budget negotiation outcomes, expectations, and perceptions. 4.5. Budget Conflict Negotiation Outcome The most important measures of our dependent variable relate to the outcome of budget bargaining behavior. We use one discrete variable and one continuous variable to measure bargaining behavior outcomes. (1) The first variable we use provides a relatively crude but useful measure of integrative behavior. As mentioned earlier, dyads that decided to settle their budget conflict by intervention during the experiment or by agreements at the end of Round 1 reflect predominantly distributive bargaining behavior. Only strictly distributive agreements (or arbitrated settlements) were possible at the end of Round 1 since logrolling was not possible, and dyadic joint earnings would at best sum to zero (a full deduction of the voucher amount). Agreements in Round 2 were potentially integrative, since logrolling was possible only in that round. Thus, the number of negotiations settled by intervention or by agreement at the end of Round 1 as well as those settled by intervention at the

124

NABIL ELIAS AND WILLIAM W. NOTZ

Table 3. Maximum Integrative Solution.
Budget Cut Amount (in $1,000) Voucher Deduction, Value, and Earnings (in $) Agriculture $0 À$8 À$21 À$2 À$31 $55 $24 Engineering À$21 À$2 $0 À$7 À$30 $55 $25 Total À$21 À$10 À$21 À$9 À$61 $110 $49

Agriculture Budget Budget Budget Budget Item Item Item Item A B C D $0 $200 $700 $100 $1,000

Engineering $700 $100 $0 $200 $1,000

Total $700 $300 $700 $300 $2,000

Total budget cut Voucher value Maximum Possible individual and dyadic earnings

end of Round 2 can serve as a proxy for distributive agreements while the number of agreements in Round 2 can be considered a reasonable proxy for integrative agreements. (2) A more refined measure of integrative behavior was the total dyadic earnings. This variable could range from strictly distributive to fully integrative. The maximum integrative agreement would result in a combined dyadic pay of $4913 as shown in Table 3.

4.6. Budget Conflict Negotiation Expectations Several measures of expectations were taken using the responses to the questionnaires administered before and after negotiations. These include: (a) Subjects’ expectations about what would constitute a good, poor, and reasonable settlement, and intended opening offer before each round of negotiations. (b) Another measure of the degree of integrative outcomes is the difference between the two offers of a negotiating dyad, which is an indicator of bargaining discrepancy (Elias, 1990). Small (large) bargaining discrepancies reflect more (less) integrative (distributive) behavior. Put another way, large bargaining discrepancies measure the level of intransigency in bargaining behavior that is associated more with distributive behavior.

The Effects of Organizational Culture on Budgetary Conflict

125

4.7. Budget Conflict Perceptual and Cognitive Variables These variables include the following: 1. Subjects’ satisfaction with the information sharing and the level of mutual trust. 2. The extent to which the situation was framed by subjects as zero-sum or joint problem solving. 3. How subjects saw themselves and their negotiating counterpart as having behaved in a flexible and innovative manner. 4. Subjects’ estimate of the likelihood that settlement by intervention of the vice president could be avoided. 5. Subjects’ satisfaction with the negotiation outcome. The next section provides the results of the effects of OC on budget conflict negotiating outcomes and behavior.

5. RESULTS
5.1. Budgeting Conflict Negotiation Outcome As described in the previous subsection, the first relatively crude measure of budget conflict bargaining outcome is the number of dyads in a cross classification of distributive or integrative outcomes that is the outcome of the EOC and TOC. As described earlier, the outcomes of dyads who reached agreement at the end of the first round of negotiations or who requested settlement through intervention (by the vice president) were classified as distributive, and those of dyads who reached agreement at the end of the second round of negotiations were classified as integrative. Dyads who settled or asked for intervention at the end of Round 1 knew there was a potential Round 2 with more information on individual budget items. Table 4 reports these data for the TOC and EOC treatments. Whereas bargaining under the TOC produced 16 distributive (11 in Round 1 and 5 in Round 2) and 5 integrative outcomes (Round 2), the EOC treatment produced a nearly symmetric reversal of 7 distributive outcomes (all in Round 1) and 14 integrative (all in Round 2). The result was statistically significant, supporting hypothesis H1A. An analysis of the subjects’ earnings in the experiment provides a more sensitive analysis of distributive vs. integrative outcomes. Table 5 provides an analysis of individual and dyadic earnings produced by the TOC and

126

NABIL ELIAS AND WILLIAM W. NOTZ

Table 4. Distributive and Integrative Outcomes by Budgetary Environment.
Traditional Organizational Culture Distributive outcomes Dyads settled through intervention in Rounds 1 and 2 or agreements in Round 1 Integrative outcomes Dyads settled through agreements in Round 2 Total number of dyads a Empowering Organizational Culture 7b

w2 Value

P-Value (1-Tail)

16a

5 21

14 21 7.78490 0.00264

This number consists of 1 agreement and 10 interventions in Round 1; and 5 interventions in Round 2. b This number consists of no agreements and 7 interventions in Round 1; no interventions in Round 2.

Table 5.

Dyadic and Individual Earnings and Pay Variables by Organizational Culture.
Traditional Organizational Culture Mean (n ¼ 21 Dyads; 42 Subjects) Empowering Organizational Culture Mean (n ¼ 21 Dyads; 42 Subjects) $29.38 F-Ratio F-Prob. (1-Tail)

Dyadic earnings (negative earnings included) Individual earnings (negative earnings included) Dyadic pay (negative earnings converted to $0) Individual pay (negative earnings converted to $0)

$15.24

4.6050

0.0190

$7.61

$14.69

3.7409

0.0283

$24.00

$34.86

4.7776

0.0174

$12.00

$17.43

5.2881

0.0120

The Effects of Organizational Culture on Budgetary Conflict

127

EOC treatments. While dyads earned an average of $15.24 (31% of the $49 maximum possible dyadic earnings) in the TOC treatment, the EOC treatment produced more integrative outcomes of nearly twice as much, $29.38 (60% of the maximum possible dyadic earnings). Since negative earnings were possible but negative pay to subjects was not, we examined dyadic pay after converting negative earnings to reflect zero pay. On that basis, average dyadic pay was $24.00 in the TOC treatment as compared to $34.86 in the EOC treatment. In both cases, the results were statistically significant and the same was true for individual earnings and individual pay. Thus, subjects who bargained under the EOC treatment produced significantly more integrative agreements and enjoyed substantially greater pay than their counterparts who bargained under the TOC treatment, thus supporting hypothesis H1B.

5.2. Budgeting Conflict Negotiation Expectations A second set of variables describes subjects’ initial states, expectations, and plans. Subjects answered questions about what they considered a reasonable, good, or poor settlement. In addition, they also provided their intended opening offer prior to the beginning of negotiations. Subjects provided this data before each of the two rounds of negotiations began. Panel A of Table 6 shows that before the first round of negotiation began, subjects’ pre-negotiation expectations reflect a pattern consistent with H2A. Subjects in the EOC treatment generally expected greater budget cuts and lower personal earnings than did their counterparts in the TOC treatment (see the difference column). Differences in subjects’ intended opening offers were statistically significant. Differences in subjects’ perceptions of a reasonable, good, or poor settlement were in the expected direction though not statistically significant. Thus, prior to the beginning of negotiations, there is only partial support for H2A. On the other hand, as Panel B of Table 6 shows, the patterns described above did not only persist but also became more pronounced and were all statistically significant at the end of the first round of negotiations (before the second round of negotiations began), lending support to H2A. This was so even in view of the smaller sample size that participated in the second round of negotiations. Table 7 shows the mean of the impasse offers subjects submitted as their request for intervention if they reached an impasse or as prompted in the experiment if they requested to continue to negotiate with a decomposed

128

Table 6.

Individual Expectation Variables Regarding Budget Cuts.
Traditional Organizational Culture (n ¼ 42) Empowering Organizational Culture (n ¼ 42) Difference F-Ratio F-Prob. (1-Tail)

Panel A: Pre-negotiation – Round 1 Reasonable (fair) settlement Good settlement Poor settlement Intended opening offer Reasonable (fair) settlement Good settlement Poor settlement Intended opening offer

Mean budget cut (in $000) 814 888 560 648 1,082 1,105 374 565 Mean individual earnings (in $) 15.57 10.17 33.40 28.52 À4.93 À8.21 42.13 33.70

74 88 23 191 5.40 4.88 3.28 8.43

3.5064 3.0195 0.1067 11.0814 2.7490 2.7400 0.5195 5.9252

0.0647 0.0860 0.7448 0.0013 0.0506 0.0509 0.2366 0.0086

NABIL ELIAS AND WILLIAM W. NOTZ

Traditional Organizational Culture (n ¼ 20) Panel B: Pre-negotiation – Round 2 Reasonable (fair) settlement Good settlement Poor settlement Next offer Reasonable (fair) settlement Good settlement Poor settlement Intended next offer

Empowering Organizational Culture (n ¼ 28)

Difference

F-Ratio

F-Prob. (1-Tail)

725 598 910 526 (n ¼ 19)

904 811 1,079 796

179 213 169 270 12.16 15.30 13.17 17.01

8.1022 16.7199 6.5588 20.1266 6.8740 15.3961 7.0407 15.1946

0.0066 0.0002 0.0138 0.0000 0.0059 0.0002 0.0055 0.0002

Mean individual earnings (in $) 21.00 8.84 31.62 17.32 6.10 À7.07 18.36 35.37 (n ¼ 19)

The Effects of Organizational Culture on Budgetary Conflict

129

Table 7.

Individual Impasse Offers Regarding Budget Cuts to Subject End of Round 1.
Mean Budget Cuts (in 000’s) Traditional Organizational Culture (N ¼ 42)a Empowering Organizational Culture (n ¼ 42) 770b Difference F-Ratio F-Prob. (1-Tail)

Impasse offers submitted for intervention or offers to be considered in case of intervention Impasse offer submitted for intervention or offer to be considered in case of intervention a b

343

427

66.2542

0.0000

Mean individual earnings (in $) 44.30 20.31c 23.99

75.8400

0.0000

This includes one dyad who reached agreement in Round 1. The EOC mean budget cut offer is nearly 2.25 times the TOC mean offer. c The EOC individual earnings mean associated with the budget cut is nearly 46% of the TOC mean.

budget.14 Consistent with H2B, the budget cut offer was substantially higher in the EOC treatment than in the TOC treatment (nearly 2.25 times) and the corresponding payment associated with the budget cut offer was substantially lower (46%), and these differences were statistically significant. 5.2.1. Budget Conflict Expectations at the Dyadic Level Another test of H2B is to examine budget expectations at the dyadic level. An important aspect of the bargaining was the subjects’ cooperation in absorbing the budget cut. We measured this variable by comparing the required two million dollar budget cut with the sum of the cuts proposed by the two members of each negotiating dyad. Since we expect the EOC treatment to produce integrative outcomes, we would also expect it to produce smaller differences between the two million dollar cut and the dyadic proposals. As Table 8 shows, the pre-negotiation closeness to the target budget cut was considerably smaller in the EOC than in the TOC treatment. This pattern was generally

130

Table 8.

Dyadic Closeness to Target Budget Cut of $2 Million by Organizational Culture.
Traditional Organizational Culture (in $000)Ã Empowering Organizational Culture (in $000)Ã (n ¼ 21 dyads) 224 705 À210 869 (n ¼ 14 dyads) 193 379 À157 407 80 (n ¼ 14) 80 (n ¼ 14) Difference (in $000)Ã EOCoTOC F-Ratio F-Prob. (1-Tail)

Panel A: Pre-negotiation – Round 1 Reasonable (fair) settlement Good settlement Poor settlement Intended opening offer Panel B: Pre-negotiation – Round 2 Reasonable (fair) settlement Good settlement Poor settlement Intended opening offer Panel C: End of Round 2 For agreements reached only; offers submitted for possible arbitration For combined agreements reached and impasse offers; offers submitted for possible arbitration and impasse offers submitted for arbitration
à Values rounded to the nearest $1,000.

(n ¼ 21 dyads) 371 881 À164 1,252 (n ¼ 10 dyads) 550 805 180 956 (n ¼ 9) 360 (n ¼ 5) 485 (n ¼ 10)

147 176 46 383 357 426 337 549 280 405

4.7154 3.3076 0.1372 11.8920 6.3463 14.5323 4.4605 29.5346 4.5419 5.4149

0.0180 0.0383 0.3565 0.0007

NABIL ELIAS AND WILLIAM W. NOTZ

0.0098 0.0005 0.0232 0.0000 0.0264 0.0159

The Effects of Organizational Culture on Budgetary Conflict

131

consistent from pre-negotiation dyadic expectations and intended opening offers (Panel A) to the end of the first round of negotiation (Panel B). In addition to having a zero divergence for all 14 EOC and only 5 TOC dyads who reached agreement at the end of the second round of negotiations, we measured dyadic divergence as the difference between the total required budget cut and the offers to be used in case agreements were not approved. The EOC treatment produced greater convergence as measured by the average closeness to the target budget cut of $80,000 while the TOC treatment produced lesser convergence as evidenced by the average closeness of $360,000. When we consider dyadic divergence of all dyads participating in the second round of negotiations (including dyads who did not reach agreement) this difference is even greater; the EOC treatment produced an average closeness to the target budget cut of $80,000 as compared to the TOC average of $485,000, which is six times larger. These differences were statistically significant (Panel C), lending support to hypothesis H2B.

5.3. Budget Conflict Perceptual and Cognitive Variables Table 9 provides a comparison by conflict environment treatment of the perceptions of subjects regarding their trust of the other subject, being trusted by the other subject, holding the view that one’s gain is the other’s loss, the need for cooperation, their plans to share information with the other subject, and their expectation that the other subject would share information. This analysis covered pre-negotiation responses conducted prior to each of the two negotiation sessions. In addition, the post-negotiation questionnaire also included relevant perceptual and cognitive variables. Panel A of Table 9 shows that before the beginning of negotiations all perceptual and cognitive variables were in the predicted direction; that is, there was a perception of greater mutual trust, greater desire for cooperation, and more plans for information sharing in the EOC than in the TOC treatment. The differences in perception in all of these variables were statistically significant. While the view that one’s gain is the other’s loss was also in the predicted direction, the difference was not statistically significant. Likewise, Panel B of Table 9 shows that all Round 2 pre-negotiation variables were in the predicted direction, but not all were statistically significant. This, only in part, may be due to the substantial shrinkage in sample size.15 At the end of negotiations, subjects in the EOC treatment reported that their behavior and their opponent’s behavior were less ‘‘rigid and uncompromising.’’ These subjects also perceived that there had been more mutual

132

Table 9.

Individual Perceptual and Cognitive Variables by Organizational Culture.
Traditional Organizational Culture Mean (n ¼ 42) Empowering Organizational Culture Mean (n ¼ 42) F-Ratio F-Prob. (1-Tail)

Panel A: Pre-Negotiation – Round 1

Extent to which subject is prepared to trust other subject Extent that other subject is prepared to trust subject View of one’s gain being the other’s loss View of situation involving joint problems requiring cooperation Extent to which subject plans to share information about budget situation Expectation that other subject will share information about budget situation

(1 not at all; 9 completely) 4.67 5.77 4.60 5.24 6.81 6.36 6.43 7.07 (1 everything; 9 nothing) 4.52 3.71 4.93 3.83

8.8671 2.7827 1.1512 2.8803 4.9702 10.7239

0.0019 0.0496 0.1432 0.0468 0.0143

NABIL ELIAS AND WILLIAM W. NOTZ

0.0008

Panel B: Pre-Negotiation – Round 2

Mean (n ¼ 20)

Mean (n ¼ 28)

Extent to which subject is prepared to trust other subject Extent that other subject is prepared to trust subject View of one’s gain being the other’s loss View of situation involving joint problems requiring cooperation Extent to which subject plans to share information about budget situation Expectation that other subject will share information about budget situation

(1 not at all; 9 completely) 4.85 5.46 4.60 5.57 7.25 6.29 6.35 7.00 (1 everything; 9 nothing) 4.30 3.57 4.65 4.21

1.1408 2.9503 2.5271 1.2286 1.5910 0.6165

0.1456 0.0463 0.0594 0.1367 0.1068 0.2182

The Effects of Organizational Culture on Budgetary Conflict

Panel C: Post-Negotiation

Mean (n ¼ 42)

Mean (n ¼ 42)

F-Ratio

F-Prob. (1-Tail)

Extent of subject’s own behavior as rigid and uncompromising Extent of other subject’s behavior as rigid and uncompromising View of one’s gain being the other’s loss View of situation involving joint problems requiring cooperation Extent to which subject provided distorted information Extent of subject’s own behavior as flexible and innovative Extent of other subject’s behavior as flexible and innovative Satisfaction with exchange of information during negotiation Extent to which subject shared information about budget situation Extent to which other subject shared information about budget situation

(1 not at all; 9 completely) 4.50Ã 3.42Ã 4.37Ã 3.53Ã 6.95 5.67 6.24 6.60 2.97Ã 2.89Ã 5.18Ã 6.16Ã Ã 5.11 5.61Ã 5.92Ã 5.24Ã (1 everything; 9 nothing) 4.88 3.88 5.21 4.45 (1 extremely dissatisfied; 9 extremely satisfied) 6.50ÃÃ 4.87Ã

5.0090 2.8462 6.5705 0.5233 0.0249 4.0793 1.0140 1.5990 5.9214 3.0713

0.0141 0.0479 0.0061 0.2358 0.4376 0.0235 0.1586 0.1050 0.0086 0.0417

Satisfaction with the outcome
à n ¼ 38 (four missing values). ÃÃn ¼ 36 (six missing values).

12.0789

0.0005

133

134

NABIL ELIAS AND WILLIAM W. NOTZ

sharing of information, and they had a weaker belief in the zero-sum nature of the bargaining than subjects in the TOC treatment. As Panel C shows, these differences were statistically significant. However, other variables, while in the predicted direction, were not statistically significant, for example, need for cooperation, and other subject’s flexible and innovative behavior. In general, subjects in the EOC treatment had different perceptions of themselves, their negotiating counterparts, and the bargaining process. We conclude that there was some support for H3. 5.3.1. Satisfaction with Budget Negotiation Outcome Table 9, Panel C shows greater satisfaction with the negotiation outcome by subjects in the EOC than in the TOC treatment. Although satisfaction with the outcome can, at least in part, be attributed to the expectation of higher pay in the EOC treatment, it is nonetheless important to report the dramatic difference in satisfaction with the outcome, in favor of the EOC. While satisfaction with the exchange of information during negotiation was also higher in the EOC treatment, this result was not statistically significant.

6. DISCUSSION AND CONCLUSIONS
Budget conflict received little explicit attention in the literature, and particularly with respect to OC and managerial style. Budgetary conflict can be costly to organizations. We proposed changes in two elements in the TOC as a means to enhance the propensity to produce more integrative budgetary conflict resolution: an empowering leadership style and an intervention process (in failed negotiations) that encourages the parties to resolve their own conflict. The results of this study supported our primary hypothesis that the TOC treatment produced budget conflict outcomes that were relatively distributive. On the other hand, the EOC treatment produced significantly more integrative budget conflict outcomes. The most dramatic evidence of this is dyadic earnings (which represent subunit utility); dyads in the EOC treatment earned nearly twice as much as dyads in the TOC treatment. The study also found that differences in budget bargaining outcomes were consistent with differences in subject and dyadic expectations. Bargainers in the EOC treatment held more reasonable expectations than those held by their counterparts in the TOC treatment before negotiations began, and those differences became greater as negotiations proceeded. Subjects in the EOC treatment expected greater budget cuts and lower personal earnings than did their counterparts in the TOC treatment; and this was persistent

The Effects of Organizational Culture on Budgetary Conflict

135

throughout the negotiation process. Dyadic expectations in the EOC treatment showed greater convergence towards budget goals than their counterparts in the TOC treatment. Finally, subjects in the EOC treatment saw themselves and their opponent as less ‘‘rigid and uncompromising,’’ more sharing of information, and less likely to define bargaining as a zero-sum conflict than their counterparts in the TOC treatment. Although the experimental results support our primary hypothesis that the EOC treatment would produce more integrative budgetary negotiation behavior and outcomes than the TOC treatment, this research cannot explain the complex causal chain that must have taken place during negotiations. For example, Galinsky and Mussweiler (2001) provide evidence that whoever makes the first offer (a buyer or seller) provides an anchor for subsequent negotiations. Although we collected data about intended opening offers at each stage of budget negotiations, we did not collect data as to who made the first offer and the role this may have had on outcome. There is a threat to the external validity of our findings. Subjects bargaining in a laboratory simulation are clearly quite different from managers bargaining in a real organization. Similarly, relationships between negotiators in organizations are relatively long term as compared to those in the laboratory, constituencies are present in real organizations but not in the laboratory, and the stakes of bargainers are much greater and more complex. However, studying real organizations and implementing the proposed changes in what we called the TOC can be costly without experimentation. The rich context of organizational settings hampers theory development (internal validity). Ultimately, the study findings must find their true test in real organizations. Our independent variable, OC, is subject to wide interpretation. Our focus was on two factors that if simultaneously changed would produce what we called an ‘‘empowering’’ OC. The construct we used consisted of the coupling of two factors that we did not independently manipulate as two separate independent variables: managerial style and managerial intervention. Such coupling was made for pragmatic and practical reasons: (a) consistency: it is more likely to see an empowering managerial style with an intervention process that expects the parties to resolve their conflict on their own, or a traditional managerial style that does not expect parties to resolve conflict on their own and applies a compromise instead, than the other two inconsistent and less likely combinations (see Note 1); (b) feasibility: if all possible combinations of these two factors were used we would have required four different experimental treatments (instead of two), doubling the sample size, and doubling the cost of the experiment. Nonetheless, future

136

NABIL ELIAS AND WILLIAM W. NOTZ

research can determine which of these two factors, if any single factor, is powerful in producing integrative budgetary conflict resolution. Our study did not consider the way in which negotiators framed the issues. There is a good deal of evidence that the framing of issues in a negotiation affects bargaining strategies and outcomes (Bazerman, Magliozzi, & Neale, 1985; Bottom & Studt, 1993; Neale, Huber, & Northcraft, 1987). Moreover, since framing affects changes in risk attitude, it follows that bargaining over losses (burdens) or gains (benefits) will affect bargaining behavior and outcomes (Sondak, Neale, & Pinkley, 1995). Our experiment framed the subject of negotiation as a budget cut, and the corresponding parallel effect on subjects as a deduction from their voucher which determined their pay (a loss). Quite possibly, if the experiment framed the subject of negotiation as a budget increase instead of a budget cut and the corresponding parallel effect on subjects as an increase in their pay (a gain), the results could be different. In this connection, it is worth noting that Sondak et al. (1995) found the negotiators allocating burdens (similar to budget cuts in our study) were more competitive and less effective in finding integrative solutions than were their counterparts who were allocating benefits. If so, one implication from their findings is that the effects of our conflict environment manipulation might be rather robust. These and other variations of framing suggest interesting interaction effects, which will require further research.

NOTES
1. It is possible to treat each of the components of our OC independent variable as a separate variable. If so, we would have four cells of possible combinations as follows: Intervention Process Managerial Style Traditional Compromise Final offer a Empowering Cell 2a Cell 4 (EOC)

Cell 1 (TOC) Cell 3a

Combination is less likely to exist.

We conclude that cells 2 and 3 are unlikely and focus on cells 1 and 4. Such focus would not enable us to determine the separate effects of each of the two components of the OC variable on conflict resolution. There is also a practical dimension to this choice. The number of negotiating dyads would have to be doubled and so would the

The Effects of Organizational Culture on Budgetary Conflict

137

cost of conducting the experiment. While the separation of the two components could provide us with useful insights, the combination of cost and the apparent inconsistency within cells 2 and 3 resulted in our decision to focus on cells 1 and 4. Future research can focus on other combinations. 2. The word ‘‘treatment’’ could be substituted with ‘‘situation,’’ ‘‘environment,’’ or ‘‘condition.’’ Since we manipulated this variable in the experiment, we use ‘‘treatment’’ in that sense. 3. Agreement does not necessarily result in an integrative resolution if both parties treated the problem as a fixed-pie conflict and did not probe or search for information to reach an integrative resolution. 4. The two-rounds of negotiations were used to simulate one aspect of search for information. 5. This is typical in real life. Both parties have to agree to negotiate before negotiation takes place. 6. Here again, subjects who search for information would opt for the second round if they were searching for integrative resolution. 7. All requests to continue negotiation received approval. The purpose of requesting an impasse offer was a means of obtaining a reading of where the subjects were at the end of the first round of negotiation. 8. At realistic budget cut levels, voucher deductions corresponding to equal budget cuts across the four budget items in Table 2 were equal to the voucher deduction resulting from a corresponding total budget cut in Table 1. For example, a budget cut of $200,000 in each of the four budget items would produce a voucher deduction of $36 (Table 2), which is equivalent to the voucher deduction corresponding to a total budget cut of $800,000 (Table 1). 9. Tape-recording was an added measure to discourage any attempt at collusion. 10. This provided the researchers with data at the end of the first session to gauge how close or far apart subjects were. 11. This was the justification to obtain a measure of the discrepancy between the two bargainers at the end of Round 1. 12. Subjects were not allowed to share their voucher deduction information in Table 2 with their negotiating counterpart. 13. The $49 amount is the maximum dyadic pay and is the amount remaining from the combined dyadic voucher value of $110 after subtracting the minimum possible combined voucher deduction of $61. See Table 3. 14. Subjects received instructions that they must submit an impasse offer that would be used by the university vice president if they did not receive approval to continue to negotiate. See Footnote 1. 15. From n ¼ 42 in each treatment who completed the round one pre-negotiation questionnaire to n ¼ 20 and n ¼ 28 who completed the pre-negotiation questionnaire for round two in the TOC and EOC treatments, respectively.

ACKNOWLEDGMENT
The authors acknowledge with gratitude stimulating discussions on this topic with Professor John Atwell. We benefited from information provided

138

NABIL ELIAS AND WILLIAM W. NOTZ

by Professor Terence Hogan, former vice president, University of Manitoba. We would also like to acknowledge the research assistance of Natalie Braun and Lucy Guest, comments of anonymous CAAA and AAA conference reviewers, Professor Ralph Greenberg’s helpful comments as AAA conference discussant of our paper, comments made by audiences at both conferences, and comments made by colleagues at research workshops at the University of Manitoba and UNC Charlotte.

REFERENCES
Anctil, R., & Dutta, S. (1999). Transfer pricing, decision rights and divisional versus firm-wide performance evaluation. The Accounting Review, 74(1), 87–104. Argyris, C. (1973). The CEO’s behavior: Key to organizational development. Harvard Business Review, 51(2), 55–64. Axelrod, R. E. (1984). The evolution of cooperation. New York: Basic Books. Baron, R. A. (1990). Organizational conflict: Introduction. In: M. A. Rahim (Ed.), Theory and research in conflict management. New York: Praeger. Bazerman, M. H., Magliozzi, T., & Neale, M. (1985). The acquisition of an integrative response in a competitive market. Organizational Behavior and Human Decision Processes, 35, 294–313. Bazerman, M. H., & Neale, M. (1983). Heuristics in negotiation: Limitations to effective dispute resolution. In: M. H. Bazerman & R. J. Lewicki (Eds), Negotiation in organizations (pp. 51–67). Beverly Hills, CA: Sage. Bazerman, M. H., & Neale, M. A. (1992). Negotiating rationally. New York: Free Press. Bottom, W. P., & Studt, A. (1993). Framing effects and the distributive aspect of integrative bargaining. Organizational Behavior and Human Decision Processes, 56, 459–474. Butler, J. K. (1994). Conflict styles and outcomes in a negotiation with fully integrative potential. The International Journal of Conflict Management, 5, 309–325. Carnevale, P. J., & Lawler, E. J. (1986). Time pressure and the development of integrative agreements in bilateral negotiation. Journal of Conflict Resolution, 30, 636–659. Carroll, J. S., Bazerman, M., & Maury, R. (1988). Negotiator cognitions: A descriptive approach to negotiators’ understanding of their opponents. Organizational Behavior and Human Decision Processes, 41, 352–370. Chalos, P., & Haka, S. (1990). Transfer pricing under bilateral bargaining. The Accounting Review, 65(3), 624–641. Chow, C. W., Cooper, J. C., & Waller, W. S. (1988). Participative budgeting: Effects of a truthinducing pay scheme and information asymmetry on slack and performance. The Accounting Review, 63(1), 111–122. DeJong, D. V., Forsythe, R., Kim, J. O., & Uecker, W. C. (1989). A laboratory investigation of alternative transfer pricing mechanisms. Accounting, Organizations and Society, 14, 41–64. Deutsch, M. (1990). A framework for teaching conflict resolution in the schools. In: B. H. Sheppard, M. H. Bazerman and R. J. Lewicki (Eds), Research on negotiation in organizations (Vol. 2). Greenwich: JAI.

The Effects of Organizational Culture on Budgetary Conflict

139

Elias, N. (1990). The effects of financial information symmetry on conflict resolution: An experiment in the context of labor negotiations. The Accounting Review, 65(3), 606–623. Elias, N., & Ezzamel, M. (1990). Financial information asymmetry and union demands: The effects of final offer arbitration. Paper presented at the European Accounting Association Annual Congress, Budapest, Hungary (April). Etherington, L. D., & Tjosvold, D. (1992). Managing budget conflicts: A field study. Hamilton, Ontario: Society of Management Accountants of Canada. Fischer, J., Fredrickson, J. R., & Peffer, S. A. (2002a). The effect of information asymmetry on negotiated budgets: An empirical investigation. Accounting, Organizations and Society, 27(1–2), 27–43. Fischer, J. G., Maines, L. A., Peffer, S. A., & Sprinkle, G. B. (2002b). Using budgets for performance evaluation: Effects of resource allocation and horizontal information asymmetry on budget proposals, budget slack, and performance. The Accounting Review, 77(4), 847–865. Follett, M. P. (1940). Constructive conflict. In: H. C. Metcalf and L. Urwick (Eds), Dynamic administration: The collected papers of Mary Parker Follett (pp. 39–40). New York: Harper. Froman, L. A., Jr., & Cohen, M. D. (1970). Compromise and logrolling: Comparing the efficiency of two bargaining processes. Behavioral Science, 15, 180–183. Galinsky, A. D., & Mussweiler, T. (2001). First offers as anchors: The role of perspective-taking and negotiator focus. Journal of Personality and Social Psychology, 81(4), 657–669. Ghosh, D. (2000). Complementary arrangements of organizational factors and outcomes of negotiated transfer price. Accounting, Organizations and Society, 25(7), 661–682. Hofstede, G. (2003). What is culture? A reply to Baskerville. Accounting, Organizations and Society, 28(7–8), 811–813. Hofstede, G., Neuijen, B., Ohayv, D. D., & Sanders, G. (1990). Measuring organizational cultures: A qualitative and quantitative study across twenty cases. Administrative Science Quarterly, 35, 286–316. Kachelmeier, S. J., & Towry, K. L. (2002). Negotiated transfer pricing: Is fairness easier said than done?. The Accounting Review, 77(3), 571–593. Kolb, D. M., & Sheppard, B. H. (1985). Do managers meditate, or even arbitrate?. Negotiation Journal, 1(4), 379–388. Lax, D. A., & Sebenius, J. K. (1986). The manager as negotiator. New York: Free Press. Martin, J. (1991). A personal journey: From integration to differentiation to fragmentation to feminism. In: P. Frost (Ed.), Reframing organizational culture. Newbury Park, CA: Sage. Morrill, C. (1995). The executive way – Conflict management in corporations. Chicago, IL: The University of Chicago Press. Neale, M. A., & Bazerman, M. H. (1983). The role of perspective-taking ability in negotiating under different forms of arbitration. Industrial and Labor Relations Review, 36, 378–388. Neale, M. A., Huber, V., & Northcraft, G. (1987). The framing of negotiations: Contextual versus task frames. Organizational Behavior and Human Decision Processes, 39, 228–241. Notz, W. W., & Starke, F. A. (1978). Final offer versus conventional arbitration as means of conflict management. Administrative Science Quarterly, 23(2), 189–203. Ott, J. S. (1989). The organizational culture perspective. Pacific Grove, CA: Brooks Cole. Pinkley, R. L., Griffith, W. I., & Northcraft, G. (1995). ‘‘Fixed pie a la mode’’: Information availability, information processing, and the negotiation of sub optimal agreements. Organizational Behavior and Human Decision Processes, 62, 101–112.

140

NABIL ELIAS AND WILLIAM W. NOTZ

Pruitt, D. G., & Lewis, S. A. (1975). Development of integrative solutions in bilateral negotiation. Journal of Personality and Social Psychology, 31(4), 621–633. Schein, E. H. (1992). Organizational culture and leadership (2nd ed.). San Francisco, CA: JosseyBass. Schiff, M., & Lewin, A. (1970). The impact of people on budgets. The Accounting Review, 45(2), 259–268. Slaikeu, K. A., & Hasson, R. H. (1998). Controlling the cost of conflict. San Francisco, CA: Jossey-Bass. Sondak, H., Neale, M. A., & Pinkley, R. L. (1995). The negotiated allocation of benefits and burdens: The impact of outcome valence, contribution and relationship. Organizational Behavior and Human Decision Processes, 64, 249–260. Thomas, K. W. (1992). Conflict and negotiation processes in organizations. In: M. D. Dunnette & L. M. Hough (Eds), Handbook of industrial and organizational psychology (2nd ed., Vol. 3). Palo Alto, CA: Consulting Psychologists Press Thompson, L. (1990). Negotiation behavior and outcomes: Empirical evidence and theoretical issues. Psychological Bulletin, 108(3), 515–532. Thompson, L., & Hastie, R. (1990). Social perception in negotiation. Organizational Behavior and Human Decision Processes, 47, 98–123. Thompson, L., & Lowenstein, G. (1992). Egocentric interpretations of fairness and interpersonal conflict. Organizational Behavior and Human Decision Processes, 51, 196–197. Walton, R., & McKersie, R. (1965). A behavioral theory of labor negotiations. New York: McGraw-Hill.

THE INTERVENING EFFECT OF INFORMATION ASYMMETRY ON BUDGET PARTICIPATION AND SEGMENT SLACK
Leslie Kren and Adam S. Maiga
ABSTRACT
The objective of this study was to extend prior research by examining subordinate–superior information asymmetry as an intervening variable linking budgetary participation and slack. The results indicate two offsetting effects of participation on slack. A significant negative indirect relation between participation and slack was found to act through information asymmetry. Thus, managers reveal private information during the budget process, reducing information asymmetry which subsequently reduces budget slack. These results provide evidence about the inability of past research to confirm a consistent direct relation between budget participation and budget slack.

The problem of budgetary slack has been extensively studied by researchers in accounting. Budget slack, defined as overstated expenses, understated revenues, or underestimated performance capabilities, allows managers to obtain excess resources and to shirk more effectively. It can also be used as a
Advances in Management Accounting, Volume 16, 141–157 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16004-4

141

142

LESLIE KREN AND ADAM S. MAIGA

hedge against uncertainties that affect outcomes (Fisher, Maines, Peffer, & Sprinkle, 2002; Kren, 1997). From the organization’s perspective, however, slack budgets do not represent managers’ best estimates of expected results so they hinder planning and control, resource allocation, and coordination of business unit activities (Baiman, 1982; Covaleski, Evans, Luft, & Shields, 2003; Choudhury, 1985). Prior field-based research attempting to link budget participation to budget slack has provided inconsistent results and whether budgetary slack is a likely outcome in participatively set budgets is a matter of conjecture (Hansen, Otley, & Van der Stede, 2003). The objective of this study is to extend prior research by examining subordinate–superior information asymmetry as an intervening variable linking budgetary participation and slack. The hypotheses developed below suggest that participation is not linked directly to a reduction in segment slack but indirectly, through information asymmetry, such that more budget participation fosters a lower information asymmetry which, in turn, decreases segment slack. The consideration of information asymmetry therefore in the model adds to the existing literature on budgeting by offering insight into budget slack. Such insight is useful as it is based on reviews of the budgeting literature that consistently emphasize the need for richer theoretical models to better explain when and how participation is effective (Covaleski et al., 2003; Dunk & Nouri, 1998; Shields & Shields, 1998; Greenberg, Greenberg, & Nouri, 1994; Shields & Young, 1993). Based on a sample of segment managers in S&P 500 firms, the results show a negative indirect relation between participation and segment slack. As managers participate, they reduce information asymmetry with their superiors and subsequently lower superior–subordinate information asymmetry leading to lower budget slack. The next section provides the literature review and develops the hypotheses. Subsequent sections contain a description of the research method, an analysis of the results, and a summary and conclusion.

LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT
Participation and Information Asymmetry Information asymmetry arises when the subordinate has information relevant to the decision process associated with budgeting that is unavailable to

Information Asymmetry on Budgetary Participation and Segment Slack

143

the superior (Evans, Hannan, Krishnan, & Moser, 2001; Kren & Liao, 1988). Subordinates’ participation in the budget-setting process gives superiors the opportunity to gain access to local information if subordinates communicate or reveal some of their private information (Baiman & Evans, 1983; Covaleski et al., 2003). Therefore, participation is an organizational solution to the information asymmetry, whereby the higher the participation in budget setting, the lower the information asymmetry is expected to be. Budget participation is expected to be negatively linked to information asymmetry because participation provides a mechanism by which superiors can learn subordinate manager’s private information. As participation increases, information becomes more valuable to the superior who increases its use. This discussion is summarized as link P21 in the research model shown in Fig. 1. Stated as a hypothesis: H1. There is a negative link between budget participation and superior– subordinate information asymmetry. Information Asymmetry and Budgetary Slack In the second link in the model (P32), information asymmetry is positively linked to budget slack. When information asymmetry is lower, superiors have better information about a budgeting manager’s performance capability allowing superiors to improve ex ante inferences about the level of budget slack. Thus, as argued by Young (1985), lower superior–subordinate information asymmetry results in reduced budget slack because the budgeting manager is then aware that the superior can directly evaluate the level of slack in the budget. Merchant (1985), for example, reported that slack is negatively related to subordinate managers’ perception of their superiors’ information asymmetry (Z2)

P21

P32

budget participation (Z1)

P31

budget slack (Z3)

Fig. 1.

Theoretical Model.

144

LESLIE KREN AND ADAM S. MAIGA

ability to detect slack. Kren (1993) later reported that formal control tools, such as pre-action reviews, contact with superiors, and variances, which presumably enhanced the ability to detect slack (reduce information asymmetry), is negatively related to slack. Chalos and Haka (1990) and Evans et al. (2001), for example, demonstrated that gains to the organization can result from a reduction of information asymmetry through reduced budget slack. H2. There is a positive relation between superior–subordinate information asymmetry and budget slack. Intervening Effect In many organizational settings, a subordinate has more accurate information than his or her superior about factors influencing performance (Evans et al., 2001; Waller & Chow, 1985). However, the theoretical research in economics (e.g., Baiman & Evans, 1983) and psychology (e.g., Locke & Schweiger, 1979; Locke & Latham, 1990) assumes that participative budgeting exists to share information between a superior and subordinate. Analytical research has demonstrated that resource allocations can be improved by the communication of the subordinate’s private information (Magee, 1989; Christensen, 1982; Baiman & Evans, 1983). Budget participation serves an informational function whereby subordinates can gather, exchange, and disseminate job-relevant information to facilitate their decision-making process and to communicate their private information to organizational decision makers (Davis, DeZoort, & Kopp, 2006; Earley & Kanfer, 1985; Campbell & Gingrich, 1986; Nouri & Parker, 1998). This private information may be incorporated into the standards or budgets against which subordinate’s performance would be assessed (Baiman & Evans, 1983). Therefore, participation provides subordinates with the opportunity to reveal private information which allows central management to improve resource allocation. This information results in more realistic plans and more accurate budgets. Magner, Welker, and Campbell (1996) for example, found that budget participation ‘‘allows subordinates to introduce private information into the budgetary process, thereby enhancing the budget’s quality.’’ This was consistent with several accounting studies (e.g., Merchant, 1981; Chow, Cooper, & Waller, 1988; Murray, 1990; Kren, 1992; Magner et al., 1996; Nouri & Parker, 1998). This perspective suggests that participation is not linked directly to slack reduction but rather acts

Information Asymmetry on Budgetary Participation and Segment Slack

145

through information asymmetry. Thus, as depicted in the theoretical model (Fig. 1), we offer the following hypothesis: H3. The relationship between budgetary participation and segment slack will be explained by an indirect effect whereby participation decreases information asymmetry, and information asymmetry is positively associated with segment slack.

METHODOLOGY
Sample The objective of the sample selection procedures was to identify executivelevel profit center managers for whom objective, archival financial information to measure segment slack was available at the profit-center level. To this end, the titles of managers listed in corporate proxy statements for all S&P 500 firms were cross-referenced with each firm’s segment-level disclosures in the Compustat Industry Segment data file. Managers were retained in the initial sample if they could be identified from their job titles as managers of profit centers that clearly corresponded to segments listed in the Compustat segment disclosures. By this procedure, an initial sample was developed of 111 managers in 70 companies who were unambiguously profit center managers of reportable segments. Segment data is disclosed in accordance with FAS 14 (AICPA, 1976), which requires separate reporting for any segment which accounts for more than 10% of consolidated sales, profits, or assets. The Compustat data file includes segment sales, capital expenditures, depreciation, employee headcount, research and development, assets, and operating profit. Each segment is assigned a four-digit SIC code by Standard & Poors. A cover letter and a questionnaire were mailed to each manager in the initial sample. A follow-up letter and another copy of the questionnaire were sent after approximately three weeks. All remaining non-respondents were later contacted by telephone and another questionnaire was mailed to the specific attention of the personal assistant that worked with the respondent. Follow-ups of the original 111 managers revealed that 19 had retired, left the company, or had changed to new positions. Of the remaining 92 potential respondents, 49 usable responses were received (a response rate of 53.3%). Forty-four different companies were represented. There were two respondents for each of five companies. Respondents were

146

LESLIE KREN AND ADAM S. MAIGA

promised anonymity, but organizational affiliation was tracked to allow measurement of segment slack, as described below. Based on segment-level SIC codes, 3 of the 49 managers in the final sample were in mining and construction (SIC 0-1799), 37 were in manufacturing (SIC 1999-3999), 1 was a utility (SIC 4800-4992), 4 were in wholesale–retail (SIC 5000-5999), 3 were in banking (SIC 6000-6399), and 1 was in miscellaneous services (SIC 6400-9999).

Measurement of Variables The appendix contains an abbreviated copy of the research questionnaire used to measure the self-reported variables in this study. The reliability coefficient (Cronbach’s Alpha) for each of the self-reported scales exceeded 0.80. Budget participation was measured using the Milani (1975) six-item measure. The validity of this scale has been assessed several times in prior research, including Brownell (1983). Factor analysis confirmed the singlefactor structure of the scale. Only one factor was present with an eigenvalue greater than 1. For subsequent analysis, the six items were summed. Information asymmetry was measured using a six-item scale in Dunk (1993) and Jaworski and Young (1992). Factor analysis revealed that only one factor was present with an eigenvalue greater than 1. For subsequent analysis, the six items were summed. A precise measure of budget slack is not possible because slack manifests in a variety of ways. Moreover, different types of slack allow managers more or less discretion, which determines the amount of slack needed to pursue personal goals. Sharfman, Wolf, Chase, and Tansik (1988), for example, identified a range of resources providing varying managerial flexibility, beginning with cash, which provides the most flexibility to managers. Thus, empirical efforts to objectively measure slack have focused on determining conditions under which slack is more likely to be present by using financial variables which are presumably antecedent indicators of budget slack (Nohria & Gulati, 1996). As proposed by Leavins, Omer, and Vilitus (1995), investigation of slack behavior can be accomplished by identifying a set of measurable variables whose pattern of behavior is expected to parallel slack behavior. To this end, researchers have operationalized slack along two, non-mutually exclusive, dimensions. One dimension focuses on investments made by managers. Greater investment signals greater availability of slack resources. Investment can take the form of capital expenditures for highly flexible

Information Asymmetry on Budgetary Participation and Segment Slack

147

machine capacity and research and development expenditures (Sharfman et al., 1988; Greenley & Oktemgil, 1998). For example, Bourgeois (1981), proposed that slack occurs when managers have resources available to retain in their operations, rather than distribute to shareholders or their superiors. Thus, he argued that greater investment by managers signals greater availability of slack resources. The other dimension of objective slack measures in the literature has focused on the level of expenditures authorized by managers. Here, slack is presumed to be related to the level of expenditures in cost of goods sold, selling and administrative, research and development, and inventories because slack can accumulate in these accounts (Bourgeois, 1981). Schiff and Lewin (1970) provided confirming evidence when they reported that these expenditures decline following organizational efforts to reduce slack. Prior empirical research has focused exclusively on organizational slack. We are not aware of previous attempts in the literature to objectively measure budget slack at the sub-unit level. Following this prior literature, our operational measure of slack was based on financial accounting measures of investments and expenditures authorized by managers. Thus, we use the sum of capital expenditures and research and development as antecedent indicators of slack. These tend to be non-repetitive transactions, which according to Leavins et al. (1995), are an appropriate ‘barometer’ of slack. The sum of capital expenditures plus research and development expenditures was divided by segment sales and averaged over the three-year observation period. Data availability was a constraint since few financial accounting measures are disclosed at the segment level in the Compustat Industry Segment database or in SEC filings. To compare each segment’s available resources with other similar segments, an average of the sum of capital expenditures plus research and development expenditures divided by segment sales over the three-year period 1995–1997 was calculated for each industry reported in the Compustat Industry Segment data file. Industry was defined as all other segments listed on Compustat in the same four-digit SIC as each sample segment. When less than three other firms were in the same four-digit SIC as a sample firm, the three-digit SIC average was used, and if fewer than three firms were available in the three-digit SIC, then the two-digit SIC average was used. A sample segment was excluded in the calculation of its corresponding industry mean. In calculating the industry mean, outlier segments with three-year average capital expenditures plus research and development expenditures exceeding twice segment sales were deleted. The slack measure for each

148

LESLIE KREN AND ADAM S. MAIGA

sample firm i, was then defined as follows: Slacki ¼ industry resources À segment resourcesi .

Empirical Procedures The hypotheses are examined using the path analysis model summarized in Fig. 1. Path analysis is appropriate for estimating the relations between a series of interrelated variables (Wonnacott & Wonnacott, 1981). For this study, it allows analysis of the direct link between participation and slack and the indirect link through information asymmetry. The path coefficients, Pij, in Fig. 1, indicate the impact of variable j in explaining the variance in variable i in units of standard deviation. A series of regressions are used to estimate the path coefficients, according to the following: Information asymmetryðZ2 Þ ¼ P21 ðparticipationÞ SlackðZ3 Þ ¼ P31 ðparticipationÞ þ P32 ðinformation asymmetryÞ (1) (2)

The path coefficients can be used to decompose the total relation between two variables (i.e., slack and participation) into direct and indirect effects (i.e., through information asymmetry). The total relation is measured with the zero-order correlation coefficient, rij. Thus, r12 ¼ P21 r23 ¼ P32 þ P31 r12 r13 ¼ P31 þ P32 r12 (3) (4) (5)

The subscripts 1, 2, and 3 refer to participation, information asymmetry, and slack, respectively (Fig. 1). Model (4) allows decomposition of the total relation between information asymmetry and slack (r23) into a direct effect (P32) and a spurious effect (P31r12). The spurious effect results from participation, which is a common antecedent of both information asymmetry and slack. Model (5) allows decomposition of the total relation between participation and slack (r13) into a direct effect (P31) and the indirect effect through information asymmetry (P32r12). H3 posits that the indirect effects of participation, through information asymmetry, will predominate.

Information Asymmetry on Budgetary Participation and Segment Slack

149

RESULTS
Table 1 provides descriptive statistics for measured variables in the study. The objective slack measure, denoted segment slack, is negative for 11 (22.4%) of sample firms. Thus, on average, most of the sample firms have a greater level of resources available than the mean resources available to other industry segments listed on Compustat. The distribution for segment slack is positively skewed, although both the mean and median fall within the quartiles. Eliminating the two most extreme observations for segment slack brought the mean close to the median but had no substantive effect on the results reported later. In addition to repeating the analysis after dropping the two observations with large segment slack measures, two other sensitivity tests were performed. The analyses were repeated after dropping the two segments in the banking industry and the segment in the utility industry since control system differences may exist due to regulatory effects. In addition, the analyses were repeated after including the natural log of sales as a control for segment size (or unmeasured variables correlated with size). These additional sensitivity tests had no substantive effect on reported results. We also controlled for industry affiliation (based on SIC code). Results (not reported) indicate that industry does not have an effect on the variables. Therefore, given degrees of freedom required when the control variable was included, we removed the industry variable in order to provide additional power for the hypotheses tests. Variable correlations are shown in Table 2. The significant negative correlation between participation and information asymmetry (À0.570; Table 1. Descriptive Statistics for Propensity to Create Slack, Segment Slack, Information Asymmetry, and Participation for 49 Sample Segmentsa.
Theoretical Range/Actual Range Segment slack Information asymmetry Participation a Mean/ Median

SD

Quartile 1 Quartile 3

na/2.6 42/36 42/35

0.16/0.05 29.0/36.0 25.4/28.0

0.39 12.4 14.4

0.002 14.0 10.0

0.20 40.0 40.0

Segment slack is measured as the sum of capital expenditures plus research and development expenditures divided by segment sales, averaged over the three-year period 1995–1997. Other variables are self-reported scales as described in the paper.

150

LESLIE KREN AND ADAM S. MAIGA

Table 2. Correlations for Propensity to Create Slack, Segment Slack, Information Asymmetry, and Participation for 49 Sample Segments (Decimals Omitted)a.
2 1. Segment slack 2. Information asymmetry 3. Participation a 3 0.003 À0.570ÃÃÃ 1.00

0.350ÃÃ – –

Segment slack is measured as the sum of capital expenditures plus research and development expenditures divided by segment sales, averaged over the three-year period 1995–1997. Other variables are self-reported scales as described in the paper. ÃÃ po0.05. ÃÃÃ po0.01.

po0.01) is consistent with H1. The positive correlation between information asymmetry and segment slack (0.350; po0.05) is also consistent with H2. These relations will be explained more clearly as the path analysis results are discussed next. The results of estimating the research models (1) and (2) are shown in Table 3 and Fig. 2. As observed in Table 2, the link between information asymmetry and budget participation (P21), is significant and the sign is negative as expected (model 1, Table 3, and Fig. 2). The negative sign indicates that as participation increases, information asymmetry decreases and suggests that managers reveal information to their superiors during budget participation, reducing the level of information asymmetry. This result is consistent with H1. This result obtained appears to be consistent with prior research results that suggest that even though managers can benefit from budget slack, the nature of the budgeting process may in fact provide incentives for (self-interested) managers to reveal their private information and reduce budget slack (Kren, 1997). First, managers may find that obtaining adequate resources for their area of responsibility requires that they disclose some of their private information and reduce budget slack. For example, Nouri and Parker (1998) propose that participative budgeting is an important means by which managers can influence the resources they receive. Hopwood (1974) and Fisher, Frederickson, and Peffer (2000) similarly suggested that budgeting is a ‘bargaining’ process. As they bargain for resources, budgeting managers will need to reveal some of their private information. Managers are also likely to disclose information and reduce budget slack while they attempt to secure job-relevant information

Information Asymmetry on Budgetary Participation and Segment Slack

151

Table 3.

Path Analysis Results for 49 Sample Segments.

Model 1 : Information asymmetry ðZ2 Þ ¼ P21 ðparticipationÞ þ . Model 2 : Slack ðZ3 Þ ¼ P31 ðparticipationÞ þ P32 ðinformation asymmetryÞ þ . Dependent Variable (t-Statistics in Parentheses) Model 1 Information Asymmetry Participation Information asymmetry R-square F-statistics À0.570 (À4.76ÃÃÃ) 0.33 22.7ÃÃÃ Model 2 Segment Slack 0.288 (1.86Ã) 0.499 (3.22ÃÃ) 0.18 5.18ÃÃ

Note: Segment slack is measured as the sum of capital expenditures plus research and development expenditures divided by segment sales, averaged over the three-year period 1995–1997. Other variables are self-reported scales as described in the paper. Ã po0.10. ÃÃ po0.05. ÃÃÃ po0.01.

P21 = -.570

Information asymmetry (Z2)

P32 = .499

Budget participation (Z1)

P31 = .288

Segment slack (Z3)

Fig. 2.

Research Model Estimates.

to evaluate alternative budget goals (Campbell & Gingrich, 1986; Early, Wojnaroski, & Prest, 1987; Lawrence & Lorsch, 1967; Kren, 1992). As their information search activities proceed, some of the manager’s own private information becomes incorporated into the budget because social norms require an information exchange as managers communicate with superiors (Fisher et al., 2000; Hopwood, 1974; Simons, 1987). To evaluate the model linkages, the research model is decomposed, as shown in Table 4. The direct path relations are also shown in Fig. 2.

152

LESLIE KREN AND ADAM S. MAIGA

Table 4.

Decomposition of Path Analysis Relations for 49 Sample Segments.
Model 3 : r12 ¼ P21 , Model 4 : r23 ¼ P32 þ P31 r12 , Model 5 : r13 ¼ P31 þ P32 r12 ,

where, Z1 is participation, Z2 information asymmetry, Z3 propensity to create slack or segment slack. Dependent Variable/Link to Model 3: Information asymmetry/participation Model 4: Segment slack/ information asymmetry Model 5: Segment slack/ participation Total Effect rij À0.570ÃÃÃ 0.350ÃÃ 0.003 Direct Effect Pij À0.570ÃÃÃ 0.499ÃÃ 0.288Ã Indirect Effect – – À0.285Ã Spurious Effect – À0.149 –

Note: Segment slack is measured as the sum of capital expenditures plus research and development expenditures divided by segment sales, averaged over the three-year period 1995–1997. Other variables are self-reported scales as described in the paper. Ã po0.10. ÃÃ po0.05. ÃÃÃ po0.01.

For the slack measure, denoted segment slack (Fig. 2; Table 4), the result indicates a significant direct path from information asymmetry (P32) to segment slack. Only a small portion of the link is spurious (0.149) relative to the direct effect (0.499). This result is consistent with H2 and supports the argument that lower superior–subordinate information asymmetry results in reduced budget slack, presumably because the superior can directly evaluate the level of slack in the manager’s budget. Combined with the negative link between budget participation and information asymmetry, the positive link between information asymmetry and segment slack result in a significant negative indirect relation (À0.285) between segment slack and budget participation. As shown in model 5, Table 4, for every standard deviation increase in participation, information asymmetry decreases by À0.570 standard deviation (model 3), and for every standard deviation decrease in information asymmetry, segment slack decreases by 0.499 standard deviation (model 4). Thus, the total indirect effect of every standard deviation increase in participation, is a decrease in segment slack of À0.285 standard deviation through information asymmetry (model 5). This negative indirect effect is the sign predicted by H3,

Information Asymmetry on Budgetary Participation and Segment Slack

153

i.e., managers have incentives to reveal private information during the budget process and concomitantly segment slack is reduced. In contrast to H3, the result also indicate a significant direct path to segment slack (P31 ¼ 0.288). This direct effect indicates that for every standard deviation increase in participation, segment slack increases by 0.288 standard deviation (model 5). This result is consistent with agency arguments that budgeting managers are motivated to bias their budget estimates for self-interest and that participation provides the mechanism for managers to insert slack into their budgets. The findings imply that, for superior to reduce segment slack in a participating budget setting, reduction in information asymmetry may be used as a control mechanism to achieve this end.

SUMMARY AND CONCLUSION
The objective of this study was to examine subordinate–superior information asymmetry as an intervening variable linking budgetary participation to budget slack. Predictions were that participation would be negatively related to information asymmetry which would subsequently be positively related to budget slack. Using path analysis, the evidence supports the negative link between budget participation and information asymmetry. This suggests that participating managers reduce information asymmetry with their superiors. The subsequent positive link between segment slack and information asymmetry was also supported. This finding is consistent with the argument that lower superior–subordinate information asymmetry leads to lower budget slack because the superior can directly evaluate the level of slack in the budget. The indirect effect was sizable. For every standard deviation increase in participation, segment slack decreases by 0.285 standard deviation. However, the result for segment slack indicates a positive significant direct effect of participation on segment slack. Overall, this is consistent with the argument that participation, through information asymmetry, does not provide the vehicle for managers to insert slack into their budgets. Limitations of this study should be mentioned. First, segment slack was measured within data availability constraints that represented resources most directly controllable by the current segment manager. We suggest that future research include additional objective slack measures. Second, of course, is the size of the sample. This is a sample of only 49 business segments. Although that is encouraging in that it provides a very conservative

154

LESLIE KREN AND ADAM S. MAIGA

test of our model and suggests that the effects we see are quite strong, we must be especially cautious in generalizing. Third, the paper only focused on one intervening variable designed to reduce budget slack. Attempt to reduce budget slack may be achieved through other variables (e.g., manager’s incentive, commitment, perception of fairness). Consequently, future research should concentrate on developing intervening variables designed to reduce budget slack. These intervening variables would be more likely help understand the conflicting results based on the direct link between budget participation and budget slack. Even with these limitations, however, the findings are both intuitively and practically significant because they demonstrate the process by which budgetary participation translates into lower budget slack. Our understanding is thus enhanced by the recognition of information asymmetry as intervening variable between the level of budget participation and budget slack. Furthermore, the findings provide practical guidance to managers involved in resource allocation decisions. These results provide some evidence about the inability of past research to confirm a consistent direct relation between budget participation and budget slack.

REFERENCES
AICPA. (1976). Financial Reporting for Segments of a Business Enterprise. FASB No. 14. Baiman, S. (1982). Agency research in managerial accounting: A survey. Journal of Accounting Literature, 1, 154–213. Baiman, S., & Evans, J. H. III. (1983). Pre-decision information and participative management control systems. Journal of Accounting Research, 21(Autumn), 371–395. Bourgeois, L. J. (1981). On measurement of organizational slack. Academy of Management Review, 6(1), 29–39. Brownell, P. (1983). Leadership style, budgetary participation and managerial behavior. Accounting, Organizations and Society, 8(4), 307–321. Campbell, D. J., & Gingrich, K. F. (1986). The interactive effects of task complexity and participation on task performance: A field experiment. Organizational Behavior and Human Decision Processes, 38, 162–180. Chalos, P., & Haka, S. (1990). Participative budgeting and managerial performance. Decision Sciences, 20, 334–347. Choudhury, N. (1985). Incentives for the divisional manager. Accounting and Business Research, 16(Winter), 11–21. Chow, C. W., Cooper, J. C., & Waller, W. S. (1988). Participative budgeting: Effects of a truthinducing pay scheme and information asymmetry on slack and performance. Accounting Review, 63, 111–122. Christensen, J. (1982). The determination of performance standards and participation. Journal of Accounting Research, 26, 589–603.

Information Asymmetry on Budgetary Participation and Segment Slack

155

Covaleski, M. A., Evans, J. H., Luft, J. L., & Shields, M. D. (2003). Budgeting research: Three theoretical perspectives and criteria for selective integration. Journal of Management Accounting Research, 15, 3–49. Davis, S., DeZoort, T. T., & Kopp, L. S. (2006). The effects of obedience pressure and perceived responsibility on management accountants’ creation of budgetary slack. Behavioral Research in Accounting, 18, 19–35. Dunk, A. S. (1993). The effect of budget emphasis and information asymmetry on the relation between budgetary participation and slack. The Accounting Review, 68(2), 400–410. Dunk, A., & Nouri, H. (1998). Antecedents of budgetary slack: A literature review and synthesis. Journal of Accounting Literature, 17, 72–96. Earley, P. C., & Kanfer, R. (1985). The influence of component participation and role models on goal acceptance, goal satisfaction and performance. Organizational Behavior and Human Decision Processes, 36, 378–390. Early, P. C., Wojnaroski, P., & Prest, W. (1987). Task planning and energy expended: Exploration of how goals influence performance. Journal of Applied Psychology, 72, 107–114. Evans, J. H., Hannan, R. L., Krishnan, R., & Moser, D. V. (2001). Honesty in managerial reporting. The Accounting Review, 76(4), 537–559. Fisher, J. G., Frederickson, J. R., & Peffer, S. A. (2000). Budgeting: An experimental investigation of the effects of negotiation. The Accounting Review, 75(1), 93–114. Fisher, J. G., Maines, L. A., Peffer, S. A., & Sprinkle, G. B. (2002). Using budgets for performance evaluation: Effects of resource allocation and horizontal information asymmetry on budget proposals, budget slack, and performance. The Accounting Review, 77(4), 847–865. Greenberg, P., Greenberg, R. H., & Nouri, H. (1994). Participative budgeting: A meta-analytic examination of methodological moderators. Journal of Accounting Literature, 13, 117–141. Greenley, G. E., & Oktemgil, M. (1998). A comparison of slack resources in high and low performing British companies. Journal of Management Studies, 35(3), 377–398. Hansen, S. C., Otley, D. T., & Van der Stede, W. A. (2003). Practice developments in budgeting: An overview and research perspective. Journal of Management Accounting Research, 15, 95–116. Hopwood, A. G. (1974). Accounting and human behavior. Englewood Cliffs, NJ: Prentice-Hall. Jaworski, B. J., & Young, S. M. (1992). Dysfunctional behavior and management control: An empirical study of marketing managers. Accounting, Organizations and Society, 17(1), 17–35. Kren, L. (1992). Budgetary participation and managerial performance: The impact of information and environmental volatility. The Accounting Review, 67(July), 511–526. Kren, L. (1993). Control system effects on budget slack. Advances in Management Accounting, 2, 109–118. Kren, L. (1997). The role of accounting information in organizational control: The state of the art. In: S. Sutton & V. King (Eds), Behavioral accounting research: Foundations and frontiers (pp. 1–48). Gainseville, FL: American Accounting Association. Kren, L., & Liao, W. M. (1988). The role of accounting information in the control of business organizations: A review of the evidence. Journal of Accounting Literature, 7, 280–309. Lawrence, P. R., & Lorsch, F. W. (1967). Organization and environment. Boston, MA: Harvard University. Leavins, J. R., Omer, K., & Vilitus, A. (1995). A comparative study of alternative indicators of budgetary slack. Managerial Finance, 21(3), 52–67.

156

LESLIE KREN AND ADAM S. MAIGA

Locke, E., & Latham, G. (1990). A theory of goal setting and task performance. Englewood Cliffs, NJ: Prentice-Hall. Locke, E. A., & Schweiger, D. M. (1979). Participation in decision making: One more look. In: B. M. Staw (Ed.), Research in Organizational Behavior. Elsevier, Amsterdam: The Netherlands. Magee, R. (1989). Equilibria in budget participation. Journal of Accounting Research, (Autumn), 551–573. Magner, N., Welker, R. B., & Campbell, T. L. (1996). Testing a model of cognitive budgetary participation processes in a latent variable structural equations framework. Accounting and Business Research, 27, 41–51. Merchant, K. A. (1981). The design of the corporate budgeting system: Influences on managerial behavior and performance. The Accounting Review, 56, 813–829. Merchant, K. A. (1985). Budgeting and the propensity to create budget slack. Accounting, Organizations and Society, 10(2), 201–210. Milani, K. (1975). The relationship of participation in budget setting to industrial supervisor performance and attitudes: A field study. The Accounting Review, 50(April), 274–284. Murray, D. (1990). The performance effects of participative budgeting: An integration of intervening and moderating variables. Behavioral Research in Accounting, 2, 104–123. Nohria, N., & Gulati, R. (1996). Is slack good or bad for innovation?. Academy of Management Journal, 39(5), 1245–1264. Nouri, H., & Parker, R. J. (1998). The relationship between budget participation and job performance: The roles of budget adequacy and organizational commitment. Accounting, Organizations and Society, 23, 467–483. Schiff, M., & Lewin, A. Y. (1970). The impact of budgets on people. Accounting Review, 45(April), 259–268. Sharfman, M. P., Wolf, G., Chase, R. B., & Tansik, D. A. (1988). Antecedents of organizational slack. Academy of Management Review, 13(4), 601–614. Shields, J. F., & Shields, M. D. (1998). Antecedents of participative budgeting. Accounting, Organizations and Society, 23, 49–76. Shields, M., & Young, S. M. (1993). Antecedents and consequences of participative budgeting: Evidence on the effects of information asymmetry. Journal of Management Accounting Research, 5, 265–280. Simons, R. (1987). Accounting control systems and business strategy: An empirical analysis. Accounting, Organizations and Society, 12(4), 357–374. Waller, W. S., & Chow, C. W. (1985). The self-selection and effort effects of standard-based employment contracts: A framework and some empirical evidence. The Accounting Review, 60, 458–476. Wonnacott, T. H., & Wonnacott, R. J. (1981). Regression: A second course in statistics. Malabar, FL: Robert E. Krieger Publishing. Young, M. S. (1985). Participative budgeting: The effects of risk-aversion and asymmetric information on budgetary slack. Journal of Accounting Research, 23(Autumn), 829–842.

Information Asymmetry on Budgetary Participation and Segment Slack

157

APPENDIX: ABBREVIATED RESEARCH QUESTIONNAIRE
Participation Response anchors: 1 ¼ strongly disagree, 7 ¼ strongly agree. Q1. Q2. Q3. Q4. Q5. Q6. I am involved in setting all of my budget. My superior clearly explains budget revisions. I have frequent budget-related discussions with my superior. I have a great deal of influence on my final budget. My contribution to the budget is very important. My superior initiates frequent budget discussions when the budget is being prepared. Information Asymmetry Response anchors: 1 ¼ strongly disagree, 7 ¼ strongly agree. Q1. In comparison to my superior, I have better information regarding the activities in my area of responsibility. Q2. In comparison to my superior, I am more familiar with the input– output relations in my area of responsibility. Q3. In comparison to my superior, I am more familiar with the performance potential of my area of responsibility. Q4. In comparison to my superior, I am more familiar technically with my area of responsibility. Q5. In comparison to my superior, I am better able to assess the impact of external factors on my area of responsibility. Q6. In comparison to my superior, I have a better understanding of what can be achieved in my area of responsibility.

DO ACCOUNTING PERFORMANCE MEASURES INDEED REDUCE MANAGERIAL AMBIGUITY UNDER UNCERTAINTY?
Frank G. H. Hartmann
ABSTRACT
Research on budget-based performance evaluation traditionally predicts that the use of accounting performance measures (APM) in complex, dynamic, and uncertain situations results in dysfunctional managerial attitudes and behaviors. Although this suggests that such situations require the use of subjective performance measures (SPM), empirical evidence is inconclusive, as APM, rather than SPM, have been found to also have a negative effect on managerial ambiguity. This suggests that APM may be more, rather than less, appropriate than SPM in situations of high uncertainty. This paper explores whether acknowledgement of different types of uncertainty may explain these apparently conflicting research findings. It develops hypotheses that predict differential interactions between the environmental uncertainty and task uncertainty and APM and SPM on managerial ambiguity. These hypotheses are tested using survey data from 250 managers in 11 organizations. Tests using moderated regression analysis provide support for the existence of different interactions between uncertainty and the use of performance measures, and provide reconciliation for the opposing findings in the extant literature.
Advances in Management Accounting, Volume 16, 159–180 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16005-6

159

160

FRANK G. H. HARTMANN

1. INTRODUCTION
The negative effects of using budgets for managerial performance evaluation forms a central theme in the accounting literature (Briers & Hirst, 1990). Empirical studies generally seek to confirm the expectation that budget-based accounting performance measures (APM) cause dysfunctional managerial attitudes and behaviors, particularly in situations of high organizational uncertainty and complexity (Hartmann, 2000). In such situations the lack of controllability and completeness generally associated with budget-based performance metrics is expected to become especially harmful, and the implication is that supervisors should rely on subjective performance measures (SPM) instead (Hopwood, 1972; Moers, 2005). Empirical evidence is inconclusive however. Recently, Marginson and Ogden (2005) therefore criticized the research emphasis on the negative sides of budgetbased performance metrics, and blame it for our current poor understanding of the more positive consequences of using APM. In contrast with standard predictions, they argue and show that budget-based performance metrics may be especially useful in uncertain and complex organizational contexts, as budget-based control may provide individual managers with the paths and goals to cope with such contexts (House, 1971; Marginson & Ogden, 2005, p. 436). In particular, budgets will serve as an antidote for managerial ambiguity (MANAAMBI), which is a lack of clarity managers in decentralized organizations may have about their role, their job objectives, and the scope of their responsibilities (e.g., Vancil, 1979; Sawyer, 1992). The arguments and findings of Marginson and Ogden (2005) are important since they allow better explanations of the continuing importance of budgets in contemporary organizations than studies that merely address the negative sides of budgeting (cf. Umapathy, 1987; Fisher, Maines, Peffer, & Sprinkle, 2002). They are problematic at the same time, however, as they contradict some earlier findings in the APM literature, without attempting their reconciliation. Indeed, Marginson and Ogden (2005, p. 435) are merely concluding ‘‘budgets [y] may be as useful to the individual as they are problematic’’. The purpose of the present paper is to provide a theoretical and empirical reconciliation of the positive and negative effects of the use of APM and SPM under uncertainty. It argues that not uncertainty per se, but rather the type of uncertainty determines whether APM or SPM have desirable or undesirable effects. In particular, it distinguishes between task uncertainty (TU) and environmental uncertainty (EU), and argues that these different types of uncertainty have a different effect on the appropriateness of APM and SPM to reduce MANAAMBI. The remainder of the

Do Accounting Performance Measures Indeed Reduce Managerial Ambiguity

161

paper is structured as follows. Section 2 provides a brief overview of the extant literature, resulting in the formulation of hypotheses about the effects of APM and SPM on MANAAMBI under task and EU. Section 3 presents the method of the empirical survey study conducted to gather data. Section 4 reports the tests of the hypotheses. Section 5 concludes this paper with a discussion of the findings and their implications.

2. LITERATURE REVIEW AND DEVELOPMENT OF HYPOTHESES
Overviews of the empirical literature that investigates the contextual appropriateness of APM suggest that predictions about its adverse effects are contingent on managers perceiving the organizational context as uncertainty (Chapman, 1997; Hartmann, 2000; Covaleski, Evens, Luft, & Shields, 2003). Perceived uncertainty in general reflects the beliefs of human actors about their inability to predict future states of the world, including the outcomes of their actions (cf. Chapman, 1997; Hartmann, 2000). Uncertainty is an important source of MANAAMBI as it concerns the impact of context on the manager’s performance, and the required managerial actions and responses (cf. Vancil, 1979; Milliken, 1987; Marginson & Ogden, 2005; Chun & Rainey, 2005). In an early study, Hirst (1983) thus suggested and found that the use of APM under high uncertainty resulted in managers experiencing high job-related tension. Building on earlier budgetary studies whose contradictory evidence suggested that any effects of APM would be context specific rather than universal (Hopwood, 1972; Otley, 1978), Hirst argued that contextual uncertainty would negatively affect managers’ beliefs about the controllability, the completeness, and the relevance of budgetary performance targets (cf. Hopwood, 1972). Under high uncertainty, stressing such targets through the use of APM would create tension between the ‘true’ and the budgetary job requirements, which would result in job-related tension. Subsequent studies that attempted to replicate Hirst’s finding were only partially successful, as they found high uncertainty to have no effects (e.g., Govindarajan, 1984; Imoisili, 1989; Ross, 1995), and even positive effects (e.g., Ezzamel, 1990), rather than the expected negative effects (e.g., Brownell, 1985). Ross (1995, p. 9) accounted for his failure to replicate the negative effect of APM under uncertainty by suggesting that replacing APM with SPM could have negative consequences as well. High uncertainty would intensify the lack of clarity and objectivity associated with subjective evaluations, which could enhance MANAAMBI about both the required

162

FRANK G. H. HARTMANN

and achieved performance levels (cf. Marginson & Ogden, 2005). Although not broadly recognized in the APM literature, this explanation of positive effects of using APM under uncertainty seems consistent with the goaltheoretic explanations on the effect of goal specificity on individuals’ motivation and task performance (Locke & Latham, 1990). Reviews of the goal-theoretic literature done by Locke and Latham (1990) and Rodgers and Hunter (1991) provide firm evidence that specific goals, rather than ‘do-your-best goals’, increase job satisfaction and performance, and decrease job-related tension and MANAAMBI. Clearly, the budgeting system is an important source of such goals as is indeed emphasized throughout the normative goal-theoretic literature (e.g., McConkie, 1979), which presents budget-based responsibility accounting systems as a specific example of applied goal-setting (see, e.g., Schuler, Beutell, & Youngblood, 1989). Earlier, Tosi (1975, p. 150), summarizing available goal-setting evidence, even suggested ‘‘y the motivating effect of the budget derives from simply the fact that it is a statement of explicit goals’’. Also Hopwood (1972, p. 173) acknowledges this potential function of budgets, stating ‘‘y one purpose of the budget is to clearly set out the objectives for a cost center. [y] and thereby add an important element of structure and clarity to the job environment’’, but he did not study how this function would work out under uncertainty. Although the use of budgets for performance evaluation under uncertainty may be thus indeed be both useful and problematic (cf. Marginson & Ogden, 2005), little knowledge is gained beyond budgets’ potential effects unless we study the exact conditions under which budgets enhance (cf. Hirst, 1983) or reduce (cf. Ross, 1995) MANAAMBI. It is very likely that part of the required reconciliation of these opposite findings may come from a further understanding of the nature of uncertainty that affects the appropriateness of APM (and SPM). As Hartmann (2000) observed in his review of this area of the empirical literature, empirical studies have used two uncertainty constructs interchangeably, EU and TU, with apparent disregard for their differential nature and their potentially rather different implications for the appropriateness of budgetbased performance evaluation (cf. Chapman, 1997; Tymon, Stout, & Shaw, 1998; Hartmann, 2000, p. 445). These two sorts of uncertainty are introduced now, after which their implications for the appropriateness of both APM and SPM will be discussed. The distinction between EU and TU originates in the work of early contingency theorists such as Thompson (1967). Analyzing uncertainty associated with managerial work, Thompson (1967, Chapters 5, 6) distinguishes between uncertainty stemming from sources outside the manager’s area of

Do Accounting Performance Measures Indeed Reduce Managerial Ambiguity

163

responsibility (i.e., EU), and uncertainty originating in the technological nature of the managerial tasks proper (i.e., TU). Regarding the latter type of uncertainty, the two determinants of this type of uncertainty are the lack of repetitiveness and the lack of programmability of tasks, as these factors both obstruct complete knowledge about cause–effect relationships when performing the task (e.g., Hirst, 1987; Campbell, 1988). When TU is high, managers have difficulty predicting the effects of their actions, which causes ambiguity concerning the courses of action that should be taken in order to obtain desired results, and performance feedback (Luckett & Eggleton, 1991). EU, instead, refers to factors – people, things, laws, and regulations – that make up the manager’s external task environment (Downey, Hellriegel, & Slocum, 1975; Ewusi-Mensah, 1981; Tymon et al., 1998). When EU is high, the managerial challenge is to define what desirable results are to which effort should be directed, given the complex and changing demands and pressures from the context on the managerial task portfolio. High EU will, therefore, cause ambiguity concerning the nature of the results to be achieved by the manager, and their prioritization. The implications of these two types of uncertainty for the appropriateness of APM and SPM are analyzed separately below, starting with the case for EU. Despite the mentioned common believe that uncertainty generally reduces the usefulness of budgets and APM, there is considerable evidence that suggests the importance of formal planning systems in general in situations characterized by EU. Horngren, Sundem, and Stratton (1996, p. 257), explaining the role of budgeting in organizations, for example note: Skeptical managers have claimed, ‘‘I face too many uncertainties and complications to make budgeting worthwhile for me’’. Be wary of such claims. Planning and budgeting are especially important in uncertain environments. A budget allows systematic rather than chaotic reaction to change. This normative, textbook, argument has received considerable empirical support over the last two decades. Merchant (1989) illustrates that EU is not merely an exogenous factor affecting (budget) plans after preparation, but also an endogenous factor providing input for the (budget) planning process, which forces organizational participants to clarify necessary actions. Simons (1987, 1995) argues that EU enhances the roles of formal controls, such as APM, as he argues ‘‘y uncertain environments require more information sharing and personal exchange. The sharing and exchange process, however, does not occur in the absence of formal control procedures’’ (Simons, 1987, p. 359). Formal controls may in fact stimulate the interactions required between organizational participants in situations of external dynamism, as they are used in an interactive fashion (cf. Simons, 1995). This explains that

164

FRANK G. H. HARTMANN

EU has been found to enhance the usefulness of formal cost controls (Khandwalla, 1972), and of budgeting systems in general (Abernethy & Brownell, 1999). At the individual managerial level, formal (budgetary) control systems take on an important additional role, which has been described as the ‘buffering function’ of budgets. Merchant (1984, p. 293) for example notes that, in the context of EU stemming from unpredictable market developments ‘‘y the budget can be used for buffering by treating it as a fixed performance commitment, regardless of how the market is changing’’. Such a buffering function of control systems has been documented in the wider organizational and administrative literatures as well (e.g., Thompson, 1967; Hayes, 1977; Olson & Rombach, 1996), and is deemed important to reduce MANAAMBI by shielding the manager’s immediate working environment from its external context (cf. Olson & Rombach, 1996; Chun & Rainey, 2005). For the individual manager, these positive effects of budget-based APM will result in less MANAAMBI, and more clarity about their managerial goals, roles, and responsibilities. This seems consistent with the recent arguments put forward by Marginson and Ogden (2005). In situations of high EU, APM serve to communicate a limited set of clear, fixed, and relatively objective performance standards to subordinate managers (Galbraith, 1977; Gordon & Narayanan, 1984; Ezzamel, 1990). The corollary of this argument is that the use of APM reduces the ambiguity associated with subjectivity performance evaluations, which have been argued to be especially harmful under high EU (see, e.g., Rizzo, House, & Lirtzmann, 1970; Ezzamel, 1990; Ross, 1995). While expecting a negative moderating effect of EU on the relationship between the use of APM and MANAAMBI, we therefore expect a positive moderating effect of EU on the relationship between the use of SPM and MANAAMBI. The following two hypotheses will be tested accordingly. H1. The effect of APM on MANAAMBI is more negative for higher levels of EU. H2. The effect of SPM on MANAAMBI is more positive for higher levels of EU. For the impact of TU on the relationship between the use of APM and SPM, and MANAAMBI the argumentation is different than for EU. Earlier, TU was defined as uncertainty caused by the inherent complexity and diversity of the managerial portfolio of tasks (cf. Hirst, 1983; Wood,

Do Accounting Performance Measures Indeed Reduce Managerial Ambiguity

165

Mento, & Locke, 1987; Campbell, 1988). Although the positive effects of formal control systems may appear just as desirable in situations of high TU as they are in situations of high EU (Hirst, 1983; Early, Prest, & Wojnaroski, 1987; Ross, 1995), the extent to which APM and SPM may provide such clarity and structure is fundamentally different. TU is an inherent job characteristic, as it relates to knowledge about cause–effect relationships that is limited because of the complexity and diversity of managerial task portfolio. Although managers may have certain goals for their job that are specific and ‘objectively’ clear, high TU will generally cause managers to feel undecided about the actions needed to attain those goals (Early, 1985). Locke and Latham (1990, p. 260) thus observe that the effectiveness of goal setting will be lower under high TU, as under such conditions effort does not necessarily pay off directly, and managers will feel ambiguous about where and how to allocate their efforts (cf. Early, 1985; Hirst, 1987). Managers confronted with high TU will thus suffer from what Milliken (1987) has labeled ‘response uncertainty’, which reflects the ambiguity managers feel about the way in which a certain objective should be achieved given the lack of precise cause–effect knowledge. Therefore, it is expected that formal systems that dictate the attainment of a set of clear and objective performance standards may in fact lead to confusion, and not clarity, over the way in which high performance may be attained (cf. Hirst, 1987; Early, 1985; Wood et al., 1987; Sawyer, 1992). Instead, high TU may provide a rationale for more subjective ways of evaluating managerial performance (Hopwood, 1972; Hirst, 1983, 1987). The use of subjective elements in their evaluation allows superiors to use their discretion when assessing managerial efforts, which allows them to provide feedback on effort and performance in cases where the mere attainment of preset goals is not informative because of the said lack of cause–effect knowledge (see, e.g., Locke & Latham, 1990; Luckett & Eggleton, 1991). Overall, therefore, this leads to the expectation that TU has a positive moderating effect on the relationship between the use of APM and MANAAMBI, and a negative moderating effect on the relationship between the use of SPM and MANAAMBI. The following two hypotheses will be tested accordingly. H3. The effect of APM on MANAAMBI is more positive for higher levels of TU. H4. The effect of SPM on MANAAMBI is more negative for higher levels of TU.

166

FRANK G. H. HARTMANN

3. RESEARCH METHOD
To test the hypotheses, data were gathered through a questionnaire survey. The choice for the survey method was made to allow comparison with the empirical studies on which this study builds, and which it attempts to reconcile, and to enable the assessment of the use subjective performance measurement, as this is not typically documented (Ittner, Larcker, & Meyer, 2003, p. 8). The survey was conducted in the Netherlands following a sampling method that had been successful in previous studies in this field (e.g., Merchant, 1984; Ross, 1995). First a sample of organizations was selected and asked for participation. Then, a sample of managers within those organizations was approached. This stratification method provides an acceptable proxy to random sampling where such is not economically feasible, or would likely result in large non-response bias (e.g., Brownell, 1995; Pedhazur & Pedhazur-Schmelkin, 1991). Random sampling, which requires the a priori establishment of the population from which a sample is drawn, and to which inferences are made, was not possible as detailed and complete data on managers with budget responsibility are simply not available publicly (cf. Alreck & Settle, 1985; Oppenheim, 1992). Moreover, the method applied ensures variation in the key independent variables as organizations were selected from various industries (Emory & Cooper, 1991, p. 275), and reduces the risk of drawing inferences from potential idiosyncrasies of single organizations (Alreck & Settle, 1985, p. 45; Pedhazur & PedhazurSchmelkin, 1991, p. 320). Out of the 12 organizations initially approached, 11 organizations agreed to take part in the study. In each organization interviews were conducted with one or more senior management officials, which served two purposes. A first purpose was to learn about the organizations’ budgeting and performance evaluation systems to assess the applicability of the research topic for each organization. A second purpose was to ask for formal organizational support for conducting the research in the organization. To maintain anonymity and avoid selection bias, in each organization one official was asked to select respondents (cf. Brownell, 1995) after being instructed to select a broad and large sample of responsibility center managers, across functional areas and positions in the organizational hierarchy, including line and staff managers, and of a single (Dutch) nationality. These managers would be in charge of a distinct area of responsibility, would have at least one functional subordinate (cf. Hirst, 1983; Ross, 1995), a separate budget, and would have experienced at least one budgeting and performance evaluation cycle. In total, 250 managers were selected. The sample size per organization ranged from 9 to 56, reflecting the

Do Accounting Performance Measures Indeed Reduce Managerial Ambiguity

167

Table 1.
Organization

Descriptives of 11 Participating Organizations.
Main Activity Production Production Retail Production Production Production Production Project development Governmental services Legal services Defense Sample 10 34 27 15 25 9 24 25 15 56 10 250 Usable Response 7 29 19 11 21 6 17 19 12 48 7 (70%) (85%) (70%) (73%) (88%) (67%) (68%) (76%) (80%) (86%) (70%)

1. Chemicals 2. Consumer electronics 3. Consumer electronics 4. Automotive 5. Food and drink 6. Food and drink 7. Electronic office equipment 8. City development 9. City administration 10. National administration 11. National government Total

196 (78.4%)

organization’s size and responsibility center structure. Table 1 below contains descriptive statistics on the sample of these 11 participating organizations.

3.1. Variable Measurement EU was measured with a scale previously used in Govindarajan (1984) and Merchant (1984). The scale contains five attributes of the respondent’s organizational environment, which are the behavior of (1) customers, (2) competitors, and (3) suppliers, (4) the rate of technological change in work area, and (5) the rate of legal and political developments. For each of these five attributes respondents were asked to indicate on six-point, fully verbally anchored Likert scales, both the perceived predictability and the perceived impact on work and performance (cf. Khandwalla, 1972), which resulted in a ten-item overall instrument. Subsequent factor analysis revealed that the ten items loaded on three factors.1 Although this apparent multidimensionality has been interpreted as support for the multidimensional nature of EU in some studies (e.g., Downey et al., 1975; Buchko, 1994), scores for this study were based on the six items that loaded on the first factor since the specific aim of this study is to sharply delineate EU from other types of uncertainty. These six items related to customers, competitors and technological developments. Cronbach’s a for the reduced scale was a satisfactory 0.77, exceeding scores obtained in previous occasions

168

FRANK G. H. HARTMANN

(Govindarajan, 1984; Merchant, 1990). Variable scores for EU were calculated using the average value of the scores on the six items. TU was measured with an instrument developed by Withey, Daft, and Cooper (1983) that asks respondents to indicate on five-point fully anchored scales their level of agreement with nine statements related to the complexity and the variety of their tasks. The instrument was used in Brownell and Hirst (1986) showing high reliability. The Cronbach’s a for the nine-item scale was a satisfactory 0.87. Subsequent factor analysis revealed two factors that, however, did not correspond to the dimensions of variety and complexity that Brownell and Hirst (1986) distinguished. Because of the high and maximal Cronbach’s a, the overall score for TU was therefore calculated as the mean of the scores on all nine items. The use of APM and SPM was measured with the scale developed and used by Hofstede (1967) and Hopwood (1972), which has become the most frequently used instrument in extant research (cf. Hartmann, 2000; Vaigneur & Peiperl, 2000). This instrument contains three items related to APM and five items related to SPM. Factor analysis attested to the two-factor structure of the instrument, as Table 2 displays. The three items related to APM loaded on one factor that explained 30.5% of variance, and had a Cronbach’s a of 0.72. The five items related to SPM loaded on the second factor that explained 25.0% of variance, and had a Cronbach’s a of 0.73. The average scores of the respective sets of items were used to form variable scores for APM and SPM.

Table 2.
Q. 1.a 1.b 1.c 2.a 2.b 2.c 2.d 2.e

Factor Analysis of Items for APM and SPM Scales (Varimax Rotation).
Wording (Abbreviated) Concern with costs My (budgetary) results Whether I (always) meet my budget Cooperation with colleagues Effort put into work Concern with quality Attitude toward work and organization Ability to handle my employees Scale APM APM APM SPM SPM SPM SPM SPM Factor 1 Factor 2 0.198 0.009 0.021 0.710 0.597 0.658 0.771 0.714 2.703 30.463 0.834 0.844 0.726 À0.075 À0.011 0.165 0.133 0.126 1.737 25.036

Eigenvalue Percentage of variance explained

Do Accounting Performance Measures Indeed Reduce Managerial Ambiguity

169

Table 3.
Q. 1.a 1.b 1.c 1.d 1.e 1.f 1.g 1.h

Factor Analysis of Items for MANAAMBI (Varimax Rotation).
Wording (Abbreviated) – R: Reversed Item Work-objectives are very clear and specific Understand which objectives are more important Work objectives are ambiguous and vague (R) Clear, planned goals, and objectives exist for job Certain about the amount of authority Know what is expected Know how to divide time over different tasks in job Know what responsibilities are Scale Factor 1 Factor 2 GC GC GC RA RA RA RA RA 0173 0.696 0.469 0.667 0.607 0.511 0.800 0.118 3.837 30.785 0.864 0.240 0.612 0.082 0.543 0.367 0.133 0.895 1.061 30.439

Eigenvalue Percentage of variance explained

MANAAMBI was measured using a scale that combined three items from Steers’ (1976) task-goal attributes questionnaire concerning goalclarity (cf. Kenis, 1979), and five items from the instrument for roleambiguity by Rizzo et al. (1970) and House (1971). Combining elements from these two existing instrument was expected to have greater validity than each of the two sets of items would have had individually, since the larger eight item scale addresses both goal- and role-related characteristics. Factor analysis revealed two factors explaining 61.2% of variance in total. These factors did not, however, correspond to the original sub-scales, as Table 3 displays. For this reason, and because of a high and maximal Cronbach’s a of 0.84 for the whole eight-item scale, the item scores were subsequently averaged to obtain a variable score for MANAAMBI. However, to allow assessing the robustness of the analyses based on this combined scales, two additional variables were constructed based on the original items. Goal-ambiguity (GOALAMBI) signifies the lack of goalclarity. It was measured by taking the (reversed) average scores on the three goal-clarity items. ROLEAMBI signifies role-ambiguity. It was measured by taking the average scores on the five ROLEAMBI items. Cronbach’s a was 0.71 and 0.73 for these scales respectively. Table 4 provides an overview of descriptive statistics of the variables measured.

170

FRANK G. H. HARTMANN

Table 4.
Variables

Overview of Descriptive Statistics of the Variables Measured.
Mean of variables 2.475 2.541 3.479 4.138 2.191 2.109 2.240 SD (n ¼ 196) 1.025 0.671 0.808 0.493 0.556 0.643 0.566 Mean Theoretical Range Actual Range a

Panel A: Descriptives EU TU APM SPM MANAAMBI GOALAMBI ROLEAMBI Variables

0.000–5.000 1.000–5.000 1.000–5.000 1.000–5.000 1.000–5.000 1.000–5.000 1.000–5.000 SD

0.009–4.167 1.889–4.889 1.000–5.000 1.800–5.000 1.000–4.125 1.000–4.000 1.000–4.200

0.766 0.869 0.721 0.726 0.836 0.705 0.728

Actual Range

Panel B: Descriptives of variables (n ¼ 196) Age 46.201 Tenure in company 17.891 Tenure in job 5.766

6.890 10.487 5.710

31–61 1–41 1–42

3.2. Organization, Response Rates, and Demographics The questionnaire was pre-tested with 5 faculty colleagues, 4 external reviewers in senior management positions, and all 11 coordinating officials in the participating organizations, after which it was field-tested with four potential respondents in two firms. Minor alterations were made in each step. Several measures from Dillman (2000) and Podsakoff, MacKenzie, Lee, and Podsakoff (2003) were used to optimize the response rate and reduce common method variance. These measures included the guarantee of strict anonymity, the use of high-quality printing with hand-written signatures on all correspondence, the use of pre-stamped envelopes and separate cards to request the study’s results, and the inclusion of a pen as token and for convenience. Overall, this may have contributed to the usable response rate of 78.4% (see also Table 1), which compares favorably to earlier and similar studies (cf. Dillman, 2000). A test for potential non-response bias was however conducted, comparing mean scores on variables for early and late responders (cf. Oppenheim, 1992; Brownell, 1995). This test was used the promise of anonymity made it impossible to obtain additional information about non-respondents for comparison with respondents (cf. Oppenheim, 1992; Brownell, 1995). The results of this analysis showed no evidence of systematic bias from non-response above chance. To test for common method variance, we performed Harman’s one factor test as advocated by Podsakoff and Organ (1986, p. 536). This procedure showed no

Do Accounting Performance Measures Indeed Reduce Managerial Ambiguity

171

Table 5.
Variables 1. EU 2. TU 3. APM 4. SPM 5. MANAAMBI 6. GOALAMBI 7. ROLEAMBI

Correlation Statistics (Pearson, Two-Tailed Significance, n ¼ 196).
1. 2. 3. 4. 5. 6.

À0.126 (0.079) 0.324 (0.000) 0.063 (0.379) À0.101 (0.158) À0.979 (0.172) À0.925 (0.197)

À0.169 (0.018) À0.070 (0.330) 0.275 (0.000) 0.260 (0.000) 0.275 (0.000)

0.183 (0.010) À0.361 (0.000) À0.342 (0.000) À0.334 (0.000)

À0.277 (0.000) À0.198 (0.005) À0.301 (0.000)

0.903 (0.000) 0.956 (0.000)

0.739 (0.000)

sign of any dominant single factor, as eight factors were extracted that explained between 6.53% (minimal) and 10.50% (maximal) of variance. Sample descriptives on managerial tenure are reported in Table 4 as well. The average respondent’s age was 46.2 years. Respondents had, on average, worked with their present employers for 18.1 years, and had been, on average, for 5.7 years in their present positions. The average number of employees in the respondents’ area of responsibility, which includes both the respondent’s department and sub-departments, was 79.0. On average, the respondents’ span of control, as measured by the number of employees under direct supervision, was 8.8. The results of the correlation analyses that were performed are reported in Table 5.

4. RESULTS
The hypotheses were tested using moderated regression analysis using the procedure recommended in Hartmann and Moers (1999). For each of the four hypotheses, regressions were run according to the following equation Y ¼ b0 þ b1 X 1 þ b2 X 2 þ b3 X 1 X 2 þ e (1) In Eq. (1) Y denotes MANAAMBI, X1 denotes the use of APM or the use of SPM, and X2 denotes EU or TU. Table 6 (panels A through D) reports on the outcomes of these tests for the four hypotheses.

172

FRANK G. H. HARTMANN

Table 6.
Variable

Effects of APM and SPM on MANAAMBI.
Coefficient Value SE t p

Panel A: Results of Hypothesis 1Ã (MANAAMBI is dependent variable) Constant APM EU APM Â EU b0 b1 b2 b3 À2.420 À0.057 0.275 À0.080 0.368 0.113 0.144 0.042 6.586 À0.506 1.905 À1.908 0.000 0.613 0.058 0.058 0.008 0.153 0.883 0.988 0.000 0.576 0.134 0.443 0.666 0.076 0.000 0.002

Panel B: Results of Hypothesis 2ÃÃ (MANAAMBI is dependent variable) Constant b0 3.584 0.899 3.985 À0.309 À0.216 À1.434 SPM b1 EU b2 À0.051 À0.342 À0.148 SPM Â EU b3 0.001 0.082 0.015 Panel C: Results of Hypothesis 3ÃÃÃ (MANAAMBI is dependent variable) Constant b0 2.050 0.629 3.260 À0.096 0.171 À0.560 APM b1 TU b2 0.364 0.242 1.503 APM Â TU b3 À0.052 0.067 À0.769 Panel D: Results of Hypothesis 4ÃÃÃÃ (MANAAMBI is dependent variable) Constant b0 À0.497 1.151 À0.432 0.487 0.273 1.785 SPM b1 TU b2 1.648 0.452 3.649 SPM Â TU b3 À0.334 0.107 À3.128
à Adjusted R2 ¼ 13.3%; F 3,192 ¼ 10.982; p ¼ 0.000; n ¼ 196. Ãà Adjusted R2 ¼ 7.0%; F 3,192 ¼ 5.858; p ¼ 0.001; n ¼ 196. ÃÃà Adjusted R2 ¼ 16.7%; F 3,192 ¼ 14.021; p ¼ 0.000; n ¼ 196. ÃÃÃÃAdjusted R2 ¼ 19.2%; F ¼ 16.452; p ¼ 0.000; n ¼ 196.
3,192

Regarding the EU hypotheses, the results suggest the existence of a negative moderating effect of uncertainty on the relationship between the use of APM and MANAAMBI, as was predicted in Hypothesis 1 (see Table 6, panel A). Hypothesis 1 indicates that the relationship between the use of APM and MANAAMBI is more negative for higher values of EU, which corresponds with the negative and significant effects found for the interaction term’s effect on MANAAMBI. In contrast, there is no positive interactive effect of the use of SPM and EU (Table 6, panel B), which contradicts Hypothesis 2. This suggests that an increasing use of SPM under high uncertainty does not result in more MANAAMBI. Regarding the TU hypotheses, the analyses show that Hypothesis 3 should be rejected. We do not find the expected positive interactive effect of the use of APM and TU on MANAAMBI (Table 6, panel C). We do find,

Do Accounting Performance Measures Indeed Reduce Managerial Ambiguity

173

however, support for Hypothesis 4 that predicted a negative effect of TU on the relationship between SPM and MANAAMBI (Table 6, panel D). This suggests that the use of SPM reduces MANAAMBI more under higher levels of TU. In addition to the straightforward analysis of the four hypotheses, two additional analyses were performed to assess the robustness of our findings. First, we reanalyzed Hypotheses 1 and 4 replacing the dependent variable MANAAMBI with two variables derived from the scores on the three items related to GOALAMBI and the five items related to ROLEAMBI. Results of these analyses are reported in Table 7 (panels A through D). These results show that the negative interactive effect of EU and APM is attributable to the enhanced clarity of goals, rather than to a reduction in role ambiguity (Table 7, panels A and B). In contrast, the negative interactive effect of TU Table 7.
Variable

Effects of APM and SPM on GOALAMBI and ROLEAMBI.
Coefficient Value SE t p

Panel A: Additional analysis of Hypothesis 1Ã (GOALAMBI is dependent variable) Constant b0 3.886 0.425 4.972 0.015 0.131 0.116 APM b1 EU b2 0.405 0.167 2.429 APM Â EU b3 À0.120 0.049 À2.463 Panel B: Additional analysis of Hypothesis 1ÃÃ (ROLEAMBI is dependent variable) Constant b0 2.604 0.380 6.852 APM b1 0.101 0.117 0.862 0.196 0.149 1.317 EU b2 APM Â EU b3 À0.056 0.043 À1.298

0.000 0.907 0.016 0.015 0.000 0.390 0.189 0.196

Panel C: Additional analysis of Hypothesis 4ÃÃÃ (GOALAMBI is dependent variable) Constant b0 0.584 1.395 0.419 0.676 SPM b1 0.200 0.331 0.604 0.547 1.086 0.548 1.983 0.049 TU b2 SPM Â TU b3 À0.196 0.129 À1.512 0.132 Panel D: Additional analysis of Hypothesis Constant b0 SPM b1 TU b2 SPM Â TU b3 4ÃÃÃÃ (ROLEAMBI is dependent variable) À1.146 1.154 À0.993 0.322 0.660 0.274 2.411 0.017 1.986 0.453 4.384 0.000 À0.417 0.107 À3.894 0.000

à Adjusted R2 ¼ 13.1%; F 3,192 ¼ 10.784; p ¼ 0.000; n ¼ 196. Ãà Adjusted R2 ¼ 10.6%; F 3,192 ¼ 8.676; p ¼ 0.000; n ¼ 196. ÃÃà Adjusted R2 ¼ 11.1%; F 3,192 ¼ 9.138; p ¼ 0.000; n ¼ 196. ÃÃÃÃAdjusted R2 ¼ 21.6%; F ¼ 18.920; p ¼ 0.000; n ¼ 196.
3,192

174

FRANK G. H. HARTMANN

and APM seems attributable to the reduction in role ambiguity, rather than an increase in the clarity of goals (Table 7, panels C and D). The substantive implications of these more detailed findings will be discussed in the final section of this paper. Results from rerunning these regressions with dummies for each of the organizations suggest that these findings cannot be attributed to systematic organizational differences. Finally, it was established that the reported results are not driven by obvious spurious factors, such as the age and experience of the managers. For this aim, all regressions were rerun including control variables that measured managers’ age, tenure in the company and tenure in their present jobs. This did not lead to substantial changes in the results, suggesting that the described effects of uncertainty are not driven by factors related to age and work experience.

5. CONCLUSIONS AND IMPLICATIONS
This paper attempted to explain the appropriateness of APM and SPM under uncertainty, which is an unresolved issue in the extant literature. For this purpose, a distinction was made between TU and EU in an attempt to reconcile findings in previous studies of positive and negative effects of using APM under uncertainty. The conclusions and implications of this study are as follows. For EU, the results support a positive effect of APM in terms of reducing MANAAMBI. As predicted, the use of APM reduces MANAAMBI more for higher levels of EU (H1). For TU, the analysis shows no such interactive effect, which suggests that the use of APM does not have the greater (negative) effect on MANAAMBI that was expected (H3). Instead, the use of SPM appeared to affect to have a more negative effect on MANAAMBI for higher levels of TU (H4). No effect was found of the interaction between EU and SPM (H2). Additional analyses revealed that the negative effects of APM and SPM under, respectively, EU and TU may regard different dependent variables as appeared when the original construct of MANAAMBI was decomposed into GOALAMBI and ROLEAMBI. Although the results of this further analysis should be interpreted with care because the factor structure of the original construct does not suggest their full independence (Table 3), it confirms that the interactive effects of the two types of performance metrics and the two types of uncertainty act along two rather different paths toward diminishing MANAAMBI. APM are especially a source of clear goals in, otherwise, turbulent external environments, as they

Do Accounting Performance Measures Indeed Reduce Managerial Ambiguity

175

inform the manager of the goals to be achieved. SPM, in contrast, reduce MANAAMBI by providing more immediate feedback on the actions that managers takes, and thus seems related to a control style that supports the manager in the actions that his role demands, rather than emphasizing the consequences of such actions which may be hard to predict. Note that these conclusions are based on interactions, rather than direct effects of uncertainty or the indicators. Some further understanding of the relevant issues may be gained, however, from observing direct (selection) effects of uncertainty on the use of performance indicators as demonstrated by the correlation statistics. Indeed, EU and TU appear to have opposite effects on the use of APM in the first place (Table 5). This provides some indication that superiors do adapt their use of APM to the context, in line with the predictions about the contextual appropriateness of APM, since the correlations between EU and TU and APM are significantly positive and negative, respectively. For SPM, no such effects are apparent. Regarding the direct effects of the use of performance measure use of MANAAMBI, both APM and SPM are able to reduce MANAAMBI. The results confirm the role of performance evaluation as a feedback device in the supervision of subordinate managers. Indeed, the intensity of such evaluations seems to be a good predictor for the direct effects of performance evaluation on subsequent MANAAMBI. Based on these findings, the following conclusions are drawn. Overall, the results seem to support the relevance of knowing the origin of uncertainty when assessing the appropriateness of APM and SPM. In particular, the results support the basic prediction that EU and TU do not affect the appropriateness of APM and SPM in the same way. When interpreting the direction of the effects, the findings suggest that the use of APM may be more appropriate under conditions of high EU. This finding is in line with the results reported by Simons (1987) and Ezzamel (1990), and appears to confirm the role of APM in structuring and ‘buffering’ the manager’s job by providing specific and clear goals (e.g., Hopwood, 1972; Hirst, 1983; Ross, 1995; Fisher et al., 2002). This finding is also in line with earlier results in budgetary control studies that addressed budgeting under uncertainty (e.g., Simons, 1987, 1995; Abernethy & Brownell, 1999), and extends these results to the larger body of APM literature. For TU, we find positive effects of increased use of SPM. This supports the general importance of subjective criteria (cf. Ross, 1995), but as the use of SPM does not seem to be directly affected by the level of uncertainty, this variable may reflect a personal trait of the superior, rather than contextually determined behavior. With these findings, some of the contradictions in the extant literature may be

176

FRANK G. H. HARTMANN

reconciled. Studies failing to find negative effects of APM under uncertainty may in fact have measured EU rather than uncertainty as such. For some studies this is clear from the scale they report to have used (e.g., Ross, 1995). Also Marginson and Ogden (2005) the source of initial ambiguity may have been environmental rather than task related. The study is not immune to traditional weaknesses associated with the survey method used, regarding internal and external validity. However, the care taken in the design of the empirical study, in obtaining the sample of respondents, in the design and pretests of the questionnaire, and in the follow-up procedures may have provided effective controls against many reliability and validity threats normally associated with survey research (cf. Young, 1996; Podsakoff et al., 2003), and may have realized the relative advantages of the survey method for studying a real-world phenomenon in a cross-sectional analysis of real-world subjects (cf. Lipe & Salterio, 2000) without publicly available data. In addition, the high response rate, the overall high reliability of the measurement instruments, and the size of the sample all compare favorably to previous accounting studies using a similar method. The fact that the sample has not been strictly randomly selected requires care when drawing inferences from this study’s results, although the sampling method was carefully designed to eliminate any obvious bias. Finally, the cross-sectional nature of the study requires care when interpreting the statistical associations as causal relationships. From the evaluation of results and implications of this study, and its strengths and weaknesses, the following directions for further research seem worth exploring. First, in this study, the use and usefulness of APM was examined in terms of subordinate’s attitudes and responses. An alternative level of analysis would be the superior, and a related question would be whether superiors’ evaluative behaviors can be explained in terms of contextual appropriateness of APM. A constructive replication of the findings in this study may explore the role of managers’ personality, in terms of their preference for uncertainty, or tolerance for ambiguity (e.g., McGhee, Shields, & Birnberg, 1978), which would complement the two types of uncertainty explored here. Second, future studies may attempt to integrate the findings from this and similar psychology-based studies with agency models, to enrich economic explanations of performance evaluation practices (cf. Prendergast, 2002; Ittner et al., 2003). This would enable this literature to recognize the wider or intermediate managerial and economic roles of performance evaluation, for example as a provider of clear feedback information (Luckett & Eggleton, 1991), and by recognizing the potential impact of different types of uncertainty (cf. Prendergast, 2002, p. S117; MacLeod,

Do Accounting Performance Measures Indeed Reduce Managerial Ambiguity

177

2003). Methodologically, this integration may imply complementing researching managerial perceptions about the use and quality of performance metrics, with archival and experimental evidence (cf. Merchant, Van Der Stede, & Zheng, 2003).

NOTE
1. All factor analyses are performed using PCA extraction, extraction factors with eigenvalues greater than 1. Interpretation of factors is done after Varimax rotation.

REFERENCES
Abernethy, M. A., & Brownell, P. (1999). The role of budgets in organizations facing strategic change: An exploratory study. Accounting, Organizations and Society, 24, 189–204. Alreck, P. L., & Settle, R. B. (1985). The survey research handbook. Homewood, IL: Irwin. Briers, M., & Hirst, M. K. (1990). The role of budgetary information in performance evaluation. Accounting, Organizations and Society, 15, 373–398. Brownell, P. (1985). Budgetary systems and the control of functionally differentiated organizational activities. Journal of Accounting Research, 23, 502–512. Brownell, P. (1995). Research methods in management accounting, Coopers and Lybrand, Accounting Research Methodology, Monograph No. 2, Melbourne, Australia. Brownell, P., & Hirst, M. K. (1986). Reliance on accounting information, budgetary participation, and task uncertainty: Tests of a three-way interaction. Journal of Accounting Research, 24, 241–249. Buchko, A. A. (1994). Conceptualization and measurement of environmental uncertainty: An assessment of the miles and snow perceived environmental uncertainty scale. Academy of Management Journal, 37, 410–425. Campbell, D. J. (1988). Task complexity: A review and analysis. Academy of Management Review, 13, 40–52. Chapman, C. S. (1997). Reflections on a contingent view of accounting. Accounting, Organizations and Society, 22, 189–205. Chun, Y. H., & Rainey, H. G. (2005). Goal ambiguity and organizational performance in U.S. federal agencies. Journal of Public Administration Research, 15, 529–557. Covaleski, M., Evens, J. H., Luft, J. L., & Shields, M. D. (2003). Budgeting research: Three theoretical perspectives and criteria for selective integration. Journal of Management Accounting Research, 15, 3–49. Dillman, D. A. (2000). Mail and internet survey: The tailored design method. New York, NY: Wiley. Downey, H. K., Hellriegel, D., & Slocum, J. W. (1975). Environmental uncertainty: The construct and its application. Administrative Science Quarterly, 20, 613–629. Early, P. C. (1985). Influence of information, choice, and task complexity upon goal acceptance, performance, and personal goals. Journal of Applied Psychology, 70, 481–491. Early, P. C., Prest, W., & Wojnaroski, P. (1987). Task planning and energy expended: Exploration of how goals influence performance. Journal of Applied Psychology, 72, 1–20. Emory, C. W., & Cooper, D. R. (1991). Business research methods. Homewood, IL: Irwin.

178

FRANK G. H. HARTMANN

Ewusi-Mensah, K. (1981). The external organizational environment and its impact on management information systems. Accounting, Organizations and Society, 6, 301–316. Ezzamel, M. (1990). The impact of environmental uncertainty, managerial autonomy and size on budget characteristics. Management Accounting Research, 1, 181–197. Fisher, J. G., Maines, L. A., Peffer, S. A., & Sprinkle, G. B. (2002). Using budgets for performance evaluation: Effects of resource allocation and horizontal information asymmetry on budget proposals, budget slack, and performance. The Accounting Review, 77, 847–856. Galbraith, J. R. (1977). Organization design. Reading, MA: Addison-Wesley. Gordon, L. A., & Narayanan, V. K. (1984). Management accounting systems, perceived environmental uncertainty and organization structure: An empirical investigation. Accounting, Organizations and Society, 9, 33–47. Govindarajan, V. (1984). Appropriateness of accounting data in performance evaluation: An empirical investigation of environmental uncertainty as an intervening variable. Accounting, Organizations and Society, 9, 125–135. Hartmann, F. G. H. (2000). The appropriateness of RAPM: Toward the further development of theory. Accounting, Organizations and Society, 25, 451–482. Hartmann, F. G. H., & Moers, F. (1999). Testing contingency hypothesis in budgetary research: An evaluation of the use of moderated regression analysis. Accounting, Organizations and Society, 24, 291–315. Hayes, D. C. (1977). The contingency theory of managerial accounting. The Accounting Review, 52, 22–39. Hirst, M. K. (1983). Reliance on accounting performance measures, task uncertainty and dysfunctional behavior: Some extensions. Journal of Accounting Research, 21, 596–605. Hirst, M. K. (1987). The effects of setting budget goals and task uncertainty on performance: A theoretical analysis. The Accounting Review, 62, 774–784. Hofstede, G. (1967). The game of budget control. Assen, The Netherlands: Van Gorcum. Hopwood, A. G. (1972). An empirical study of the role of accounting data in performance evaluation. Journal of Accounting Research, 10, 156–182. Horngren, C. T., Sundem, G. L., & Stratton, W. O. (1996). Introduction to management accounting. Upper Saddle River, NJ: Prentice-Hall. House, R. J. (1971). A path goal theory of leader effectiveness. Administrative Science Quarterly, 16, 321–339. Imoisili, O. A. (1989). The role of budget data in the evaluation of managerial performance. Accounting, Organizations and Society, 14, 325–335. Ittner, C. D., Larcker, D. F., & Meyer, M. W. (2003). Subjectivity and the weighing of performance measures: Evidence from a Balanced Scorecard. The Accounting Review, 78, 725–753. Kenis, I. (1979). Effects of budgetary goals characteristics on managerial attitudes and performance. The Accounting Review, 54, 707–721. Khandwalla, P. N. (1972). The effect of different types of competition on the use of management controls. Journal of Accounting Research, 10, 275–285. Lipe, M. G., & Salterio, S. E. (2000). The balanced scorecard: Judgmental effects of common and unique performance measures. The Accounting Review, 75, 283–298. Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task performance. Englewood-Cliffs, NJ: Prentice-Hall. Luckett, P. F., & Eggleton, I. R. (1991). Feedback and management accounting: A review of research into behavioural consequences. Accounting, Organizations and Society, 16, 371–394.

Do Accounting Performance Measures Indeed Reduce Managerial Ambiguity

179

MacLeod, W. B. (2003). Optimal contracting with subjective evaluation. The American Economic Review, 93, 216–241. McGhee, W., Shields, M. D., & Birnberg, J. G. (1978). The effects of personality on a subject’s information processing. The Accounting Review, 53, 681–697. Marginson, D., & Ogden, S. (2005). Coping with ambiguity through the budget: The positive effects of budgetary targets on managers’ budgeting behaviours. Accounting, Organizations and Society, 30(5), 435–456. McConkie, M. L. (1979). A clarification of the goal setting and appraisal processes in MBO. Academy of Management Journal, 4, 29–40. Merchant, K. A. (1984). Influences on departmental budgeting: An empirical examination of a contingency model. Accounting, Organizations and Society, 9, 291–307. Merchant, K. A. (1989). Rewarding results: Motivating profit center managers. Boston, MA: Harvard Business School Press. Merchant, K. A. (1990). The Effects of financial controls on data manipulation and management myopia. Accounting, Organizations and Society, 15, 297–313. Merchant, K. A., Van Der Stede, W. A., & Zheng, L. (2003). Disciplinary constraints on the advancement of knowledge: The case of organizational incentive systems. Accounting, Organizations and Society, 28, 251–286. Milliken, F. J. (1987). Three types of perceived uncertainty about the environment: State, effect, and response uncertainty. Academy of Management Review, 12, 133–143. Moers, F. (2005). Discretion and bias in performance evaluation: The impact of diversity and subjectivity. Accounting, Organizations and Society, 30(1), 67–80. Olson, O., & Rombach, B. (1996). The treasurer’s department as a buffer organization. Financial Accountability and Management, 12, 245–259. Oppenheim, A. N. (1992). Questionnaire design, interviewing and attitude measurement. London, UK: Pinter. Otley, D. T. (1978). Budget use and managerial performance. Journal of Accounting Research, 16, 122–149. Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903. Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational research: Problems and prospects. Journal of Management, 12, 531–544. Pedhazur, E. J., & Pedhazur-Schmelkin, L. (1991). Measurement, design and analysis: An integrated approach. Hillsdale, MI: Erlbaum. Prendergast, C. (2002). Uncertainty and incentives. Journal of Labor Economics, 20, S115–S137. Rizzo, J. R., House, R. J., & Lirtzmann, S. I. (1970). Role conflict and ambiguity in complex organizations. Administrative Science Quarterly, 15, 151–163. Rodgers, R., & Hunter, J. E. (1991). Impact of management by objectives on organizational productivity. Journal of Applied Psychology, 76, 322–336. Ross, A. (1995). Job related tension, budget emphasis and uncertainty: A research note. Management Accounting Research, 6, 1–11. Sawyer, J. E. (1992). Goal and process clarity: Specification of multiple constructs of role ambiguity and a structural equation model of their antecedents and consequences. Journal of Applied Psychology, 77, 130–142. Schuler, R. S., Beutell, N. J., & Youngblood, S. A. (1989). Effective personnel management. St. Paul, MN: West Publishing Company.

180

FRANK G. H. HARTMANN

Simons, R. (1987). Planning, control and uncertainty: A process view. In: W. J. Bruns & R. S. Kaplan (Eds), Accounting and management, field study perspectives (pp. 339–362). Boston, MA: Harvard Business School Press. Simons, R. (1995). Levers of control. Boston, MA: Harvard Business School Press. Steers, R. M. (1976). Factors affecting job attitudes in a goal-setting environment. Academy of Management Journal, 19, 6–16. Thompson, J. D. (1967). Organizations in action. Social science base of administrative theory. New York, NY: McGraw-Hill. Tosi, H. L. (1975). The human effects of managerial budgeting systems. In: J. L. Livingstone (Ed.), Managerial accounting: The behavioral foundations. Columbus, OH: Grid. Tymon, W. G., Stout, D. E., & Shaw, K. (1998). Critical analysis and recommendations regarding the role of perceived environmental uncertainty in behavioral accounting research. Behavioral Research in Accounting, 10, 23–46. Umapathy, S. (1987). Current budgeting practices in U.S. industry. New York, NY: Quorum. Vaigneur, K., & Peiperl, M. (2000). Reconsidering performance evaluative style. Accounting, Organizations and Society, 25, 511–525. Vancil, R. F. (1979). Decentralization: Managerial ambiguity by design. Homewood, IL: Dow Jones-Irwin. Withey, M., Daft, R. L., & Cooper, W. H. (1983). Measures of Perrow’s work unit technology: An empirical assessment and a new scale. Academy of Management Journal, 26, 45–63. Wood, R. E., Mento, A. J., & Locke, E. A. (1987). Task complexity as a moderator of goal effects: A meta-analysis. Journal of Applied Psychology, 72, 416–425. Young, S. M. (1996). Survey research in management accounting: A critical assessment. In: A. J. Richardson (Ed.), Research methods in accounting. Vancouver, Canada: CGA Research Foundation.

CAPACITY UTILIZATION AND THE BEFCU MODEL: A FIELD STUDY
Mohamed E. Bayou and Alan Reinstein
ABSTRACT
The few management accounting pricing methods in the management accounting literature are ineffective in helping small firms use their idle capacity during lingering economic recessions, and some of these methods may even worsen this problem. Extending the traditional break-even-cost-volume-profit model, we derive a more effective pricing method, the break-even-full-capacity-utilization (BEFCU) model, to handle this problem. Seeking full capacity utilization, the BEFCU model has two characteristics: (a) highlighting the importance of the exigent fixed cost category for utilizing idle capacity and (b) using a functional cost structure that focuses on a hierarchy of value drivers in the firm’s value creation process. Accordingly, under the BEFCU model, management has an instrumental pricing continuum extending from the minimum acceptable BEFCU sale price to the regular sale price. To demonstrate its practicality, the authors apply the BEFCU model to an actual job shop. This model integrates certain strategies based on builtin flexibility in commitments with suppliers and customers and maintaining a mode of conservatism in accounting for plant assets. The model can also help small tooling companies currently seeking entrance into China; it may take a while for these companies to gain a foothold in this new market because copyrights and other legalities are rarely enforced (Bunkley, 2004).
Advances in Management Accounting, Volume 16, 181–203 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16006-8

181

182

MOHAMED E. BAYOU AND ALAN REINSTEIN

In a competitive economy, idle capacity costs can significantly affect an organization’s performance (Klammer, 1996; Dodd, Lavelle, & Margolis, 2002). In economic recessions, however, utilization of idle capacity is often the key to survival. This is especially true for such small firms as job shops and medium-size firms since they often do not have enough resources to enable them endure lingering economic recessions and they often depend on a few large customers who themselves may suffer decreasing market shares and huge losses. During the current decade’s economic recession, many such firms went out of business (Mulligan, 2004). The key to their survival is effective capacity management (McNair, 1994; Klammer, 1996; Paranko, 1996). Unfortunately, traditional measures of capacity in management accounting literature do little to highlight idle capacity (Klammer, 1996, p. 28). McNair (1994) contends that capacity management is one of the most complex and troublesome areas in management practice. A review of many research studies in this literature reveals that capacity management is a twofold problem. First, no generally accepted definition of capacity and capacity management currently exists. Klammer (1996, p. 3) discusses several types of idle capacity. Second, traditional costing systems fail to measure the costs of unused capacity (Dodd et al., 2002; Buchheit, 2001; Klammer, 1996). Consequently, as Braucch and Taylor (1997) found, very few of their sampled firms seek to measure the extent or costs of unused productive capacity. Using an experimental methodology, Buchheit (2001) concludes that explicitly reporting idle capacity costs can lead to biased performance evaluation. This paper focuses on the pricing decision’s effect on utilizing idle capacity. Management accounting literature includes several pricing methods for capacity utilization. However, due to their rigid structures and lack of insight into the idle capacity’s twofold problem, these methods do not offer much help to solve small firms’ idle capacity problem. Many pricing methods in the management accounting literature are inadequate to help organizations utilize their idle capacity. They usually focus on either the short- or long-run decisions (Paranko, 1996; most managerial cost accounting textbooks). In the short-run idle capacity case, the literature is replete with discussions of the special-offer pricing method where the minimum acceptable special price covers the variable costs and any incremental fixed cost. In the long-run case, the sale price either covers the full cost of the product or the idle capacity is recommended for elimination (downsizing). But these two cases do not deal with situations when economic recessions are expected to last more than one year and the firm resists shrinking its capacity. In this case neither of the short-run nor the long-run case is applicable.

Capacity Utilization and the BEFCU Model: A Field Study

183

This paper focuses on small firms’ idle capacity utilization during an economic recession that lasts a few years, as many small firms have experienced since the start of this decade. After this introduction, the first section briefly reviews several pricing methods and shows their weaknesses in helping utilization of idle capacity. The second section shows how the pricing decision can help implement this goal. The third section explains how to modify a traditional cost-volume-profit (CVP) model, thus mitigating weaknesses of traditional pricing methods. The fourth section applies this CVP method to a job shop using information gathered by interviewing a job shop’s top manager. Finally, a summary and conclusions are presented.

PRICING METHODS TO HELP UTILIZE IDLE CAPACITY
The management accounting literature offers few techniques to help companies deal with idle capacity issues during economic slowdowns and recessions, explained as follows. Cost-Plus Pricing US firms often use cost-plus pricing, whose popularity stems mainly from management’s concern that a product’s sale price covers its cost base to generate a target markup (Saccomano, 1998; Shim & Sudit, 1995). Eq. 1 shows the general form of this method. Sale price ¼ Cost base þ Markup (1) Thus, if the cost base of manufacturing a product is $100 per unit and the target markup is 150% of the cost base, the sale price is $250 ($100+[$100 Â 150%]). The markup covers two elements: costs excluded from the cost base and a target profit. The cost-plus method has several basic weaknesses (Saccomano, 1998; Rutledge, 1996). If the cost base includes a fixed-cost element, then decreased demand and consequent decreased production result in increased cost base per unit. A vicious circle arises when increases in the cost base raise the sales price per unit, which, in turn, depresses the weakening demand and so on until the firm goes out of business, a dynamic sometimes called the ‘‘death spiral.’’ Some job shops are especially vulnerable to this spiral dilemma since much of their costs are fixed. In short, the cost-plus technique can augment – rather than reduce – the idle-capacity problem.

184

MOHAMED E. BAYOU AND ALAN REINSTEIN

Contribution-Margin Pricing Contribution margin, the excess of sales revenues over variable expenses as computed by Eq. 2, provides a different instrument to utilize idle capacity. Contribution margin per unit ¼ Sale price per unit À Variable cost per unit (2) The minimum acceptable sale price is determined when it produces a zero contribution margin (Eq. 3). Minimum acceptable sale price per unit ¼ Variable expenses per unit (3) With a short-run focus, this instrument appears appropriate to handle the idle-capacity problem. Idle capacity and cost behavior (i.e., for variable, fixed or mixed costs) normally exist only in the short-run because in the long run all costs tend to become variable and idle capacity is either reduced or eliminated by either downsizing or full utilization. Heavily automated, many job shops’ costs escape the contribution margin caption. For example, a sale price set near variable expenses per unit hardly contributes toward the coverage of any nonvariable cost, which offers an illusive solution to idle capacity problems during a chronic economic recession. Moreover, the job shop company we discuss later in this paper uses an activity-based costing (ABC) system that may prove ineffective since much of the fixed manufacturing costs are facility-sustaining costs, which the ABC system does not allocate to output units. ABC also does not include the cost of idle capacity in product costing (Garrison, Noreen, & Brewer, 2006).

Capital-Based Pricing Rutledge (1996, p. 50) asserts: ‘‘Don’t use cost-based pricing. Use capitalbased pricing.’’ This strong opposition stems from reasoning that rather than being merely an operating decision, bidding on projects requires a capital investment. Thus, he argues, ‘‘A low-margin project can be terrific if it uses very little capital. A high-margin project can be a loser if it requires a lot of capital. You can’t tell the difference until you do the work to estimate the balance sheet impact of the project and estimate the resulting returns.’’ He adds that a cost-plus price brings in ‘‘a mediocre return on capital while giving all of the value of any competitive advantage to the

Capacity Utilization and the BEFCU Model: A Field Study

185

customer without charge.’’ Eq. 4 shows the general form of capital-based pricing method. Sale price of a project ¼ f ðIncremental investment in the project; Incremental operating costs; Cost of capitalÞ ð4Þ Rutledge’s (1996) argument has two problems. First, contrary to his claim that the cost-plus price formula ignores the cost of investment, the markup in this formula typically includes a return on capital (Anthony & Govindarajan [A & G], 2004, p. 249). This markup component arises by multiplying the net investment (capital) of the firm by its required return on investment (ROI) to produce the total earnings required on the investment. Technically, this required return considers the firm’s capital structure. Dividing this amount of earnings by the number of units to be produced (or by such capacity input measures as total machine hours or total direct labor hours) determines the profit required per unit (or per hour) in the markup in the cost-plus price formula. Second, in pursuing accounting for capital structure, the capitalbased pricing method ignores the firm’s cost structure. Different firms may allocate their total investments differently among unit-level, batch-level, product-sustaining and facility-sustaining costs. Since the risk characteristics of these cost-structure elements differ, they may require different ROI rates. Revenue Management Revenue management links three functions: (1) demand forecasting capacity, (2) utilization and (3) adjusting prices to best meet and influence demand (Kroll, 1999, p. 26). Applied first in the late 1970s by airline companies, revenue management has gained wide acceptance in the service industry and has attracted strong attention in the manufacturing industry. Eq. 5 shows the general form of the revenue management method. Sale price ¼ f ðCustomer mix; Demand forecasting; SupplyÞ (5) Kroll (1999, p. 26) summarizes the core concepts of revenue management as follows: (1) the focus is on price rather than on costs to balance supply and demand, (2) market-based pricing should replace cost-plus pricing, (3) the method is more appropriate for sales to segmented micro-markets than to mass markets, (4) products should be saved for the most valuable customers, (5) decision making should be based on knowledge instead of supposition,

186

MOHAMED E. BAYOU AND ALAN REINSTEIN

(6) the firm should exploit each product’s value cycle, and (7) the firm should continually evaluate its opportunities. Revenue management is useful when the following conditions are met (Kroll, 1999, p. 28): (1) (2) (3) (4) (5) (6) short-term capacity is limited, marginal costs are low, demand varies over time, pricing is flexible, management can control supply allocation, and data on historical demand is obtainable.

Revenue management has several weaknesses. First, its benefits disappear if demand forecasts are uncertain. Second, charging different prices for what appears to be the same product can raise legal and customer problems. Third, revenue management is ‘‘not gouging’’ (Kroll, 1999) since aggressive revenue management can backfire and cause customers to seek alternative sources to satisfy their needs. Finally, by focusing on revenues and ignoring costs, the method’s mechanism becomes unable to adjust prices for different cost structures. In short, revenue management is inapplicable to such a small firm as a job shop that has very few large manufacturing customers and faces an extended period of weak or uncertain demand.

Analytical Modeling Analytical modeling considers capacity, pricing and costing concepts. For example, Balakrishnan and Sivaramakrishnan (2002) discuss how combining analytical and numerical methods can consider capacity-planning and product-pricing problems to clarify full costing decisions, suggesting that future research explore optimal time revisions and consider the costs associated with price and capacity revisions. Banker and Hansen (2002) develop a fullcost-based pricing heuristic model to simulate how a service firm determines each period’s amount of capacity, a price and a price discount. Their model finds that accuracy is more important for capacity than for pricing decisions. Banker, Hwang, and Mishra (2002) next develop a model to analyze product-costing and pricing decisions in a dynamic information environment under long-term capacity commitment, finding that the average expected optimal charge for an activity resource equals its expected full costs. Gox (2002) analyzes a capacity-planning and pricing problem of a monopolist facing uncertain demand. He finds that different types of capacity constraints

Capacity Utilization and the BEFCU Model: A Field Study

187

affect the firms’ obtained capacity, but not how it sets prices. However, none of these models fully accounts for cost hierarchies, which are critical for many different types of business. Segment-Margin Pricing Segment-margin pricing, as Eq. 6 shows, stresses more cost coverage than the contribution-margin method because the former covers fixed costs. Segment margin ¼ Sales revenues À ðVariable expenses þ Direct fixed expensesÞ Thus, Sales revenues ¼ Variable expenses þ Direct fixed expenses þ Segment margin ð7Þ Direct fixed expenses can be easily and economically traced to the segments. The minimum acceptable sale price per unit produces a zero segment margin in Eq. 7. But the segment-margin pricing method, as commonly presented in the management accounting literature, rarely differentiates between committed and discretionary fixed costs, as explained below. Also, when presented on a unit basis, it suffers from the death spiral dilemma as explained above. The following section explains the weaknesses of these pricing methods.

ð6Þ

LIMITATIONS OF THE TRADITIONAL PRICING METHODS TO MITIGATE THE IDLE CAPACITY PROBLEM
The six pricing approaches contain two limitations that impair their mitigating the idle capacity problem, explained as follows. Ignoring Urgent Costs for Utilization of Idle Capacity Fixed cost and idle capacity concepts are interrelated so that one can hardly exist without the other. The management accounting literature often classifies fixed cost into committed and discretionary fixed costs, presenting them as a dichotomy: current year’s fixed cost incurrence is either unavoidable (committed) or avoidable (discretionary). But many fixed costs are neither

188

MOHAMED E. BAYOU AND ALAN REINSTEIN

committed nor discretionary. A prudent cost manager would regard them as necessary or essential although they are not bound by contracts or by law. For example, maintenance costs, usually considered discretionary, contain parts that may be deemed necessary and wisely unpostponable. Similarly, some insurance coverage on factory facilities is not committed, but urgent – as are minimum advertising, necessary employee training and some software and hardware updating. For a lack of a better term, we label these portions of costs ‘‘exigent fixed costs.’’ Thus, during an economic recession, the minimum sale price should not be set below the sum of variable expenses and exigent fixed expenses.

Ignoring the Hierarchy of Value Drivers in the Value Creation Process Most firms’ manufacturing costs (direct material, direct labor, variable and fixed manufacturing overhead) are hierarchical, where some costs dominate or rule over others from their significance and functional role in the value creation process. For example, in such service firms as CPAs, lawyers or consultants, direct labor occupies a higher level of priority than raw materials, if any, and overhead in the cost hierarchy. In these businesses, direct labor determines the type, quantity and quality of raw materials and overhead in the production process. On the other hand, in a jewelry-manufacturer, direct materials (e.g., diamonds and gold) may have the highest priority in the cost hierarchy. In many high-tech manufacturing firms, factory overhead plays a dominant role since advanced robotic assets and high-technological systems often determine the type of labor skills and the form of prefabricated raw materials needed for the manufacturing process. Thus, in these firms factory overhead occupies the highest functional hierarchy in their manufacturing cost structure. The significance of the functional hierarchy of resources for idle capacity utilization purposes can be explained as follows: when the dominant resource becomes active in the value creating process, other resources become active and profitable. The dynamics of this functional hierarchy follow a ‘‘dynamo principle’’: when the dominant resource becomes idle, other related resources in the hierarchy become idle also. The key to reach full utilization of idle capacity is to keep the dominant resource active. A complete shutdown of a major resource also creates another problem: the cost of bringing back this resource to an operating level can be substantial. The following section presents a full-capacity utilization-pricing model that overcomes these limitations.

Capacity Utilization and the BEFCU Model: A Field Study

189

COST-VOLUME-PROFIT BREAK-EVEN MODELS
Many advances in management accounting have occurred since Jaedicke and Robichek (1964) developed a CVP model to incorporate uncertainty, including Jarrett (1973), Hilliard and Leitch (1975) and Yunker and Schofield (2005). Yunker and Yunker (2003) recently developed an economic demand function relating the expected sales level to price, making price the entity’s basic decision variable, and Yunker and Schofield (2005) apply this similar model to determine enrollment fees for training and development programs. Several extensions of the CVP model to incorporate production variables include Hayes and Wheelwright’s (1984) product life cycle as a key variable for products in later or ‘‘mature’’ stages of production. Hanna and Newman (1993) develop the cost volume flexibility break-even analysis (CVBA) to consider both economies of scale and economies of scope to compare among equipment alternatives. Their model considers that increased fixed costs could result in both lower variable costs and lower setup costs, suggesting that future research consider broader sets of parameters for manufacturing systems to adopt. Kortge (1984) argues that since break-even analysis forms a basis to compare various price alternatives (rather than a tool to make industrial price decisions), such analysis should include such variables as optimum sales forecasts (to help ascertain proper quantities) and the proper sequence of items through the production process. Our model considers and improves upon these approaches, explained as follows.

THE BREAK-EVEN FULL-CAPACITY-UTILIZATION (BEFCU) PRICING METHOD
The essential form of this BEFCU pricing method is developed as follows: Xf ¼ Ef ðPf À bÞ (8)

Where Xf, full capacity utilization in input units, e.g., machine hours, at the annual capacity level; Ef, total amount of exigent fixed cost at the annual capacity level; Pf, sale price per unit of input at the full-capacity utilization level; b, variable cost per unit of input.

190

MOHAMED E. BAYOU AND ALAN REINSTEIN

Eq. 8 is a modified version of the familiar break-even point formula. Solving for Pf yields:   Ef Pf ¼ b þ (9) Xf Simplifying Eq. 9 yields: Pf ¼ b þ ef (10) Where ef, exigent fixed cost per unit of input at the full capacity level. Eq. 10 shows that to fully utilize an asset’s capacity, the lowest acceptable price, Pf, in a bidding context during an economic recession is equal to the variable cost per unit plus the exigent fixed cost per unit.1 Pf represents an extreme point on a continuum whose other extreme is the regular sale price, P.
BEFCU Price Pf Regular Price P

This proposed method has three advantages. First, it explicitly accounts for the exigent costs, i.e., urgent and necessary costs that cannot be reduced to zero for establishing the minimum price for full-capacity utilization. Second, the modified CVP pricing method recognizes the hierarchy of value creation. Finally the BEFCU’s continuum points to the fuzzy nature of the pricing decision when some capacity is idle due to economic recessions, as demonstrated in the following application.

APPLICATION OF THE BREAK-EVEN FULL-CAPACITY UTILIZATION PRICING METHOD
The Custom Tool and Die (CTD) Company, a small job shop, has suffered significant losses due to their 60% idle capacity. Highly automated, much of their manufacturing costs are fixed. Management recently considered two options: (1) keeping several, specialized expensive machines completely idle for one year, i.e., until the current economic recession is over or (2) keeping these machines working with the hope the losses can somehow be sustained. Management chose not to dispose of any of these machines, expecting the recession not to last for several years. While some discretionary fixed costs may be decreased, management feels that finding an appropriate selling price is a key factor in selecting between Options One and Two. However,

Capacity Utilization and the BEFCU Model: A Field Study

191

the company is unsure of the appropriate price for deciding between the two options. With annual sales of approximately $40 million before the current economic recession, CTD operated through several plants in various U.S. locations; it produces customized tools and dies for automakers and a few other manufacturing companies. Contracts with these customers are generated mainly through price bidding practices. CTD needs assistance in utilizing its idle capacity and improving its pricing mechanism. During our recent tour of the company’s main plant and from lengthy interviews of its chief financial officer, we noted that high-tech machinery costs have the highest priority in their cost hierarchy. Their machining cost category alone is larger than direct-material and direct-labor categories combined. They customize most of their products to customer specifications using a high-tech job-order costing system, where manufacturing overhead plays a major role in the value creation process. Hence, when their machines become active to produce customers’ orders, their labor, materials and administrative resources also become active, and vice versa – when their machines are idle, nearly every other resource in the firm is idle (Smith, 2002, p. A1). The key then is to trigger the working of equipment in a job shop infected with idle capacity.

The Idle-Capacity Problem Since CTD decided not to dispose of its machines, the key question becomes: what pricing method should it use to decide whether to keep sophisticated machines idle or working? CTD bought several high-tech machines to serve a few large customers, entailing significant amounts of committed and discretionary annual fixed costs. Who should bear these costs? A and G (2004, p. 250) suggest that the company charge these fixed costs to these customers. But this is inapplicable to CTD. A & G’s suggestion applies to transactions between sellers and buyers in the same company. The CTD officials told us that if they charge these fixed costs to their key customers, the latter would switch to competition; external buyers do not have the same obligations to buy from an independent job shop as they feel toward internal sellers in their corporations. In short, finding an appropriate pricing method becomes essential in deciding between the two options outlined above. We helped management choose between options One and Two by classifying the cost-pricing techniques reviewed above in one group and setting our modified BEFCU pricing method as another group. Using several

192

MOHAMED E. BAYOU AND ALAN REINSTEIN

profitability measures and capacity utilization levels, we compare the two pricing groups’ effects on the two options in a matrix such as the following: Option 1: Keep Machines Idle Cost-based pricing methods Contribution margin; discretionary segment margin; total segment margin Contribution margin; discretionary segment margin; total segment margin Option 2: Keep Machines Working Contribution margin; discretionary segment margin; total segment margin Contribution margin; discretionary segment margin; total segment margin

Break-even-fullcapacity-utilization method

Estimates of the Key Capacity Costs The job shop uses high-tech manufacturing operations where several large machines provide the highest level in the firm’s hierarchical capacity. To simplify the presentation, we use only five machines (Machines 1–5) to represent this hierarchical level. Machine 3 is specialized and sophisticated with a regular sale price of $24 per hour of work on this machine. Table 1, Panel A shows that idle capacity is 60% (line (f)). Two major cost elements, manufacturing maintenance and rent, are allocated to these five machines as follows. Allocating Maintenance Costs Table 1 shows the maintenance cost allocation for the five machines. Variable maintenance costs of $1.80 per machine hour multiplied by the budgeted capacity on line (c) of Panel A provide the total variable maintenance cost of each machine in Panel B. Thus, variable maintenance costs of Machines 1–5 are $6,453, $4,500, $567, $2,880 and $1,800, respectively, a total of $16,200. The total discretionary maintenance fixed cost per year is normally $62,550. Dividing this amount by the full capacity of 22,500 h yields line (h) in Table 1. However, of the $62,550 amount, management deems only $27,000 is necessary, i.e., exigent, as explained above. Accordingly, the exigent fixed maintenance cost per hour is $1.20 ($27,000/ 22,500 h). Multiplying the maintenance exigent fixed cost of $1.20 per machine hour by the full capacity (line (a)) of Machines 1–5 (Table 1) yields

Capacity Utilization and the BEFCU Model: A Field Study

193

Table 1.

Allocating Maintenance Costs.
Machine 1 Machine 2 Machine 3 Machine 4 Machine 5

Total

Panel A: Long-term and short-term capacity levels Long-term capacity a 22,500 8,100 (machine hours) Weights b 100% 36% Master-budget c 9,000 3,585 (machine hours) Weights d 100% 40% Idle capacity: a–c e Idle capacity f percentage: e/a Panel B: Maintenance costs Variable cost ($1.80 g per hour)  c Discretionary fixed h costsà Exigent fixed costsà i 13,500 60% 4,515 56%

5,400 24% 2,500 28% 2,900 54%

3,150 14% 315 4% 2,835 90%

4,500 20% 1,600 18% 2,900 64%

1,350 6% 1,000 11% 350 26%

$16,200 $62,550 $27,000

$6,453 $22,518 $9,720

$4,500 $15,012 $6,480

$567 $8,757 $3,780

$2,880 $12,510 $5,400

$1,800 $3,753 $1,620

à The total amount of discretionary maintenance fixed cost is $62,550 per year, of which only

$27,000 is considered necessary, i.e., exigent and may not be postponed.

the allocated maintenance cost ($9,720, $6,480, $3,780, $5,400 and $1,620, respectively, a total of $27,000). Allocating Manufacturing Rent Cost Table 2 shows the allocation of manufacturing rent costs to the five machines. The $100,000 annual rent is split into two parts, 90% used by the manufacturing function and 10% used by the administrative and general functions. Allocating the $90,000 amount to the five machines depends on the size of the area each machine needs to operate. CTD’s actual practice follows. How can we determine this area? The four panels of Fig. 1 show how to ascertain this area for one of the machines, as explained at the bottom of Fig. 1. Since CTD’s Machines differ in shape, design, function and space needed for operations and operators’ safety, we allocate the machines’ factory floor space using the equipment’s center of gravity, which, in turn, forms the focal point to determine the needed space to be occupied. After determining the area of each machine, the factory rent of $90,000 is allocated to the five machines based on the areas they occupy (Table 2). The rent cost rate of $67 per sq. ft. multiplied by each machine’s area shows the allocated costs of $21,184, $13,561,

194

MOHAMED E. BAYOU AND ALAN REINSTEIN

Table 2.

Allocating Manufacturing Rent Cost.
90% 10% $100,000 $90,000 $10,000

Annual rent (based on ten-year lease) Percentage allocated to the manufacturing area Percentage allocated to the administrative and other areas

Total Machine boundary (sq. ft.) Rent cost (see above) Rent rate per sq. ft.a Allocated rent a Machine 1 314

Machine 2 201

Machine 3 452

Machine 4 254

Machine 5 113

1,334 $90,000 $67 $90,000

$21,184

$13,561

$30,495

$17,136

$7,624

Computed as follows: $90,000/1,334 sq. ft. ¼ $67.

$30,495, $17,136 and $7,624 to Machines 1–5, respectively. Given these and other cost elements, the following section produces a profitability report for CTD Company. A Segment Report of the First Hierarchical Capacity Level Table 3 shows a proforma income statement arranged by segments, where the five machines represent five segments. This report shows three profitability levels: contribution margin, discretionary segment margin and total segment margin. The income statement shows negative total segment margins (losses) for all five machines ($71,142, $37,713, $78,742, $47,646, $14,287 for Machines 1–5, respectively, a total of $249,530) (Table 3). This result is due mainly to the large idle capacity of 60% (Table 1). Using the BEFCU method, Table 4 shows a pro forma segment income statement for Machine 3 only, the focus of Options One and Two. Table 4 is based on the following decisions: (1) The company decides to spend only the exigent fixed cost for the following items: fixed maintenance, $27,000 (Table 1); insurance, $45,000 and updating, $31,500, a total of $103,500. Machine 3’s share of these costs are: maintenance, $3,780; insurance, $6,300 and updating, $4,410 (Table 4). (2) The BEFCU price, Pf, of $8.60 is determined as follows: Given the total variable cost per machine hour of $4.00, the total exigent fixed cost is $103,500 and Eq. 10, the BEFCU price per hour of work on

Capacity Utilization and the BEFCU Model: A Field Study
Panel (a): A Bird’s Eye View of a Machine’s Footprint

195

Panel (b): Drawing the Machine Boundary A Rotating Machine

Center of Gravity

. .
.
Panel (c): A Stationary Machine : A Rectangular
4 2

.

. .

Panel (d): A Stationary Machine : Odd Shape

Fig. 1. Allocation of Rent Cost Among Factory Machinery. Note: The marked square inside the machine represents the location of the machine’s center of gravity. Panel (a) is a bird’s eye view of the machine footprint. Panel (b) shows two circles around the machine that rotates around its center of gravity during operations. Determining the inner circle requires making the machine’s center of gravity the center of a circle whose radius, the length of the line connecting this center of gravity measures r, and the furthest point on the machine’s footprint. Panel (b) shows that r equals 4 ft. Adding 2 ft. to r, an additional space for safety purposes, gives a radius of 6 ft. for the outer circle in Panel (b). Hence, the area of the outer circle around the machine is 113.097 sq. ft. For a stationary machine, as in Panel (c), its circle area can be converted into a rectangular (Panel (c)), a square, a triangle or an odd shape (Panel (d)) depending on the active side(s) of the machine that requires more space for machine and labor activities.

Machine 3, Pf3, is computed as follows: Pf3 ¼ b þ ef3 $103; 500 22; 500 ¼ 4:00 þ 4:60 ¼ $8:60 ¼ $4:00 þ

ð11Þ

196

MOHAMED E. BAYOU AND ALAN REINSTEIN

Table 3.

Proforma Segmented Income Statement for the Year Ending 12/31/2004.
Total Machine 1 3,585 $16 $57,360 $6,453 7,170 717 $14,340 $43,020 Machine 2 2,500 $17 $42,500 $4,500 5,000 500 $10,000 $32,500 Machine 3 315 $24 $7,560 $567 630 63 $1,260 $6,300 Machine 4 1,600 $16 $25,600 $2,880 3,200 320 $6,400 $19,200 Machine 5 1,000 $15 $15,000 $1,800 2,000 200 $4,000 $11,000

Master budget hours Regular sale price per hour Sales Variable expenses Maintenance ($1.80/ h) (Table 1) Power ($2.0/h) Cleanup ($0.20/h) Total variable cost ($4.00 per hour) Contribution margin Discretionary fixed expensesa Maintenance (Table 1) Insurance Updating Total direct discretionary fixed expenses Discretionary segment margin Direct committed fixed expenses Depreciation Rent (Table 2) Total direct committed fixed expenses Total segment margin a 9,000

$148,020 $16,200 18,000 1,800 $36,000 $112,020

$62,550 90,000 58,500 $211,050

$22,518 32,400 21,060 $75,978

$15,012 21,600 14,040 $50,652

$8,757 12,600 8,190 $29,547

$12,510 18,000 11,700 $42,210

$3,753 5,400 3,510 $12,663

$99,030

$32,958

$18,152

$23,247

$23,010

$1,663

$60,500 90,000 $150,500

$17,000 21,184 $38,184

$6,000 13,561 $19,561

$25,000 30,495 $55,495

$7,500 17,136 $24,636

$5,000 7,624 $12,624

$249,530

$71,142

$37,713

$78,742

$47,646

$14,287

The CTD company decides to spend the full amount of discretionary fixed cost of each of these items.

Therefore, the pricing continuum for Machine 3 extends from a minimum of $8.60 to a maximum of $24 per hour.
BEFCU Price Pf3 = $8.60 Regular Price P = $24.00

Capacity Utilization and the BEFCU Model: A Field Study

197

Table 4. Proforma Segmented Income Statement for Machine 3 for the Year Ending 12/31/2004.
Machine 3 Full capacity in hours BEFCU sale price per hour Sales Variable expenses Maintenance ($1.80/h) (Table 1) Power ($2.0/h.) Cleanup ($20/h) Total variable cost ($4.00/h) Contribution margin Exigent fixed expensesa Maintenance (Table 1) Insurance Updating Total direct discretionary fixed expenses Exigent segment margin Direct committed fixed expenses Depreciation Rent (Table 2) Total direct committed fixed expenses Total segment margin a 3,150 $8.60 $27,090 $5,670 6,300 630 $12,600 $14,490 $3,780 6,300 4,410 $14,490 – $25,000 30,495 $55,495 $55,495

The allocation of each exigent fixed expense to the five machines is based on the exigent fixed cost multiplied by each machine’s full capacity hours of activity as shown in Line (b) in Table 1. Thus, maintenance’s exigent cost per hour is $1.20 ($27,000 divided by 22,500 h); insurance’s exigent cost per hour is $2.00 ($45,000 divided by 22,500 h); updating expenses’ exigent cost per hour is $1.40 ($31,500 divided by 22,500 h).

ANALYSIS OF RESULTS AND IMPLICATIONS
Tables 5 and 6 show the contribution margin, discretionary (or exigent) segment margin and total segment margin in total and per hour for each of Option One (keep Machine 3 idle) and Option Two (keep Machine 3 working). All pricing methods favor keeping Machine 3 working since the three profitability criteria are better under Option Two than under Option One. However, the BEFCU pricing method shows stronger values for the three criteria; hence, the method provides more confidence for management choice.

198

MOHAMED E. BAYOU AND ALAN REINSTEIN

Table 5. Implications of the Cost-Based Pricing Methods for Capacity Options One and Two of Machine 3 (Regular Sale Price Per Unit is $24.00).
Profitability Measures Panel A: Contribution margin Option One (keep Machine 3 idle) Option Two (keep Machine 3 working) Panel B: Discretionary segment margin Option One (keep Machine 3 idle) Option Two (keep Machine 3 working) Panel C: Total segment margin Option One (keep Machine 3 idle) Option Two (keep Machine 3 working) a b

Total

Per Hourf

$– 6,300a $29,547b 23,247c $85,042d 78,742e

$– 20.00 Undefined 74.44 Undefined 249.97

The contribution margin as shown in Table 3. This amount is Machine 3’s share of the discretionary fixed costs (Table 3). c From Table 3. d This amount is the sum of discretionary ($29,547) and committed ($55,495) fixed costs of Machine 3 (Table 3). e This amount is Machine 3’s contribution margin ($6,300) less its discretionary ($29,547) and committed ($55,495) fixed cost. f The denominator for Machine 3 under Option Two is 315 h (Table 1).

Table 6. implications of the Cost-Based Pricing Methods for Capacity Options One and Two of Machine 3 (The BEFUC Price Per Unit is $8.60).
Profitability Measures Panel A: Contribution margin Option One (keep Machine 3 idle) Option Two (keep Machine 3 working) Panel B: Discretionary segment margin Option One (keep Machine 3 idle) Option Two (keep Machine 3 working) Panel C: Total segment margin Option One (keep Machine 3 idle) Option Two (keep Machine 3 working) a b

Total

Per Hourf

0 $14,490a $14,490b 0c $69,985d 55,495e

0.00 4.60 Undefined 0.00 Undefined 17.62

The contribution margin as shown in Table 4. This amount is Machine 3’s share of the exigent fixed costs (Table 4). c From Table 4 d This amount is the sum of exigent and committed fixed costs (Table 4). e This amount is Machine 3’s contribution margin ($14,490) less its exigent and committed fixed cost (Table 4). f The denominator under Option Two is Machine 3’s full capacity of 3,150 h (Table 1).

Capacity Utilization and the BEFCU Model: A Field Study

199

While using an ABC system, CTD’s management found that the system solves little of its idle capacity problem. As Yang and Wu (1996, p. 34) explain, ABC assumes full (100%) capacity utilization when the system calculates unit cost of resources, and that ‘‘[c]apacity ABC is unrealistically optimistic in pricing application’’ because ‘‘Fixed and variable expense rates usually increase an operation approaches full capacity due to over-time, maintenance, or fatigue.’’ This argument further supports including exigent costs in the BEFCU model. Generally, ABC systems are inapplicable to solve idle capacity problems because as Cooper and Kaplan (1992, p. 12) explain, ‘‘activity-based cost systems are not models of how expenses or spending vary in the short-run.’’ However, idle capacity is often a short-run and rarely a medium-run problem because as Klammer (1996, p. 16) explains, in the long-run, idle capacity becomes an ‘‘idle not-marketable’’ capacity and is target for abandonment. When selecting Option Two, CTD may reduce its capacity and/or stress flexibility in its contracts with its suppliers and customers as follows: 1. Contingent commitments. Certain discretionary fixed costs such as maintenance, training and system updating costs may not be incurred or committed until customers sign, or are about to sign, order contracts. Costs are made contingent upon receiving customers’ orders. CTD can use this strategy if it makes discretionary fixed costs conditional on receiving bidding offers. 2. ‘‘Total package’’ commitment. CTD can benefit from such a total package deal as that compressor manufacturer Atlas Copco offers. Baker (2004, p. 24) explains that this offer is a flexible range of funding packages that can cover equipment acquisition, planned maintenance and the supply of replacement parts, which contain three major benefits:  to help facilitate financial planning, costs are guaranteed for the period of the contract, with no hidden extras;  prices include all services according to the manufacturer’s recommendations; and  depending on the nature of the contracts, the equipment need not appear as an asset on the job shop’s balance sheet. 3. Flexible commitments. Leasing instead of owning plant assets with the option to sublease these assets may allow a company to utilize its idle capacity to cover some of its committed fixed costs, which assumes an existing demand for the subleasable space. The increasing trends of leasing and hiring employees on part-time rather than full-time basis are consistent with flexible commitment strategies.

200

MOHAMED E. BAYOU AND ALAN REINSTEIN

4. Order now, pay later. A job shop may encourage its ‘‘illiquid’’ customers to place orders during the economic recession without paying for the orders until after the recession ends. The job shop can then discount the accounts receivable at a bank with or without recourse. Such retail stores as furniture companies have followed this strategy for years. 5. Switching emphasis on resources. Through price adjustments, a company may encourage customers to switch their orders from production on the less sophisticated equipment to the more sophisticated ones. For example, increasing the price per hour of Machine 1 and decreasing the price of Machine 3 may help increase utilization of Machine 3. 6. Full-cost recovery. When significant idle capacity risks are imminent, CTD may use the full-cost recovery approach for such specialized resources as Machine 3. Under this conservative approach, the firm recognizes no profit until it recovers fully its invested cost. It can later decrease the price on this asset when depreciation charges no longer arise.

LIMITATIONS OF THE BEFCU METHOD
The BEFCU method has several limitations. (1) The paper’s assumption that price adjustments can help increase capacity utilization may be untenable during chronic economic recessions. Some small firms have solved this problem by selling some of their products in China. Bunkley (2004, p. 1C) observes that China is ‘‘giving many small firms across Michigan a huge opportunity to expand. y The small engineering and development firm [Managed Programs LLC] for the tooling industry has made approximately $1.5 million in China since 2002.’’ He quickly notices the risks a small firm may face when doing business in China: ‘‘getting a foothold in China is not easy and requires something of a gamble y The Chinese government does little to protect intellectual property rights, and business is usually conducted more slowly. Costs can mount quickly as prospective deals drag on, and innovative designs can fall into the hands of greedy, unethical competitors.’’ (2) The line demarcating committed and discretionary fixed costs can be a fine one for some cost elements, e.g., software registration fees and insurance. (3) Large manufacturing customers may play favorites among competing job shops. A job shop may suffer from these maneuvers regardless of the pricing method used.

Capacity Utilization and the BEFCU Model: A Field Study

201

(4) During an economic recession, a small job shop may be at the mercy of large manufacturing customers who use the ‘‘buyers’ market’’ mentality to their advantage. The pricing methods used may become irrelevant in this case.

SUMMARY AND CONCLUSIONS
During economic recessions, many job shops have several characteristics that can lead to huge losses from idle capacity: using high-tech operations, relying on a few large manufacturing customers and raising most of their revenues through price-bidding practices. After reviewing and critiquing several pricing methods in the management accounting literature, we find these methods containing two limitations that impair their mitigating the idle capacity utilization problem. By classifying fixed costs into committed and discretionary categories, they ignore a cost element that can be important for idle capacity utilization decisions. Exigent fixed costs, e.g., the minimum maintenance, insurance and technology updating, are necessary and urgent, but neither committed nor totally discretionary. The pricing methods also ignore the cost structure and hierarchy of value drivers in the value creation process of the firm. We derive a new pricing technique – the ‘‘break-even-full-capacityutilization’’ (BEFCU) method, which does not contain these limitations, to help job shops to cope with idle capacity dilemmas. The BEFCU method has two characteristics: (a) highlighting the importance of the fixed exigent cost for idle-capacity utilization and (b) using a hierarchy of value drivers in the value creation process. These characteristics help to develop an instrumental pricing continuum whose end points are the minimum acceptable (BEFCU) sale price and the regular sale price. We applied the BEFCU method to an actual job shop needing help to improve its pricing mechanism and capacity utilization during an economic recession. Illustrating this model, we pose two options: Option One: keep factory machinery idle, and Option Two: keep the factory machinery active. Using hypothetical data to compare the performance of the traditional pricing method on the one hand and the BEFCU pricing method on the other, the paper suggests that the BEFCU pricing method works better, especially if the firm follows certain capacity strategies with built-in flexibility and is conservative in the face of the uncertainty of idle capacity problems. China provides substantial opportunities for small companies, which may find the BEFCU method a useful pricing tool

202

MOHAMED E. BAYOU AND ALAN REINSTEIN

since a small firm usually needs time to establish a strong foothold in this huge market. Further research can incorporate several iterations for the values of the BEFCU method into a simulation program, thus providing more realistic results for making informed decisions. To enrich Klammer’s (1996) CAM-I capacity model, we recommend adding the above explained hierarchy of value drivers, as a new dimension to the three kinds of idle capacity (marketable, not-marketable and idle off-limit),2 thus making the sources of capacity idleness more visible to management. For example, a utilization strategy for utilizing idle-marketable capacity depends on whether the primary value drive is direct materials, direct labor or manufacturing overhead.

NOTES
1. The management accounting literature usually presents variable cost per unit as a constant within the relevant range, and the total fixed cost is a constant subject to modification. For example, if the annual capacity utilization is 100,000 machine hours and total exigent fixed cost is $200,000, exigent fixed cost per hour, ef, is $2.00. If capacity utilization is expected to be 80,000 h (i.e., 80% of full capacity), exigent fixed cost per hour, e80%, would be $2.50 ($2.00/0.8). Similarly, at 40% capacity utilization, e40%, becomes $5.00 ($2.00/0.4) per hour. 2. Klammer (1996, p. 16) explains that idle capacity includes:  idle marketable: a market exists but capacity is idle because of competitor’s market share, product substitutes or distribution constraints;  idle not marketable: a market does not exist or management decides not to participate in the market; and  idle off-limit: capacity unavailable because of holidays, contracts or management’s policies or strategies.

REFERENCES
Anthony, R. N., & Govindarajan, V. (2004). Management control systems (11th ed.). Boston, MA: McGraw-Hill/Irwin. Baker, P. (2004). Unlock the cash from your compressors. Works Management, 57(1), 24–27. Balakrishnan, R., & Sivaramakrishnan, K. (2002). A critical overview of the use of full-cost data for planning and pricing. Journal of Management Accounting Research, 14, 3–31. Banker, R. D., & Hansen, S. C. (2002). The adequacy of full-cost based price heuristics. Journal of Management Accounting Research, 14, 33–57. Banker, R. D., Hwang, I., & Mishra, B. K. (2002). Product costing and pricing under long-term capacity commitment. Journal of Management Accounting Research, 14, 79–97.

Capacity Utilization and the BEFCU Model: A Field Study

203

Brauch, J. M., & Taylor, T. C. (1997). Who is accounting for the cost of capacity? Management Accounting, 79, 44–50. Buchheit, S. (2001). Outcome effects and capacity cost reporting. Managerial Finance, 27(5), 3–16. Bunkley, N. (2004). Mich. Firms look to China. The Detroit News, (November 28), 1C, 3C. Cooper, R., & Kaplan, R. S. (1992). Activity-based systems: Measuring the costs of resource usage. Accounting Horizons, 6(3), 1–13. Dodd, G. D., Lavelle, W. K., & Margolis, S. W. (2002). Driving improved profitability with activity-based costing. An Executive White Paper, (May). Garrison, R., Noreen, E., & Brewer, P. C. (2006). Managerial accounting (11th ed.). Chicago, IL: McGraw-Hill/Irwin. Gox, R. F. (2002). Capacity planning and pricing under uncertainty. Journal of Management Accounting Research, 14, 59–77. Hanna, M. D., & Newman, W. R. (1993). Academic traditional breakeven analysis to modern production economics: Simultaneously modeling economies of scale and scope. International Journal of Production Economics, 29, 187–201. Hayes, R. H., & Wheelwright, S. C. (1984). Restoring our competitive edge: Competing through manufacturing. New York: Wiley. Hilliard, J. E., & Leitch, R. A. (1975). CVP analysis under conditions of uncertainty: A log-normal approach. The Accounting Review, 50(1), 69–80. Jaedicke, R. K., & Robichek, A. A. (1964). Cost-volume-profit analysis under conditions of uncertainty. The Accounting Review, 39(4), 917–926. Jarrett, J. (1973). An approach to cost-volume-profit analysis under conditions of uncertainty. Decision Sciences, 4(3), 405–420. Klammer, T. (1996). Capacity measurement and improvement. Chicago, IL: Irwin/Professional Publishing. Kortge, G. D. (1984). Inverted breakeven analysis for profitable marketing decisions. Industrial Marketing Management, 13, 219–224. Kroll, K. H. (1999). A new tool for manufacturers. Industry Week, (May 3), 25–28. McNair, C. J. (1994). The hidden costs of capacity. Journal of Cost Management, 8(1), 12–25. Mulligan, A. (2004). Issues for small manufacturing enterprises, in New Directions in Manufacturing, National Academy of Sciences, pp. 46–48. http://books.nap.edu/ openbook.php?record_id=11024&page=46 Paranko, J. (1996). Cost of free capacity. International Journal of Production Economics, 46–47, 469–476. Rutledge, J. (1996). ‘‘Pricing for growth.’’ Forbes, (October 7), 158(8): 50 Saccomano, A. (1998). The price is right. Traffic World, (August 24), 50. Shim, E., & Sudit, E. F. (1995). How manufacturers price products. Management Accounting, 76, 37–39. ´ja Smith, R. A. (2002). Enron’s rise and fall gives some scholars a sense of de ` vu. Wall Street Journal, (February 4), A1, A6. Yang, G. Y., & Wu, R. (1996). Strategic costing and ABC. Management Accounting, 74(11), 33–37. Yunker, J. A., & Schofield, D. (2005). Pricing training and development programs using stochastic CVP analysis. Managerial and Decision Economics, 26(3), 191–208. Yunker, J. A., & Yunker, P. J. (2003). Stochastic CVP analysis as a gateway to decision-making under uncertainty. Journal of Accounting Education, 21, 339–365.

THE APPLICATION OF PERCEPTUAL BIAS TO NEGATIVE COMPENSATION SITUATIONS IN MANAGEMENT ACCOUNTING RESEARCH
Harry Z. Davis, Solomon Appel and John Y. Lee
ABSTRACT
In this article, we provide evidence that even when Murphy’s Law is objectively untrue, because of sampling bias, people perceive the law as true, and this perceptual bias has far-reaching implications in management accounting research. A corollary to Murphy’s Law is: ‘‘The other lane always moves faster than my lane.’’ A manager who is aware of this perceptual bias will try to structure her budget cutbacks and all other ‘‘negative compensations’’ in such a way that her employees perceive that the cutback applies to everyone, not just to themselves. The findings of our study support the wisdom that, whenever managers must implement managerial plans that will be perceived as ‘‘negative,’’ the plans should be implemented all at once. Spreading the implementation over a period of time produces more discontent on the part of the personnel affected. The findings lend credence to a generalization that peoples’ discontent is minimized when the number of observations

Advances in Management Accounting, Volume 16, 205–215 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16007-X

205

206

HARRY Z. DAVIS ET AL.

(and thus the number of chances for forming a negative perception) of undesirable events is minimized.

INTRODUCTION
The original formulation of Murphy’s Law in 1949 was: ‘‘If there is a wrong way to do something, then someone will do it.’’ The original formulation was a principle in designing machines to make them idiot-proof and thus help prevent accidents. If a machine is designed so that a connecting cable can be inserted either correctly or incorrectly, someone somewhere sometime will insert the cable incorrectly. The machine is thus poorly designed. In a well-designed machine, it is impossible to insert the cable incorrectly (Forward, 1983). Over time the law morphed into its classical formulation, which is a pessimistic Weltanschauung of a malevolent nature: ‘‘If something can go wrong, it will.’’ There are many corollaries to the classical formulation.1 Some of the corollaries have been validated in scholarly articles (Matthews, 1997c). ‘‘If a location can lie in an awkward part of the map, it will’’ has been explained using geometry and probability theory (Matthews, 1997b). ‘‘Toast falls off the kitchen table buttered side down’’ was validated by analyzing the gravitational torque of gravity and table heights (Matthews, 1995, 2001). ‘‘The notorious ubiquity of unpaired – ‘odd’ – socks’’ has been demonstrated using combinatorics (Matthews, 1996). ‘‘If a rope can become knotted, it will’’ has been validated using a recently discovered theorem in topology (Matthews, 1997a). One set of corollaries is of the form: ‘‘I always suffer the worst of all possible outcomes.’’ An example of this corollary law is: ‘‘The other lane always moves faster than my lane.’’ Regardless of the lane I switch into, my lane is always the slowest. In this paper we analyze this corollary in two ways: one, the objective reality, and two, the subjective perception of reality. We show that because of statistical sampling biases, people perceive that the law is valid, even when it has no basis in objective reality. Finally, we present the managerial implications.

THE MODEL
Assume a highway with two lanes, I, the lane I am in, and O, the ‘‘other lane,’’ both traveling in the same direction. Define Gj as the percentage of

Management Accounting Research

207

time that Lanej is ‘‘Going,’’ and define Sj as 1–Gj, i.e., the percentage of time that Lanej is ‘‘Stopped.’’ Further, assume GI ¼ GO, both lanes go (and stop) the same percentage of time. This allows us to drop the subscript, and just write G as the percentage of time the lanes are moving, and S as the percentage of time the lanes are stopped. Thus, by assumption, neither lane is slower or faster than the other. Further, assume that at any given point in time, there is no correlation between the two lanes – that is, at any given point in time, the conditional probability that I is stopped if O is stopped equals the conditional probability that I is stopped if O is going.2 At any given point in time there are four possible states of the world. One, both lanes are stopped, ISOS, which happens S2. Two, both lanes are going IGOG, which happens G2. Three, I is going and O is stopped, IGOS, which happens G Â S. Four, I is stopped and O is going, ISOG, which happens S Â G. Table 1 describes all states of the world from S ¼ 0% to S ¼ 100% in increments of 10%. In states ISOS and IGOG, the two lanes are equal. Thus, they provide no evidence for or against Murphy’s Law.3 For a driver in I, the state IGOS is a refutation of Murphy’s Law, since the other lane is the worse lane. ISOG is support for Murphy’s Law, since the other lane is the better lane. Define M, the Murphy Index (see Appendix A): M¼ ðI S OG À I G OS Þ ðI S OG þ I G OS Þ (1)

When ISOG ¼ IGOS, M ¼ 0%, there is no evidence for or against Murphy’s Law. When ISOG ¼ 100%, IGOS ¼ 0% so that M ¼ 100% and all the data support Murphy’s Law. When IGOS ¼ 100%, ISOG ¼ 0% so that M ¼ –100% and all the data refute Murphy’s Law.

Objective Reality For a driver randomly sampling the state of the two lanes, M ¼ 0% for all values 0%oSo100%. This follows because, by construction, at any level of S, ISOG ¼ IGOS. At the extreme values S ¼ 0% and S ¼ 100%, M is not defined, since there are no observations supporting or refuting Murphy’s Law – both lanes are always in the same state. Thus, objectively, for all values of S, there is no support or refutation for Murphy’s Law.

208

HARRY Z. DAVIS ET AL.

Table 1.
S Traffic Stoppage ISOS Both Lanes Stopped

Joint Traffic Probabilities (Real).
IGOG IGOS I (My Lane) Going ISOG I (My Lane) Stopped M Murphy’s Index

Both Lanes Going

O (Other Lane) O (Other Lane) Stopped Going S (%) S2 (%) (1–S)2 (%) S(1–S) (%) S(1–S) (%) (ISOG–IGOS)/ (ISOG+IGOS) (%) Undefined 0 0 0 0 0 0 0 0 0 Undefined

0 10 20 30 40 50 60 70 80 90 100

0 1 4 9 16 25 36 49 64 81 100

100 81 64 49 36 25 16 9 4 1 0

0 9 16 21 24 25 24 21 16 9 0

0 9 16 21 24 25 24 21 16 9 0

Happiness: Absolute or Relative In the standard economic model of consumer behavior, each consumer’s utility depends only on the quantity of goods the consumer possesses. By definition, one consumer’s utility is independent of the quantity of goods that any other consumer possesses. Each consumer calculates his or her utility function without comparing himself or herself to any other consumer. Psychologists and even casual observers of human behavior are familiar with concepts such as envy and jealousy. ‘‘Misery seeks company’’ implies that a person who is unhappy is comforted (and therefore less unhappy) in knowing that other people are also unhappy. If other people are also doing badly, knowing that relative to others they are not doing badly comforts them. If they observe that others are doing well, they conclude that not only are they themselves doing badly in absolute terms, but also they are doing badly in relative terms. The emotion of a person saying: ‘‘The other lane always moves faster than my lane’’ implies that the person’s happiness depends not only on the state they are in, but also on their state relative to the state of others. Furthermore, if there are only two states in the world, good and bad, the statement

Management Accounting Research

209

can be recast in the following form: ‘‘When my state is bad, the state of other people is good.’’ The statement says nothing about what happens when the state of the speaker is good. It is plausible that the underlying assumption is that when people are content, they tend not to compare themselves to other people. However, when they are discontent, they look at how other people are doing. We, therefore, allow for the possibility that a driver is more likely to sample how others are doing when the driver himself is stopped than when the driver himself is going. Subjective Perception First, consider a driver who only samples the state of the lanes when he is stopped, IS. Such a driver will only observe two states of the world: ISOS and ISOG. This effectively eliminates columns IGOS and IGOG from the sample, because sampling does not occur in those situations. Table 2 describes all states of the world from S ¼ 0% to S ¼ 100% in increments of 10%. For a driver randomly sampling the state of the two lanes, M ¼ 100% for all values 0%oSo100%. This follows because at any level of S, the driver sometimes observes ISOG, but the driver never observes IGOS. Thus, subjectively for all values of S, there is 100% support for Murphy’s Law.4 Table 2. Joint Traffic Probabilities (Biased).
S Traffic Stoppage ISOS Both Lanes Stopped ISOG I (My Lane) Stopped O (Other Lane) Going S (%) 0 10 20 30 40 50 60 70 80 90 100 S (%) 0 10 20 30 40 50 60 70 80 90 100 1–S (%) 0 90 80 70 60 50 40 30 20 10 0 ISOG/ISOG (%) Undefined 100 100 100 100 100 100 100 100 100 Undefined M Murphy’s Index

210

HARRY Z. DAVIS ET AL.

Biased Perception: Partially Objective, Partially Subjective Define b (0%pbp100%), the bias level of the driver, as the percentage that the driver samples from Table 2 and not from Table 1. A driver with bias level 0% samples only from Table 1; a driver with a bias level of 100% samples only from Table 2. A driver with bias level 0%obo100% samples b from Table 2, and 1–b from Table 1. For such a driver (see Appendix B): M¼ b ðb þ 2S À 2SbÞ (2)

Table 3 calculates M for b ¼ 25% to b ¼ 75% at 25% intervals. Table 4 is a summary of the three panels in Table 3, and M for b ¼ 0% and b ¼ 100%, which are calculated in Tables 1 and 2, respectively.

Results There are two interesting results in Table 4. One, as expected, when the sampling bias increases, the Murphy Index increases (@M/@b>0, see Appendix C). This result is intuitively obvious, because the greater the bias, the less observations of data which contradict Murphy’s Law. Two, as S increases, the Murphy Index decreases (@M/@So0, see Appendix D). This result is somewhat surprising. Since Murphy’s Law deals with a person in stopped traffic, why would the increase in the percentage of time the driver is stopped result in a lower M? The answer lies in the fact that M decreases only if the driver observes OS, conversely M increases only if the driver observes OG. As S increases, it becomes less likely that the driver will observe OG. Thus, paradoxically, as the objective reality of the world improves (S decreases), people with 0%obo100% become more convinced that Murphy’s Law is valid.

MANAGERIAL IMPLICATIONS
Assume a manager expects to make cutbacks in all her departments. The manager has the option of spreading the cutbacks over a year, or making all the cutbacks simultaneously. If the manager spreads the cutbacks over the whole year, each time a department is cut back, they will observe that other departments are not cut back (the equivalent of my lane stopped and the

Management Accounting Research

211

Table 3.
S Traffic Stoppage ISOS Both Lanes Stopped

Joint Traffic Probabilities (Partial Bias).
IGOG Both Lanes Going IGOS I (My Lane) Going ISOG I (My Lane) Stopped M Murphy’s Index

O (Other Lane) O (Other Lane) Stopped Going S (%) bS+(1–b)S2 (%) (1–b)(1–S)2 (%) (1–b)S(1–S) (%) b(1–S)+(1– b)S(1–S) (%) b/(b+2S– 2Sb) (%)

Panel A: Bias (b) ¼ 25% 10 3 20 8 30 14 40 22 50 31 60 42 70 54 80 68 90 83 Panel B: Bias (b) ¼ 50% 10 6 20 12 30 20 40 28 50 38 60 48 70 60 80 72 90 86 Panel C: Bias (b) ¼ 75% 10 8 20 16 30 25 40 34 50 44 60 54 70 65 80 76 90 88

61 48 37 27 19 12 7 3 1 41 32 25 18 13 8 5 2 1 20 16 12 9 6 4 2 1 0

7 12 16 18 19 18 16 12 7 5 8 11 12 13 12 11 8 5 2 4 5 6 6 6 5 4 2

29 32 33 33 31 28 23 17 9 50 48 46 42 38 32 26 18 10 70 64 58 51 44 36 28 19 10

63 45 36 29 25 22 19 17 16 83 71 63 56 50 45 42 38 36 94 88 83 79 75 71 68 65 63

212

HARRY Z. DAVIS ET AL.

Table 4. Summary of Murphy’s Index.
S Traffic Stoppage (%) 10 20 30 40 50 60 70 80 90 0% 0 0 0 0 0 0 0 0 0 25% 63 45 36 29 25 22 19 17 16 b (Sampling Bias) (%) 50% 83 71 63 56 50 45 42 38 36 75% 94 88 83 79 75 71 68 65 63 100% 100 100 100 100 100 100 100 100 100

other lane going). When a different department is cut back and this department is not cut back (the equivalent of my lane going and the other lane stopped), it makes much less of an impression. Thus, because of the perceptual bias, each department believes that it has been singled out for cutbacks more often than others. The manager is thus better off making all her cutbacks at once (the equivalent of my lane stopped and the other lane stopped) so that each one of her departments will not feel that the other departments always get fewer cutbacks. The managerial implications of our findings are that, whenever managers have ‘‘bad news’’ to relay or must implement managerial plans that will be perceived as ‘‘negative,’’ the news should be delivered or the plans implemented all at once. Spreading the delivery or implementation over a period of time produces more discontent on the part of the personnel affected. The findings lend credence to a generalization that peoples’ discontent is minimized when the number of observations (and thus the number of chances for forming a negative perception) of undesirable events is minimized. Our findings have wide-reaching managerial implications. Management must deal with expected discontent of organizational personnel whenever actions involving organizations’ limited resources affect peoples’ perception. Based on our results, researchers and practitioners can focus on the number of times people observe undesirable announcements and/or events rather than just on the amount of resources reduced. This can have far-reaching implications in various management related disciplines.

Management Accounting Research

213

CONCLUSION
In our simple model, we show that due to a sampling bias, people perceive that there is evidence supporting Murphy’s Law even when there is no such evidence in objective reality. The model applies to many situations. As long as people compare themselves to others more often when they are in a losing situation than when they are in a winning situation, even if there is no objective basis for Murphy’s Law, people will perceive that Murphy’s Law is valid. This would be true if two friends are taking an exam and one fails, he will then want to know how his friend did. If the friend also fails, he will be less upset. However, if his friend passes, he will perceive that Murphy’s Law is working against him. As people go through life, there will be many situations in which the biased perception will seem to provide evidence for Murphy’s Law. As they accumulate, the person will be more and more convinced that Murphy’s Law is valid. The implication for a manager is to try to schedule cutbacks across department simultaneously, so that each department does not feel that it is suffering the most. Based on our findings we conclude that people’s discontent is minimized when the number of observations of undesirable events is minimized, because it reduces the number of chances for forming a negative perception (bias). The results of our study provide a basis for future research involving the number of times people observe undesirable announcements and/or events versus the amount of resource reduction managers must deal with. The application of our findings is not limited to a single social-science discipline.

NOTES
1. An internet search yields over a quarter million sites. Two sites with a long list of corollaries are: Murphy’s Laws and Murphy Laws Site. 2. Relaxing this assumption by allowing for a correlation between the states of the two lanes changes the ratio of (ISOS+IGOG)/(IGOS+ISOG). Since the evidence for or against Murphy’s Law is the ratio IGOS/ISOG, the results of the paper remain unchanged. 3. Alternatively, a state in which the two lanes are equal could be considered evidence against Murphy’s Law. 4. At the extreme values S ¼ 0% and S ¼ 100%, M is not defined, since there are no observations supporting or refuting Murphy’s Law.

214

HARRY Z. DAVIS ET AL.

REFERENCES
Forward, R. L. (1983). Murphy lives. PPC-UCI News, March/April, 30 Matthews, R. A. J. (1995). Tumbling toast, Murphy’s Law and the fundamental constants. European Journal of Physics, 16, 172–176. Matthews, R. A. J. (1996). Odd socks: A combinatoric example of Murphy’s Law. Mathematics Today, March-April, 39–41. Matthews, R. A. J. (1997a). Knotted rope: A topological example of Murphy’s Law. Mathematics Today, 33, 82–84. Matthews, R. A. J. (1997b). Murphy’s Law of maps. Teaching Statistics, 19, 34–35. Matthews, R. A. J. (1997c). The science of Murphy’s Law. Scientific American, April, 72–75. Matthews, R. A. J. (2001). Testing Murphy’s Law: Urban myth as a source of school science projects. School Science Review, 83, 23–28.

APPENDIX A. THE MURPHY INDEX
The basic index is: ISOG/(ISOG+IGOS), since ISOG measures the data supporting Murphy’s Law, and IGOS measures the data contradicting Murphy’s Law. To refine the measure so that it is always between 100% and À100%, subtract a constant (50%) and multiply by a constant (2):   I S OG À 50% (A.1) M ¼2Â ðI S OG þ I G OS Þ Simple algebraic manipulation yields: M¼ ðI S OG À I G OS Þ ðI S OG þ I G OS Þ (A.2)

APPENDIX B. THE MURPHY INDEX WITH BIAS
Since I S OG ¼ bð1 À SÞ þ ð1 À bÞSð1 À SÞ and I G OS ¼ ð1 À bÞSð1 À SÞ M¼ ½bð1 À SÞ þ ð1 À bÞSð1 À SÞ À ð1 À bÞSð1 À Sފ ½bð1 À SÞ þ ð1 À bÞSð1 À SÞ þ ð1 À bÞSð1 À Sފ (B.2) (B.3) (B.1)

Management Accounting Research

215

M¼ M¼

b ðb þ 2ð1 À bÞSÞ b ðb þ 2S À 2SbÞ

(B.4) (B.5)

APPENDIX C. PARTIAL DERIVATIVE OF MURPHY INDEX WITH RESPECT TO BIAS
M¼ b ðb þ 2S À 2SbÞ (C.1) (C.2)

@M ½ðb þ 2S À 2SbÞ À bð1 À 2Sފ ¼ @b ðb þ 2S À 2SbÞ2 @M 2S ¼ @b ðb þ 2S À 2SbÞ2

(C.3)

The denominator is always positive, and for S>0 the numerator is also positive, so for S>0, @M 40 @b (C.4)

APPENDIX D. PARTIAL DERIVATIVE OF MURPHY INDEX WITH RESPECT TO TRAFFIC STOPPAGE
M¼ b ðb þ 2S À 2SbÞ (D.1) (D.2)

@M ½Àbð2 À 2bފ ¼ @S ðb þ 2S À 2SbÞ2 @M 2bðb À 1Þ ¼ @S ðb þ 2S À 2SbÞ2

(D.3)

The denominator is always positive, and for 0>b>1 the numerator is negative, so for 0>b>1, @M/@So0.

ACTIVITY-BASED COST MANAGEMENT AND MANUFACTURING, OPERATIONAL AND FINANCIAL PERFORMANCE: A STRUCTURAL EQUATION MODELING APPROACH
Adam S. Maiga and Fred A. Jacobs
ABSTRACT
This study uses structural equation modeling to investigate the impact of ABC implementation factors (management support, clarity and consensus of ABC objectives, non-accounting ownership, and training) on quality, cost, and cycle time improvements, the relations among quality, cost, and cycle time improvements and, the influence of quality, cost, and cycle time improvement on financial performance at the business unit level. Overall, the results of the structural analyses support the theoretical model indicating that ABC implementation factors influence quality, cost, and cycle time, and partial support for the relations among quality, cost, and cycle time improvement and their effect on financial performance. When these relationships are further analyzed within the context of ABC implementation stage, adoption of advanced manufacturing practices, industry characteristics and plant size to determine if these contextual

Advances in Management Accounting, Volume 16, 217–260 r 2007 Published by Elsevier Ltd. ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16008-1

217

218

ADAM S. MAIGA AND FRED A. JACOBS

factors impact the model constructs and the relationships between the variables in the theoretical model, the results show that these contextual factors do not affect the model constructs, however, they affect the model relations.

1. INTRODUCTION
The ability of a manufacturing company to compete effectively in the global market is determined to a large extent by the cost and quality of its products (Gunasekaran et al., 1994) and getting the product to market much more quickly (Milligan, 1999). To this end, more and more companies are implementing activity-based costing (ABC) (Kaplan & Norton, 1992, 1993, 1996; Kennedy & Affleck-Graves, 2001; Krumwiede, 1998; Shields, 1995; Shields & Young, 1994) in response to the new global competitive environment, and advocates argue that ABC provides cost data needed to make appropriate key decisions (Cooper & Kaplan, 1991). However, reservations have been expressed regarding the benefits of ABC (Innes et al., 2000; Malmi, 1997; Morrow & Connolly, 1994). In some cases firms have failed to complete their ABC projects and in others they have failed to gain benefits expected from the ABC systems they have installed (Lyne & Friedman, 1996). For example, Ittner, Lanen, and Larcker (2002) find modest evidence that ABC use is positively associated with manufacturing performance. However, on average, they find that extensive ABC use is positively associated with higher quality levels, greater decreases in cycle time, and larger increases in first pass quality. Additionally, their path analysis also indicates that ABC use has a positive indirect association with manufacturing cost reductions through improvements in quality and cycle time. Despite the evidence of association between ABC use and certain improvements in manufacturing processes, Ittner et al. (2002) find that, on average, extensive ABC use has no significant association with return on assets (ROA). Instead, they find some evidence that the relation between ABC and profits varies with the extent to which the decision to use ABC ‘‘matches’’ the plant’s operational characteristics. These findings, and particularly the strength of the relationships, may have been impacted by the choice of measurement metrics and method of analysis. Shields (1995) suggests that since ABC is embedded in a behavioral and organizational context that defines the programs and innovations that are implemented and succeed and fail, it is important that an ABC implementation strategy be focused on these behavioral and organizational

Activity-Based Cost Management

219

variables. However, very little empirical information exists about firms’ use of these behavioral and organizational variables and how they impact performance. Such information could be useful in helping to identify how companies benefit from ABC implementation. Additionally, contextual factors may be relevant in ABC implementation and success (Krumwiede, 1998). Hence, more empirical studies are needed as the population of ABC adopters increases to provide and test alternative empirical studies for ABC success. The first contribution of this study is to re-examine the findings of Ittner et al. (2002) and to further investigate whether there is an association between ABC use and manufacturing operational and financial performance using different measurement metrics and analyses methods in order to test the robustness of their findings. Thus, this study differs from Ittner et al. (2002) in several important ways. First, Ittner et al. (2002) measure ABC based on responses to a question asking whether ABC is extensively used in the plant (1 ¼ yes and 0 ¼ no), while this study focuses on ABC implementation factors, i.e., top management support, clarity and consensus of ABC objectives, non-accounting ownership, and training. Second, in addition to ROA used by Ittner et al. (2002) as the financial performance measure, this study also uses return on sales (ROS) as ABC adoption may affect both income and assets. Consequently, we use improvements in a plant’s ROA and ROS to evaluate the effect on financial performance. Third, Ittner et al. (2002) use ordinary least squares regression and average standardized responses of the performance variables to assess the association between the use of ABC and manufacturing plant performance, while this study uses structural equation modeling (SEM) (1) to estimate the impact of ABC implementation factors on process performance variables, (2) to examine the relations among the process performance variables, and (3) to assess the impact of process performance variables on financial performance. SEM incorporates all of the variables in the model to aid our understanding of their linkages. SEM allows the examination of the entire model simultaneously, rather than one relation at a time (Kline, 1998). SEM also avoids the potential for simultaneous equation bias, which arises when endogenous explanatory variables in the system of equations are correlated with disturbance terms with the result that classic ordinary least square estimators are not consistent (Gujarati, 1995). Recent research has examined the association between contextual factors and innovation diffusion process. For example, research has suggested that the benefits of ABC are more readily realized under contextual factors (Krumwiede, 1998; Cagwin & Bouwman, 2002). Hence, the second

220

ADAM S. MAIGA AND FRED A. JACOBS

contribution of this study is to investigate the model constructs and model relationships within the context of ABC implementation stages, adoption of manufacturing practices, industry, and size. To our knowledge, no prior study has assessed these relationships within the context of these variables. Overall, the results indicate support for the theoretical framework posited. ABC implementation factors significantly influence both quality and cost. However, their impact on cycle time improvement is rather tenuous. Quality improvement has a negative effect on both cost improvement and cycle time improvement which is found to positively impact cost improvement. Both quality improvement and cost improvement have significant positive effects on financial performance, while the impact of cycle time improvement on financial performance is not significant. Further analysis shows that the direct relations between ABC implementation factors and business unit financial performance are not significant, indicating that manufacturing performance measures (i.e., quality, cost, and cycle time) are intervening variables that mediate the relationship between ABC implementation factors and financial performance. Next, tests of contextual effects of ABC implementation stage, adoption of manufacturing practices, industry characteristics, and plant size on the construct levels and relationships were carried out. Results show that although these contextual factors do not affect the model constructs, they affect the model relationships. Therefore, this study adds to the literature by indicating contexts under which ABC implementation may be beneficial. The paper is organized as follows. First, the literature review is discussed. Next, a discussion of the research methods is conducted. After the empirical results are reported, a summary and discussion are presented.

2. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT
Anderson and Young (1999) argue that, ‘‘while there is broad agreement that ABC implementation factors are associated with successful outcomes, a difficulty exists in developing hypotheses because existing theories do not relate specific ABC implementation factors to particular aspects of success, and empirical studies vary in terms of effectiveness constructs, duration of implementation and units of analysis.’’ Furthermore, Shields and Young (1989) argue that, ‘‘consistent with other administrative innovations, the successful implementation of a cost management system does not depend on

Activity-Based Cost Management

221

technical resources, such as whether or what type of software is used, whether the system is integrated with other accounting systems or stands alone, or whether external consultants are used.’’ In this study, four key multidimensional constructs of implementation are derived from empirical analysis, based on existing dimensions considered in the literature to be associated with ABC success. The ABC implementation factors concern top management support, training, non-accounting ownership, and clarity of objectives. Prior research supports the potential role of these four implementation factors (e.g., Krumwiede, 1998) in ensuring that ABC will be useful for quality, cost, and cycle time improvement. Management decisions in the areas of product quality improvement, cost management, and cycle time improvement tend to be important strategically as they specify organizational direction and involve significant reengineering and cost reduction programs (Chenhall, 2004). Fig. 1 depicts the model under study. Top management must send clear signals to various parts of the organization about the importance of a project (McGowan & Madey, 1998). Transformational leadership theory suggests that senior management can encourage the pursuit of change by formulating and communicating a vision for the future and reinforcing values that support the vision (Tichy & Devanna, 1986). This suggests that top management support will help focus efforts toward the realization of organizational benefits and lend credibility to functional managers responsible for a project implementation and use (Doll, 1985). This support is necessary to change the culture of the organization if a work environment conducive to employee involvement is to be created (Burack et al., 1994; Daft, 1998; Hamlin, Reidy, & Stewart, 1997; Snell & Dean, 1992). Also, employees are more apt to work harder and contribute ideas to change process if management establishes a culture that supports employees (Burack et al., 1994). According to Shields (1995), support from top management provides the vehicle through which resources are controlled, goals are set and monitored, and political forces are generated to support the innovation. In addition top management can institute ABC performance-based incentive and procedures for employees. Compensation and rewards that fit with the improvement strategy will encourage employees to work toward the organization’s goals (Bonito, 1990). Empirical studies indicate that top management support can encourage practices and behaviors that lead to superior quality performance (Anderson, Rungtusanathan, Schroeder, & Dearaj,1995; Flynn, Schroeder, & Sakakibara, 1995; Saraph, Benson, & Schroeder, 1989), cost management (Chenhall, 2004), and cycle time improvement (Simers & Priest, 1989; Schilling & Hill, 1998).

222

Management Support Quality Improvement Training

Cycle Time Improvement

Financial Performance Improvement

ADAM S. MAIGA AND FRED A. JACOBS

Nonaccounting Ownership Cost Improvement Clarity of Objectives

Fig. 1.

Theoretical Framework.

Activity-Based Cost Management

223

This study suggests that with management support and visible signaling of their commitment to quality, cost, and cycle time improvement, the ABC implementation is likely to be successful. Hence, H1. ABC implementation factor of top management support has a significant impact on (1) quality improvement, (2) cost improvement, and (3) cycle time improvement. Goal theory suggests that acceptance is enhanced and individuals will expend effort in trying to make systems work, if they are provided with the specific goals of the initiatives (Locke, Shaw, Saari, & Latham, 1981). This suggests that before embarking upon an ABC initiative, it is imperative that organizations develop clear objectives of the system adoption. Clarity of objectives may enhance understanding of, and focus on, the purposes of ABC, and is likely to show how ABC aims to link operations to strategy, thereby enhancing the organizational validity of the systems (Chenhall, 2004). Contained therein should be the expected benefits that will result from the ABC adoption and steps for achieving these benefits (Jeffery & Morrison, 2000). Management should also formally communicate the implementation objectives to employees and help everyone understand their contribution to the process as well as implications their decisions have on the value of the organization (Bradford & Roberts, 2001). This implies that companies that embark upon an ABC adoption with clear and concise expectations of what the package will do for them will arguably realize greater organizational performance. Hence, by providing clarity of objectives that lends to understanding the process, ABC systems are said to encourage employers and manager toward more innovative problem solving techniques leading to changes in the cost, quality of production, and waste (McGowan, 1994). In summary, we argue that incorporating clarity of objectives into the ABC implementation process provides opportunities and focus for endusers to agree on organizational direction and technical characteristics of the organization. Therefore, the usefulness of ABC for quality, cost, and cycle time improvement will be enhanced if it is clear how ABC can improve these types of strategic decisions. Therefore, H2. ABC implementation factor of clarity of objectives has a significant impact on (1) quality improvement, (2) cost improvement, and (3) cycle time improvement. Training refers to the process of teaching job-related skills and knowledge to the employees in an organization and emerges as a crucial element of

224

ADAM S. MAIGA AND FRED A. JACOBS

workforce management (Mintzberg, 1979). Training, in addition to providing a stepping stone toward the efficient use of the information provided by the system, is vital in promoting acclamation to the system (Tait & Vessey, 1988). Training in designing, implementing, and using ABC is an important way to interrelate ABC among strategy, performance evaluation, and ABC objectives (McGowan, 1994). Choi (1995) suggests that to enable continuous improvement in an organization, workforce training in techniques necessary for improving process should be an ongoing activity. An ongoing application training based on the results of the ABC analysis is necessary for employees to use the ABC in changing operations and managing the business, using such tools as performance measurement (Shields, 1995; Sharman, 1993; Chenhall, 2004). Training also allows employees to more effectively undertake the goals that have been established, facilitates a better understanding of the detailed information produced by these ABC systems, allows employees to track activities involved in each process, and more clearly identify sources of waste (Choi, 1995). The influences of training on ABC implementation success are captured in the following hypothesis: H3. ABC implementation factor of training has a significant impact on (1) quality improvement, (2) cost improvement, and (3) cycle time improvement. Designating a cross-functional team, developing and enforcing guidelines regarding access to ABC data and authority to update ABC systems helps avoid confusion, misuse of ABC, and turf problems (Tatikonda, 2003). It is also plausible that ABC would be accepted and more readily promoted if there is non-accounting ownership of the systems (Cooper, Kaplan, Maisel, Morrissey, & Oehm, 1992). Such ownership derives from the centrality of ABC to the individuals’ jobs and generates a proclivity to champion the cause of ABC (Anderson, 1995), and to demonstrate commitment to the ABC system in their decisions and interactions with others in the organization (Chenhall, 2004). Therefore, this study suggests that non-accounting ownership may be associated with improved information about activities and cost drivers that is expected to enhance improvement in quality, cost, and cycle time by identifying non-value added activities, waste, and activities caused by poor quality and the drivers of these problems. Hence, H4. ABC implementation factor of non-accounting ownership has a significant impact on (1) quality improvement, (2) cost improvement, and (3) cycle time improvement.

Activity-Based Cost Management

225

Existing literature disagrees as to the compatibility of product quality and cost position. One view predicts inherent trade-offs between quality and cost position; another suggests that no inherent trade-offs exist and that the pursuit of high quality strategy may actually have synergistic effects on cost position. For example, both Juran (1988) and Crosby (1979) have consistently argued that better quality practices can reduce cost. The more obvious cost reductions from higher quality are achieved through increased output of defect-free products and lower expenditure on scrap, rework, inspection, and warranty and repair (Ittner, 1994; Kaynak, 2003). Perhaps even more significant are the less obvious indirect cost-reduction effects, such as fewer disruptions in operations due to out-of-conformance purchases and production, elimination of buffer inventories held to compensate for poor quality, improved machine utilization, and reductions in qualityrelated schedule changes, congestion, and downtime (Ittner, 1994). However, it has also been widely believed that higher quality entails the adoption of more expensive production technology including machinery, labor, and materials. Furthermore, achieving high quality position may require higher expenditures in other areas beyond the direct costs of manufacturing or distribution. Therefore, higher quality is viewed as being incompatible with lower per-unit manufacturing costs. Under this view, it would be difficult, if not impossible, for a plant to simultaneously pursue high quality levels for its products and low manufacturing costs. Based on the above competing arguments, we develop the following hypothesis in the null form: H5. Quality improvement does not have a significant impact on cost improvement. The relationship between quality and cycle-time has been viewed from two different perspectives. One view is that cycle time must be traded off against improvements in quality. That is, quality-improvement processes can be implemented only at the expense of longer cycle-times. However, an alternate view is that quality improvement and faster cycle-time can be simultaneously attained by reducing defects and rework (Crosby, 1979; Deming, 1986; Nandakumar, Datar, & Akella, 1993). Also, the empirical research on the impact of quality on cycle time is inconclusive. For example, in a study of information technology firms, Harter et al. (2000) find that higher quality products exhibit significant reductions in cycle-time; while Ittner et al. (2002) find that quality improvement significantly and positively impacts cycle-time.

226

ADAM S. MAIGA AND FRED A. JACOBS

Given the above conflicting relationship between quality and cycle-time, we develop the following hypothesis in null form: H6. Quality improvement does not have a significant impact on cycle time improvement. Achieving cost efficiency is increasingly critical in globally competitive markets (Young & Selto, 1993). Decreasing cycle time means eventually decreasing non-value added time and inventory, thereby lowering product cost (e.g., reduced overhead) (Ittner et al., 2002). Assuming that even with reduced capacity product volume remains constant, lowering overall operating costs by reducing cycle time would reduce unit product costs (Campell, 1995). We contend that that cycle improvement will exhibit a positive relationship with cost improvement. This leads to the following hypothesis: H7. Cycle time improvement has a significant impact on cost improvement. One of the challenges associated with making strategic decisions about quality is that its conceptualization varies by discipline. In marketing, quality tends to mean quality as perceived by the customer (e.g., Bolton & Drew, 1991; Parasuraman et al., 1985). In operations and quality management, quality tends to mean the efficiency and reliability of internal processes (e.g., Crosby, 1979; Deming, 1986), even if those processes are invisible to the customer (Ramaswamy, 1996). Therefore, depending on how quality is defined, different kinds of quality improvement efforts are likely to be appropriate, and most important, they are likely to have different pathways to profitability (Rust et al., 2002). Our conceptualization is based on the latter viewpoint. However, the direct impact of quality on financial performance has been inconclusive. For example, while Tatikonda and Montoya-Weiss (2001) found that technical product quality defined from operational perspective does, in fact, translate into financial performance, other studies in operations management (Dale & Lightburn, 1992; Madu & Kuei, 1995; Voss, Blackmon, Hanson, & Oak, 1995) that examined the impact of quality performance on overall business performance and have reported mixed results (Gale, 1994; Powell, 1995). Consequently, we develop and test the following hypothesis in null form: H8. Quality improvement does not have a significant impact on financial performance.

Activity-Based Cost Management

227

Prior literature strongly endorses the view that improved manufacturing performance will translate into higher profits (Garvin, 1988; Hayes et al., 1988). Roth and Borthick (1989), e.g., support the view that manufacturing performance, i.e., product cost, is an important key to improved business performance. Low cost is linked to competitive strategy because having a low-cost position yields the firm above-average returns relative to competitors by achieving lower relative direct costs (Phillips et al., 1983; Porter, 1980). Cost reduction programs transfer their savings to the bottom line directly (Rust et al., 2002). Gatignon and Xuereb (1997) suggest that the lower the cost, the greater the potential for profits, either by setting higher margin or by penetrating the market with lower price. This is in support of prior studies that suggest that businesses which primarily compete with the low-cost approach tend to achieve high market share through the offering of low prices, made possible by scale economies (Hambrick, 1983; Henderson & Henderson, 1979; Porter, 1980, 1985). Tatikonda and Montoya-Weiss (2001) found that cost does not have a significant effect on customer satisfaction, but does have a significant direct effect on relative sales. Therefore, in this study, we investigate the direct link between cost improvement and plant financial performance by developing and testing the following hypothesis. H9. Cost improvement has a significant impact on financial performance. Manufacturers are under pressure to produce and get the product to market much more quickly (Milligan, 1999). Thus, cycle time is increasingly becoming a critical variable in many business decisions (Stalk, 1988; Stalk & Hout, 1990; Meyer, 1993). This confirms the importance of cycle time in helping firms to not only compete effectively but also attain a competitive advantage because the focus on cycle time translates into bottom-line profits (Sharland, Eltantawy, & Giunipero, 2003). However, Stalk and Hout (1990) found that in industries where most companies follow a strategy of rapid product development, and rapid production and/or delivery, firms may incur the costs of accelerated cycle-time without the corresponding financial benefits. Research studies in the literature dealing with the effect of manufacturing cycle time on financial performance are inconclusive. Thus, there is a need to further investigate this relationship. Therefore, the following hypothesis is developed and tested in null form: H10. Cycle time does not have a significant impact on financial performance.

228

ADAM S. MAIGA AND FRED A. JACOBS

3. RESEARCH DESIGN
3.1. Sample To address the hypotheses, we used a survey to collect data from a crosssection of U.S. manufacturing plants that have adopted ABC. Our primary source is the Industry Week series on manufacturing excellence. Additional sources include The Wall Street Journal, Journal of Cost Management, Management Accounting, Harvard Business Review, various industrial engineering journals, and periodical indices for articles in any journal that might produce a case report or other information to determine if ABC is adopted. A total of 2,317 manufacturing units1 were randomly selected from the above sources and the names of managers were gathered using Dun & Bradstreet. The process resulted in 611 responses of which 5972 are usable or 25.77% is the usable response rate.3

3.2. Activity-Based Costing Implementation Factors ABC factors found significant in past studies include top management support, linkage to competitive strategies, adequacy of resources, non-accounting ownership, linkage to performance evaluation and compensation, implementation training, clarity of objectives, and number of purposes for ABC (Foster & Swenson, 1997; McGowan & Klammer, 1997; Shields, 1995; Krumwiede, 1998). However, none of these studies used SEM to link ABC implementation factors to performance outcomes. The composite ABC factors (see appendix) are measured using questions adapted from Krumwiede (1998).4

3.3. Plant Performance Measures We use four measures to assess plant performance based on Ittner et al. (2002) and Gatignon and Xuereb (1997) and relate to changes in quality, costs, cycle time, and financial performance over the last 5 years. The first variable, quality improvement, is captured using responses to two questions on product quality (‘‘finished product first pass quality yield in percentage terms’’ and ‘‘scrap and rework costs as a percentage of sales’’). The second variable, cost improvement, is captured using three questions borrowed from Gatignon and Xuereb (1997) and modified for the purpose of this

Activity-Based Cost Management

229

study (‘‘manufacturing/operations,’’ ‘‘research and development,’’ and ‘‘marketing costs’’).5 The third variable, cycle time improvement, is measured using responses to two questions (‘‘manufacturing cycle from start of production to completion of product in hours’’ and ‘‘standard lead-time from order entry to shipment in days’’). Two measures capture improvement in plant financial performance: ROA and ROS.6 These performance measures relate to changes (improvement) over the last 5 years with higher values for the four change variables that represent greater improvement.7

3.4. Contextual Factors The specification of the theoretical framework (see Fig. 1) consists of a set of hypotheses suggesting structural relationships among variables in the framework. Although ABC factors affect implementation, contextual factors may still play a part (McGowan & Klammer, 1997). Below, we identify some relevant contextual factors and their potential impact on the levels of the constructs relationships in the model as well as on the model relationships. In this study, we are interested in six contextual factors. The first contextual factor is Stage of ABC Implementation. For this study, the three stages of the Cooper and Zmud (1990) model (acceptance, routinization, and infusion) are employed (Foster & Swenson, 1997). Acceptance is achieved when ABC is used at least somewhat by non-accounting management for decision making (Anderson, 1995). Routinization is achieved when ABC is commonly used by non-accounting management for decision making and considered a normal part of the information system. Infusion is defined as not only using ABC extensively but also integrating it with the primary financial system (Reeve, 1996; Kaplan, 1990). The second contextual factor is Advanced Manufacturing Practices (denoted AMP where, 1 ¼ if advanced manufacturing and 0 ¼ if no advanced manufacturing) such as just-in-time, total quality management, MRP, ERP. Studies indicate that these elements interact to improve performance, suggesting that advanced manufacturing is a total system solution rather than a set of independent techniques (Milgrom & Roberts, 1995; MacDuffie, 1995; Chenhall & Langfield-Smith, 1998).8 The third contextual factor is Industry. Type of industry could have moderating effects on the model relationships for several reasons. First, industries differ in terms of types of products and production processes. For example, the chemical industry primarily uses batch and continuous manufacturing processes whereas the automotive or computer industry

230

ADAM S. MAIGA AND FRED A. JACOBS

relies heavily on modular assembly line production. Second, adoption and implementation rigor of technological and managerial innovations has been linked to industry structural characteristics such as domestic and global competitive environment (Porter, 1980). The higher the volatility and global competitive challenges of an industry, the higher the incentive for implementing such initiatives. For example, the automotive and computer industries have adopted contemporary operations improvement strategies such as TQM to a great extent as compared with more stable industries such as pulp and paper or clay and glass (Dreyfus, Ahire, & Ebrhimpour, 2004). Schmenner (1986) classified industry SIC groups into three categories based on the logistical complexity of their production processes (measured by number of steps in the production process). He concluded that the inherent nature of processes could impact the ability of various industries to implement individual elements of productivity improvement techniques. Funk (1995) further used Schmenner’s classification and argued that the logistical complexity of a production system will affect the relevance of various operations improvement techniques. For example, teamwork and cooperation are of greater significance in logistically complex production systems such as automobile and computer assembly than in low logistical complexity production processes such as chemicals or food processing (Funk, 1995). Schmenner (1986) and Funk (1995) classification coded SIC groups 20 through 33 as low logistical complexity industries and SIC 34 through 38 as high logistical complexity industries. Coincidentally, the high logistical complexity industries (electrical machinery, fabricated metal products, industrial machinery, transportation equipment, electronics, and instrumentation) have also experienced fiercer global competition and use of advanced manufacturing technologies as compared with the low logistical complexity industries (Kotha & Vadlamani, 1995). Moreover, these categories are synonymous with classification of process-type production industries (SIC 20 through 33) versus discrete-type production industries (SIC 34 through 38) (Swamidas & Kotha, 1998). Hence, we test the impact of the differing intensity of competition, differing logistical complexity of production, and differing production processes across industries on our model. The fourth factor is Size. Size measures the number of employees at the plant. Plant size9 is an important contingency factor for several reasons. First, larger plants have more market clout, capital resources, and professional managerial expertise (Finch, 1986), and are likely to adopt ABC (Innes & Mitchell, 1995). On the other hand, smaller plants have flatter organizational structures and more informal communication channels

Activity-Based Cost Management

231

(Sirinopolis, 1994). Thus, because smaller factories are more manageable, researchers have associated the smaller size and informal organizational structures with their abilities to encourage and implement innovation (Sirinopolis, 1994).

3.5. Analysis and Results In this section we first present the descriptive statistics. Next, the structural equation model, which consists of two parts, is discussed: the measurement model and the structural model. The measurement model considers the adequacy of the various measures used for theoretical constructs employed in the study, while the structural model specifies the relationships between the various constructs. Strength of structural equation analysis is that multiple indicators are used to represent each unobserved latent construct and that it provides an efficient technique for estimating interrelated dependence relationships, such as those proposed in this study. The contribution of each scale item is incorporated into the estimation of the independent and dependent relationships of the model. This procedure is similar to performing a factor analysis of the scale items and using the factor scores in a regression analysis. Finally, the results of hypotheses testing are presented.

3.6. Descriptive Statistics The descriptive statistics in Table 1, Panel A provide the profile of the responding companies, showing that they constitute a broad spectrum of manufacturers as defined by the two-digit SIC code. The sample composition has the largest representation in electronic and electrical and other (16.080%), followed by industrial equipment (11.223%), primary metals (10.385%), motor vehicles (8.878%), and instruments and related products (8.543%). Additional information on respondents’ characteristics is provided in Table 2, Panel B. The length of ABC implementation has a mean of 8.56 years. The respondents to the question regarding the number of years with the manufacturing plant have a mean of 12.34 years in their current position. To the number-of-years-in-management question, respondents indicated a mean of 18.92 years. While the number of employees range from less than 100 to greater than 350. It appears from their positions and tenure that the respondents are knowledgeable and experienced with access to

232

ADAM S. MAIGA AND FRED A. JACOBS

Table 1.

Descriptive Statistics.

Panel A: Distribution of Two-Digit Industry Classifications SIC Industry Code 20 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 Organization Type Number of Business Units Used in the Study 39 23 24 13 12 13 15 23 17 15 14 21 62 39 67 96 53 51 597 %

Food and kindred Textile mill products Apparel and other fabricated textile products Lumber and wood products Furniture Paper and allied products Printing and publishing Chemical and allied products Petroleum and coal products Rubber and plastics Leather and leather products Stone, clay and glass products Primary metals Fabricated metals Industrial equipment Electronic and other electric equipment Motor vehicles Instruments and related products

6.533 3.853 4.020 2.178 2.010 2.178 2.513 3.853 2.848 2.513 2.345 3.518 10.385 6.533 11.223 16.080 8.878 8.543

Panel B: Other Characteristics of Respondents Minimum Maximum Mean Standard Deviation Length of ABC implementation (years) Length at present position (years) Length in management (years) Number of employees o100 ¼ 24 101–150 ¼ 49 151–200 ¼ 23 201–250 ¼ 105 251–300 ¼ 95 301–350 ¼ 97 >350 ¼ 204 7 9 15 11 13 25 8.56 12.34 18.92 1.93 3.32 3.73

Activity-Based Cost Management

233

Table 2.

Respondents’ Characteristics.

Panel A: Results of Factor Analysis and Measurement Characteristics of the ABC Implementation Variables Factor 1 Eigenvalues Percent variance explained (80.138%) Cronbach alpha Item loadings ABC receives strong active support from top management Upper management has provided adequate resources to the ABC implementation effort ABC has been closely tied to the competitive strategies of the business unit Adequate training was provided for designing ABC Adequate training was provided for implementing ABC Adequate training was provided for using ABC Departments outside accounting (e.g., manufacturing, marketing, etc.) have shown personal ownership for ABC’s success The ABC implementation team was (is) truly cross-functional ABC has been linked to performance evaluations of non-accounting personnel When the ABC initiative began, there was consensus about its specific objectives When the ABC initiative began, its purpose was clear and concise 2.520 22.911 .892 .877 .910 Factor 2 2.410 21.909 .861 .092 .034 Factor 3 2.343 21.301 .850 À.144 .032 Factor 4 1.542 14.017 .701 .084 À.048

.916

À.016

À.038

.101

.038 .032 .031 À.158

.768 .951 .929 .013

À.123 .036 À.017 .880

À.026 .091 .075 À.003

.057 À.044

.010 À.146

.876 .861

.085 .057

À.078

.143

À.011

.882

.207

À.034

.141

.848

234

ADAM S. MAIGA AND FRED A. JACOBS

Table 2.

(Continued )

Panel B: Measurement Characteristics of Plant Performance Item Loading Variance Extracted Reliability (Cronbach Alpha) .711

Quality improvement Finished product first pass quality yield in percentage terms Scrap and rework costs as a percentage of sales Cost improvement Manufacturing/operations costs Research and development costs Marketing costs Cycle time improvement Manufacturing cycle time from start of production to completion of product in hours Standard lead time from order entry to shipment in days Financial performance Return on asset Return on sales

72.249% .849 .851 58.243 .701 .795 .796 83.584 .924

.709

.827

.904 78.312 .889 .880 .819

information upon which to provide reliable perceptions and otherwise well qualified to provide the information required.

3.7. Analysis of Measurement Model We assessed the measurement model using SPSS. The measurement model describes the relation between the latent variables or constructs identified in Fig. 1 and the indicator variables (i.e., scale items). We first checked for sampling adequacy for the benchmarking measures using the Bartlett Test of Sphericity (w2 ¼ 3,659.552, significance ¼ .000) and the Kaiser–Meyer–Olkin (KMO) Measure of Sampling Adequacy (.672). Next, to examine the extent to which these categories are interrelated, we used factor analysis with varimax rotation to determine whether the

Activity-Based Cost Management

235

ABC implementation items used in this study can be grouped according to Krumwiede (1998). Four factors with eigenvalues greater than one emerged from the analysis, with the varimax rotation factor solution retaining 80.138% of the total variance in the data. Three measures vary with top management support: ‘‘ABC receives strong active support from top management,’’ ‘‘Upper management has provided adequate resources to the ABC implementation effort,’’ and ‘‘ABC has been closely tied to the competitive strategies of the business unit.’’ Three measures capture training: ‘‘Adequate training was provided for designing ABC,’’ ‘‘Adequate training was provided for implementing ABC,’’ and ‘‘Adequate training was provided for using ABC.’’ Three measures are associated with non-accounting ownership: ‘‘Departments outside accounting (e.g., manufacturing, marketing, etc.) have shown personal ownership for ABC’s success,’’ ‘‘The ABC implementation team was (is) truly cross-functional,’’ and ‘‘ABC has been linked to performance evaluations of non-accounting personnel.’’ Finally, two measures vary with clarity of objectives: ‘‘When the ABC initiative began, there was consensus about its specific objectives,’’ and ‘‘When the ABC initiative began, its purpose was clear and concise.’’ Factor loadings, eigenvalues, and corresponding Cronbach alphas are provided in Table 2, Panel B. The factor solutions for the defined ABC implementation constructs support the construct validity of the survey instrument. Convergent validity is demonstrated by each factor having multiple-question loadings in excess of .50. In addition, discriminant validity is supported, since none of the questions in the factor analysis have loadings in excess of .40 on more than one factor. No item was dropped because each loading correlated highly with its respective factor, indicating that each item was well reflective of the underlying construct. The factor patterns are consistent with the four factors identified by Krumwiede (1998). Next, in order to assess the consistency or reliability of responses across the endogenous and exogenous items, composite reliability was calculated. Composite reliability measures the internal consistency of the construct’s indicators, similar to coefficient alpha (Fornell & Larcker, 1984). This coefficient is based on the correlations among the responses comprising a scale. Coefficients were quite high for each of the factors. For the endogenous variables, the alpha coefficients were .891 for management support, .862 for training, .850 for non-accounting ownership, and .711 for clarity of objectives. For the exogenous variables, the alpha coefficients were .701 for quality improvement, .735 for cost improvement, .828 for cycle time improvement, and .732 for financial performance. Hence, all measures

236

ADAM S. MAIGA AND FRED A. JACOBS

demonstrated acceptable reliabilities, with coefficients above .70. Nunnally (1967) among others have noted that this is an acceptable standard for the reliability of measures. The Cronbach alphas for the variables and their measurement characteristics are provided in Table 2. Overall, these tests support the validity of the measures representing the constructs used in this study. 3.8. Analysis of the Structural Model Our analysis focused on understanding the nature of the relationships among the constructs under study. Through SEM, we tested our specified framework (see Fig. 1 for a diagram of the model and the testable paths). We evaluated our measurement model and considered the relationship between the observed measures and the latent constructs for both the overall sample and the sub-groups. In this section, we first assess the measures of fit (see Table 3). This fit is expressed using measures of Goodness-of-Fit (GFI). At present, there is no consensus on a single or even on a set of measures of fit (Maruyana, 1998). Thus, it is standard practice to report several measures. We outline below some of the most common measures used in the literature and in this study. (1) The ratio w2 test statistic over the degrees of freedom (w2/df). Good fitting models evidence a ratio of 3.0 or less (Wheaton, Muthen, Alwin, & Summers, 1977). (2) GFI (Bentler & Bonnt, 1980) is based on a w2 likelihood test of the hypothesized model with a null model (no relationships among constructs). Typically, GFI numbers greater than .9 indicate a good fit. (3) Comparative Fit Index (CFI) and Normed Fit Index (NFI) (Bentler & Bonnet, 1980). Both these measures compare the research model specified with the null model (no relationships). The NFI can be viewed as a percent improvement over the null model but does not adjust for the number of parameters in the model. The CFI is based on the w2 distribution. Both NFI and CFI range from 0 to 1 with values exceeding .9 considered good. (4) Root Mean Square Error of Approximation (RMSEA) is computed as the difference between the residuals in the estimated and specified models (Steiger, 1990). Although RMSEA is sensitive to model complexity, it is one of the most informative criteria as to an absolute fit (Byrne, 1998). A value less than .1 is considered a good fit, and a value less than .05 is considered a very good fit of the data to the research

Activity-Based Cost Management

237

Table 3.
Statistical Tests w df w2/df Fit indices GFI CFI NFI Residual analysis RMSEA
2

Overall Fit Summary.
Acceptable Fit Standard N/A N/A o3.0 >.90 >.90 >.90 o.10

Proposed (Tested) Model 266.089 94 2.831 .960 .969 .954 .055

model. For the overall sample measurement model analysis, the overall fit statistics in Table 4 reveal that the proposed (tested) model fits reasonably well the data from all respondents. First, the w2 test statistic associated with the null hypothesis that the proposed model can effectively reproduce the observed covariance is 266.089 with 94 degrees of freedom resulting in a ratio of 2.831. Good fitting is achieved since the ratio is less than 3.0 (Wheaton et al., 1977). Second, the various measures of relative and absolute fit index (ranging from 0 to 1, with 0 implying poor fit and 1 indicating perfect fit) including the GFI, the CFI, and the NFI exceed .9 without any exceptions. Noting that different fit indices have different strengths and weaknesses, this consistent evidence of exceeding the target value of .9 for good-fitting models is encouraging. Third, Table 3 indicates that the difference between reproduced and observed covariances is rather small as evidenced by the RMSEA of .055. Also, to ensure that specification error is not biasing the results, we rerun the factor analysis to allow the errors (i.e., ds) of the measures to covary. According to Hughes et al. (1986), one would expect that if an unobservable error biases the data, a common error variance would be generated between items actually measured. The absence of a significant improvement in overall model fit when these constraints are released would demonstrate the absence of such a bias. We find no significant difference between the fit of the new and original factor analysis models at a confidence level of a ¼ .05. Thus, the proposition that omitted variables are generating biases at the overall model level is rejected at a ¼ .05.

238

ADAM S. MAIGA AND FRED A. JACOBS

Table 4.
Hypotheses Tests

Overall Standardized Path Coefficients and Significance.
Standardized Coefficient Significance

H1: Management support Quality improvement Cost improvement Cycle time improvement H2: Training Quality improvement Cost improvement Cycle time improvement H3: Non-accounting ownership Quality improvement Cost improvement Cycle time improvement H4: Clarity of objectives Quality improvement Cost improvement Cycle time improvement H5: Quality improvement and cost improvement H6: Quality improvement and cycle time improvement H7: Cycle time improvement and cost improvement H8: Quality improvement and financial performance H9: Cost improvement and financial performance H10: Cycle time improvement and financial improvement Explained variances R2 for quality improvement R2 for cost improvement R2 for cycle time improvement R2 for financial performance

.120 .208 À.070 .210 .103 À.489 .305 .166 À.151 .337 À.168 .738 À.273 À.315 .152 .210 .727 .104 .275 .196 .381 .174

.021 .000 .139 .000 .005 .068 .000 .001 .264 .000 .020 .048 .002 .000 .011 .075 .028 .875

3.9. Structural Model Results Before testing the specified hypotheses, we first confirmed the overall model by calculating w2 difference tests to identify any statistically significant paths that are not in the original conceptual model. This procedure has been recommended by Bollen (1989) and others (e.g., Hayduk, 1987; Joreskog & Sorbom, 1993; Medsker, Williams, & Holahan, 1994). None of the w2 difference tests is significant at the .05 level. Therefore, we do not include any additional paths in fully saturated models. Hence, all the hypothesized

Activity-Based Cost Management

239

paths are confirmed, and there are no significant paths between the variables that were not originally identified that would further explain the impacts on financial performance.10 Next, to test the hypotheses, we rely on the standardized parameter estimates for the measurement model (see Table 4 and Fig. 2). The effects of ABC implementation factors on manufacturing performance (quality, cost, and cycle time improvement) show that management support is significantly related to quality improvement (z ¼ .120, p ¼ .021) and cost improvement (z ¼ .208, p ¼ .000), but it is not significantly related to cycle time improvement (z ¼ À.070, p ¼ .139). Training significantly impacts both quality improvement (z ¼ .210, p ¼ .000) and cost improvement (z ¼ .103, p ¼ .005), however training negatively impacts cycle time improvement (z ¼ À.489, p ¼ .068). Non-accounting ownership significantly impacts both quality improvement (z ¼ .305, p ¼ .000) and cost improvement (z ¼ .166, p ¼ .001), but does not significantly affect cycle time improvement (z ¼ À.151, p ¼ .264). Clarity of objectives has significant positive impact on both quality improvement (z ¼ .337, p ¼ .000) and cycle time improvement (z ¼ .738, p ¼ .048); however, its impact on cost improvement is significantly negative (z ¼ À.168, p ¼ .020). The results also show that quality improvement has a significant negative impact on both cost improvement (z ¼ À.273, p ¼ .002) and cycle time improvement (z ¼ À.315, p ¼ .000). Cycle time improvement has a significant positive impact on cost improvement (z ¼ .152, p ¼ .011). Finally, both quality improvement and cost improvement have a significant positive impact on financial performance (z ¼ .210, p ¼ .075; z ¼ .727, p ¼ .028, respectively), while cycle time improvement is not significantly related to financial performance (z ¼ .104, p ¼ .875). Also, the squared multiple correlations (R2) indicate that the model explains (27.50%) variance in quality improvement, 19.60% in cost improvement, 38.10% in cycle time improvement, and 17.40% in financial performance. Further analyses (see Table 5) show that the direct relations between ABC implementation factors and financial performance are not significant, indicating the mediating effects of manufacturing performance between ABC implementation factors and financial performance.

3.10. Contextual Factors Before testing the model relationships for each contextual factor, it is necessary to evaluate its fit to the sub-group samples and its invariance across

240

Management Support

.120* .210*

Quality Improvement .337* -.315* .210*

.305* Training .103* -.151 Nonaccounting Ownership .738* -.273* Clarity of Objectives .166* -.168* .208* Cost Improvement .152* -.489* -.070 Cycle Time Improvement .104 Financial Performance Improvement

ADAM S. MAIGA AND FRED A. JACOBS

.725*

Fig. 2.

Model Path Significance Results and Explained Variances.

Activity-Based Cost Management

241

Table 5. Analyses of Non-Hypothesized Paths. w2 (df ¼ 93) Management support - financial performance Training - financial performance Non-accounting ownership - financial performance Clarity of objectives - financial performance 265.442 265.431 265.598 264.871 w2 Difference .647 .658 .491 1.218 P-value .754 .810 .521 .237

sub-groups (Marsh, 1987; Bollen, 1989). Following the recommendations of Doll, Hendrickson, and Deng (1998), we examined the adequacy of the baseline measurement model for each of the sub-groups as follows: the measurement model was executed and the ratio w2/degree of freedom, GFI, CFI, NFI, and RMSEA values are used to assess the model fit for each subgroup. The fit statistics in Table 6 reveal that the proposed model fits reasonably well the data of each sub-group. First, the ratio ‘‘Chi-square test statistic/ degree of freedom’’ for each sub-group results in a ratio less than 3.0, indicating good fitting (Wheaton et al., 1977). Second, the measures of relative and absolute fit indices exceed .90, and the RMSEA for each sub-group is less than .10. These results demonstrate the overall adequacy of the baseline measurement model for the sub-groups. The results for the sub-group construct model support further analyses of the structural path model. Hence, a LISREL analysis was conducted on the sub-group samples corresponding to each contextual factor.11 The withingroup, completely standardized, coefficient estimates were reviewed for each sub-group of the corresponding contingency models. The within-group completely standardized path coefficient estimates can be used to compare the relative magnitudes of various direct effects within each sub-group. Results of the sub-group structural model analysis show the standardized path coefficient estimates for each sub-group. As in the measurement model, the estimates for each sub-group were obtained in SEM by standardizing the latent variables within a sub-group to unit variance for each sub-group, and are similar to standardized regression coefficients in multiple regression. First, we investigate whether ABC implementation stages affect the model relationships. As shown in Table 7, management support has a significant impact on quality improvement only in the infusion stage, a significant impact on cost improvement in both routinization and infusion stage. Training has a significant negative impact on quality improvement in the acceptance stage, while it has a significant positive impact on quality improvement in both the routinization and

242

ADAM S. MAIGA AND FRED A. JACOBS

Table 6.
Sub-Groups Acceptance stage (n ¼ 120) Routinization stage (n ¼ 206) Infusion stage (n ¼ 271) ABC with AMP (n ¼ 365) ABC without AMP (n ¼ 232)

Sub-Group Model Fit Analyses. w2 df w2/df GFI .944 .959 .947 .956 .958 .959 .954 .960 .961 CFI NFI RMSEA .000 .004 .052 .049 .027 .039 .047 .048 .000

75.177 94 .800 94.318 94 1.003 161.922 94 1.723 175.311 94 1.865 109.382 94 1.164

1.000 1.000 1.000 1.000 .974 .942 .976 .993 .984 .979 .977 1.000 .950 .953 .950 .951 .953 .955

Industry (n ¼ 291) (SIC 20 through 33) 135.111 94 1.437 Industry (n ¼ 306) (SIC 34 through 38) 156.295 94 1.663 Large plants (n ¼ 396) Small plants (n ¼ 201) 178.675 94 1.901 87.953 94 .936

infusion stages, significant positive impact on cost improvement in both routinization and infusion stages. Non-accounting ownership has a significant positive impact on both quality improvement and cost improvement at all stages. Clarity of objectives is significant related to quality improvement at all stage, and cycle time improvement only at the acceptance stage. The impact of quality improvement on cost improvement is negative in all stages. The association between quality improvement and cycle time is negative at both the acceptance stage and the routinization stage. Cycle time improvement is positively related to cost improvement at the infusion stage. Quality improvement has a significant positive impact on financial performance at the infusion stage, while cost improvement significantly affects financial performance at both the routinization stage and the infusion stage. Second, the results for the model relations under adoption of advanced manufacturing practices context in Table 8 show that management support has a significant positive impact on both quality improvement and cost improvement for the sub-group adopting advanced manufacturing practices, and impact on cost improvement in the sub-group not adopting advanced manufacturing practices. Training has significant impact on both quality improvement and cost improvement for the sub-group adopting advanced manufacturing, and significant impact on quality improvement for sub-group not adopting advanced manufacturing practices. Nonaccounting ownership has significant impact on both quality improvement and cost improvement in both sub-groups although the impact is more pronounced for the sub-group adopting advanced manufacturing practices. Clarity of objectives is significantly related to quality improvement in both

Activity-Based Cost Management

243

Table 7.

Standardized Path Coefficients and Significance for ABC Implementation Stages Sub-Groups.
Acceptance Stage Routinization Infusion Stage (n ¼ 120) Stage (n ¼ 206) (n ¼ 271) Stand. Coef.a Sign.b Stand. Coef. Sign. Stand. Coef. Sign.

H1: Management support Quality improvement Cost improvement Cycle time improvement H2: Training Quality improvement Cost improvement Cycle time improvement H3: Non-accounting ownership Quality improvement Cost improvement Cycle time improvement H4: Clarity of objectives Quality improvement Cost improvement Cycle time improvement H5: Quality improvement and cost improvement H6: Quality improvement and cycle time improvement H7: Cycle time improvement and cost improvement H8: Quality improvement and financial performance H9: Cost improvement and financial performance H10: Cycle time improvement and financial improvement Explained variances R2 for quality improvement R2 for cost improvement R2 for cycle time improvement R2 for financial performance a .092 .052 À.045

.368 .690 .647

.107 .237 À.065

.224 .011 .410

.218 .213 À.063

.011 .086 .359

À.231 .048 À.165

.018 .612 .359

.181 .130 À.528

.024 .052 .235

.211 .158 À.893

.009 .070 .242

.226 .252 .282

.023 .083 .367

.309 .195 À.185

.001 .026 .431

.371 .203 .960

.000 .087 .454

.346 À.387 .266 À.380 À.534 .116 .024 .362 À.258

.006 .096 .046 .075 .002 .568 .892 .242 .152

.350 À.079 .195 À.308 À.305 .167 .358 .095 .031

.002 .392 .309 .039 .017 .103 .125 .092 .889

.321 .061 .431 À.347 À.060 .422 .709 .937 .038

.006 .405 .480 .099 .577 .025 .004 .042 .582

.253 .446 .217 .103

.271 .585 .203 .506

.340 .635 .297 .446

b

Standardized coefficient. Significance.

sub-groups, while negatively related to cost improvement in the sub-group not adopting advanced manufacturing practices. The impact of quality of improvement on cost improvement and quality improvement on cycle time is significant but negative for both sub-groups. Cycle time improvement is significantly related to cost improvement in adopting advanced manufacturing practices sub-group. Both quality improvement and cost

244

ADAM S. MAIGA AND FRED A. JACOBS

Table 8. Standardized Path Coefficients and Significance for ABC SubGroups with and without Advanced Manufacturing Practices (AMP).
With AMP (n ¼ 365) Stand. Coef. H1: Management support Quality improvement Cost improvement Cycle time improvement H2: Training Quality improvement Cost improvement Cycle time improvement H3: Non-accounting ownership Quality improvement Cost improvement Cycle time improvement H4: Clarity of objectives Quality improvement Cost improvement Cycle time improvement Sign. Without AMP (n ¼ 232) Stand. Coef. Sign.

.162 .229 À.041 .183 .115 À.465 .350 .137 À.317 .364 .006 .143

.020 .001 .514 .004 .017 .224 .000 .012 .072 .000 .894 .493 .005 .035 .059 .008 .027 .471

.081 .182 À.082 .243 .090 À.587 .264 .183 À.031 .327 À.238 .516 À.298 À.464 .148 .129 .558 À.064

.308 .022 .257 .001 .120 .153 .001 .025 .891 .002 .064 .259 .045 .000 .142 .404 .260 .595

H5: Quality improvement and cost improvement À.321 H6: Quality improvement and cycle time improvement À.210 H7: Cycle time improvement and cost improvement .157 H8: Quality improvement and financial performance .610 H9: Cost improvement and financial performance .652 H10: Cycle time improvement and financial .161 improvement Explained variances R2 for quality improvement R2 for cost improvement R2 for cycle time improvement R2 for financial performance .319 .575 .154 .106

.256 .699 .254 .148

improvement are significantly related to financial performance for subgroups adopting advanced manufacturing practices. Third, Table 9 reports the model relations for industry sub-groups. Management support has significant impact on quality improvement for SIC 34-38 sub-group, and significant impact on cost improvement for both

Activity-Based Cost Management

245

Table 9.

Standardized Path Coefficients and Significance for Industry Sub-Groups.
SIC 20-33 (n ¼ 291) Stand. Coef. Sign. SIC 34-38 (n ¼ 306) Stand. Coef. Sign.

H1: Management support Quality improvement Cost improvement Cycle time improvement H2: Training Quality improvement Cost improvement Cycle time improvement H3: Non-accounting ownership Quality improvement Cost improvement Cycle time improvement H4: Clarity of objectives Quality improvement Cost improvement Cycle time improvement

.098 .189 À.013 .237 .105 À.392 .330 .160 À.135 .306 À.188 .462

.161 .019 .834 .000 .075 .195 .000 .038 .381 .000 .076 .218 .107 .000 .233 .400 .249 .988

.186 .249 À.127 .194 .151 À.715 .289 .217 .787 .383 .017 .108 À.405 À.160 .448 .866 .931 .052

.018 .064 .057 .009 .074 .253 .000 .076 .470 .000 .831 .500 .089 .140 .016 .003 .045 .520

H5: Quality improvement and cost improvement À.194 H6: Quality improvement and cycle time improvement À.412 H7: Cycle time improvement and cost improvement .095 H8: Quality improvement and financial performance .076 H9: Cost improvement and financial performance .444 H10: Cycle time improvement and financial improvement .001 Explained variances R2 for quality improvement R2 for cost improvement R2 for cycle time improvement R2 for financial performance .277 .507 .142 .143

sub-groups. Training significantly impacts both quality improvement and cost improvement for both sub-groups. Non-accounting ownership has a significant impact on both quality improvement and cost improvement for both sub-groups. Clarity of objectives significantly impacts quality improvement for both sub-groups, and negatively impacts cost improvement for SIC 20-33 sub-group. Quality improvement is negatively related to cost

246

ADAM S. MAIGA AND FRED A. JACOBS

improvement for SIC 34-38 sub-group, and negatively related to cycle time improvement for SIC 20-33 sub-group. Cycle time improvement is positively related to cost improvement for SIC sub-group. Quality improvement and cost improvement are both significantly related to financial performance. Fourth, the analysis of the model within the context of size shows in Table 10 that management support affects quality improvement for large Table 10. Standardized Path Coefficients and Significance Plant Size Sub-Groups.
Large Plants (n ¼ 396) Stand. Coef. H1: Management support Quality improvement Cost improvement Cycle time improvement H2: Training Quality improvement Cost improvement Cycle time improvement H3: Non-accounting ownership Quality improvement Cost improvement Cycle time improvement H4: Clarity of objectives Quality improvement Cost improvement Cycle time improvement Sign. Small Plants (n ¼ 201) Stand. Coef. Stand.

.114 .204 À.067 .210 À.104 À.447 .306 .168 À.121 .339 À.171 .865

.075 .000 .250 .000 .023 .159 .000 .005 .473 .000 .058 .107 .013 .001 .044 .141 .071 .871

.131 .216 À.075 .210 .103 À.563 .303 .162 À.202 .336 À.160 .539 À.278 À.331 .159 .220 .737 .006 .277 .197 .839 .097

.139 .027 .357 .012 .108 .251 .001 .052 .376 .003 .189 .249 .076 .011 .124 .289 .203 .972

H5: Quality improvement and cost improvement À.272 H6: Quality improvement and cycle time improvement À.305 H7: Cycle time improvement and cost improvement .148 H8: Quality improvement and financial performance .208 H9: Cost improvement and financial performance .733 H10: Cycle time improvement and financial improvement .018 Explained variances R2 for quality improvement R2 for cost improvement R2 for cycle time improvement R2 for financial performance .275 .196 .795 .074

Activity-Based Cost Management

247

plants, and has significant impact on cost improvement for both sub-groups. Training significantly impacts quality improvement for both plants, but has a negative impact on cost improvement for large plants. Non-accounting ownership significantly affects both quality improvement and cost improvement for both sub-groups. Clarity of objectives is significantly related to quality improvement for both sub-groups, while it is negatively related to cost improvement for large plants. Quality improvement is negatively related to cost improvement for both sub-groups. Cycle time improvement is also negatively related to cycle time improvement for both sub-groups. Cycle time improvement is significantly related to cost improvement for large plants. Finally, cost improvement has a significant effect on financial performance for large plants. In summary, the results show that the model relations vary within the context of analysis. Revisiting Table 7, we note that the ABC plants at the infusion stage performed better then those in the routinization stage, followed by those in the acceptance stage. Also, ABC plants accompanied by advanced manufacturing practices seem to have outperformed those that did not adopt advanced manufacturing practices. Also, SIC 34-38 subgroups seem to have performed better than SIC 20-33. Finally, the analysis of the model within the context of size shows that large plants, overall, outperform small plants. Thus, the results provide support for the notion that ABC implementation is affected by contextual factors.

4. SUMMARY AND DISCUSSION
Ittner et al. (2002) find that extensive ABC use is positively associated with higher quality levels, greater decreases in cycle time, and larger increases in first pass quality. Additionally, their path analysis also indicates that ABC use has a positive indirect association with manufacturing cost reductions through improvements in quality and cycle time. However, they find that, on average, extensive ABC use has no significant association with ROA, rather the relation between ABC and profits varies with the extent to which the decision to use ABC ‘‘matches’’ the plant’s operational characteristics. Our study uses SEM and seeks to provide empirical evidence about the relations between ABC implementation factors and business unit manufacturing performance measures (quality improvement, cost improvement, and cycle time improvement), the relations among manufacturing performance measures, and the impact of manufacturing performance measures on financial performance. The measurement model is adequate for the entire

248

ADAM S. MAIGA AND FRED A. JACOBS

sample as well as for the sub-groups based on four contextual factors (ABC implementation stage, adoption of advanced manufacturing practices, industry, and size). Overall, except for cycle improvement, the hypotheses are supported. However, although the results demonstrate adequacy of the baseline measurement model and sub-group constructs, the results influence the strength of path relations using the contextual factors, indicating that the model relations vary within the context of analysis. These findings are of particular interest to manufacturing units. The importance of ABC implementation in achieving high product quality while reducing product costs and improving cycle time in order to achieve favorable financial outcome becomes critical to production managers. Thus, this study contributes significantly to the literature by improving our understanding of the relationships among factors leading to improved financial performance in strategic business units. This is an important finding as prior studies relating ABC implementation to business financial performance have been mixed. Thus, this study provides strong evidence to suggest that the impact of ABC implementation factors on business unit financial performance could be mediated by manufacturing performance. The study also provides evidence of the importance of contextual factors in the success of ABC implementation. Prior to this research there has been no comprehensive explanation of the plant-specific conditions under which ABC is associated with positive performance results. For example, our results show that when ABC is used concurrently with advanced manufacturing practices, plants have a net improvement in financial performance greater than that obtained from those without use of advanced manufacturing practices. This study provides needed empirical evidence to support analytical and theoretical research regarding the conditions favorable to obtaining benefits from ABC (MacDuffie, 1995; Milgrom & Roberts, 1995; Chenhall & Langfield-Smith, 1998; Krumwiede, 1998). However, there are a number of limitations in this study. For example this study only sampled ABC implementers. Further studies should include firms which have attempted to implement ABC but failed. These studies can then look for differences in firm characteristics, or other factors to explain success or failure with ABC. Also, an interesting approach for empirical research would be to use a research design that captures the longitudinal aspects of the design, implementation, and use of cost management systems. Related, since all research methods have strengths and weaknesses, future research should use multiple methods between studies and within a study (Birnberg, Shields, & Young, 1990). Also, we suggest further study using a more complex model that would include other

Activity-Based Cost Management

249

variables such as firm’s efficiency and effectiveness, customer satisfaction, customer loyalty, and price. Despite the limitations, the results of this study have several implications for managers and researchers. The evidence strongly suggests that the path analytical model offers a useful way for managers to approach ABC success. In particular, given the contextual factor(s), ABC implementation should be incorporated into development of quality improvement, cost improvement and cycle time improvement, and the justification of attaining higher financial performance. The results of this study should enhance practitioners’ confidence in their implementation of ABC and improvement efforts as enablers of financial performance.

NOTES
1. Because of budget constraint, only 2,317 manufacturing units were randomly selected. 2. Of the 14 non-usable responses, 5 business units have mentioned that they have abandoned ABC, 3 responses were returned blank, and 6 were incomplete. 3. To investigate the possibility of non-response bias in the data, the surveys were tested for statistically significant differences in the responses between the early and late waves of returned surveys, with the last wave of surveys received considered to be representative of non-respondents (Armstrong & Overton, 1977). t-tests were performed to compare the mean scores of the early and late responses. The t-tests yielded no statistically significant differences among the survey items, suggesting that non-response bias was not a problem in this study. 4. Linkage to competitive strategies and adequacy of resources are assumed to be closely related to top management support and have been combined as part of top management support. Because non-accountants may be more likely to take ownership for ABC if it is linked to their personal welfare, the linkage of ABC to performance evaluation is combined with the non-accounting ownership factor. Clarity and consensus includes both clarity of purpose and consensus for the objectives of ABC. Training reflects the level of training relating to the design, implementation, and usage of ABC. These three training phases have been combined due to their high correlation with each other and with ABC success in Shields’ (1995) study. 5. Consistent with the literature, the commercial performance of a product can be measured based on multiple items Moenaert, Souder, De Meyer, and Deschoolmeester (1994). 6. Higher values for the four change variables represent greater improvement. 7. See appendix for an abbreviated copy of the research questionnaire used to measure the self-reported variables in this study. 8. For example, Steimer (1990) has suggested that ABC is ideally suited to TQM because it encourages a better analysis of activities through which non-value added activities can be reduced or even eliminated completely.

250

ADAM S. MAIGA AND FRED A. JACOBS

9. To our knowledge, only Ahire and Golhar (1996) and Ahire and Dreyfus (2000) have compared TQM implementation in large firms (more than 250 employees) and small firms (250 or fewer employees). 10. The detailed results of the analyses are available from the authors. 11. Significant effects (positive or negative) are discussed (all p-values o.10 are considered significant). Please see appropriate Tables for other path coefficients.

REFERENCES
Ahire, S. L., & Dreyfus, P. (2000). The impact of design management and process management on quality: An empirical investigation. Journal of Operations Management, 18(5), 549–575. Ahire, S. L., & Golhar, D. Y. (1996). Quality management in large and small firms. Journal of Small Business Management, 34(2), 1–13. Anderson, J. C., Rungtusanathan, M., Schroeder, R. G., & Dearaj, S. (1995). A path analytic model of a theory of quality management underlying the Deming management method: Preliminary empirical findings. Decision Sciences, 26(5), 637–658. Anderson, S. W. (1995). A framework for assessing cost management system changes: The case of activity based costing implementation at General Motors. Journal of Management Accounting Research, 7, 1–51. Anderson, S. W., & Young, S. (1999). The impact of contextual and process factors on the evaluation of activity-based costing systems. Accounting, Organizations and Society, 24, 525–559. Armstrong, J. S., & Overton, T. S. (1977). Estimating nonresponse bias in mail survey. Journal of Marketing Research, 14(3), 396–421. Bentler, P. M., & Bonnet, D. G. (1980). Significance tests and goodness-of-fit in the analysis of covariance structures. Psychological Bulletin, 88, 588–600. Birnberg, J., Shields, M., & Young, S. M. (1990). The case for multiple methods in empirical management accounting research (with an illustration from budget setting). Journal of Management Accounting Research, 2, 33–66. Bollen, K. A. (1989). Structural equations with latent variables. New York, NY: Wiley. Bolton, R. N., & Drew, J. M. (1991). A multistage model of customers’ assessments of service quality and value. Journal of Consumer Research, 17(4), 875–884. Bonito, J. G. (1990). Motivating employees for continuous improvement efforts – Part 3: Additional critical success factors. Production and Inventory Management Review APICS News, 10(8), 32–33. Bradford, M., & Roberts, D. (2001). Does your ERP system measure up? Strategic Finance, 83(3), 30–35. Burack, E., Burack, M., Miller, D., & Morgan, K. (1994). New paradigm approaches in strategic human resource management. Group and Organization Management, 19(2), 141–159. Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS: Basis, concepts, application, and programming. Mahwah, NJ: Lawrence Erlbaum. Cagwin, D., & Bouwman, M. J. (2002). The association between activity-based costing and improvement in financial performance. Management Accounting Research, 13, 1–39.

Activity-Based Cost Management

251

Campell, R. J. (1995). Steering time with ABC or TOC. Management Accounting, 76(7), 1–36. Chenhall, R. H. (2004). The role of cognitive and affective conflict in early implementation of activity-based cost management. Behavioral Research in Accounting, 16, 19–44. Chenhall, R. H., & Langfield-Smith, K. (1998). The relationship between strategic priorities, management techniques and management accounting: An empirical investigation using a systems approach. Accounting, Organizations and Society, 23(3), 243–264. Choi, T. Y. (1995). Conceptualizing continuous improvement: Implications for organizational change. Omega, 23, 607–624. Cooper, R., & Kaplan, R. (1991). The design of cost management systems. Englewood Cliffs, NJ: Prentice-Hall. Cooper, R., Kaplan, R., Maisel, L., Morrissey, E., & Oehm, R. (1992). Implementing activitybased cost management. Montvale, NJ: Institute of Management Accountants. Cooper, R. B., & Zmud, R. W. (1990). Information technology implementation research: A technological diffusion approach. Management Science, 36(2), 123–139. Crosby, P. B. (1979). Quality is free. New York, NY: McGraw-Hill. Daft, R. L. (1998). Organization theory and design. Cincinnatti, OH: South-Western College. Dale, B. G., & Lightburn, K. (1992). Continuous quality improvement: Why some organizations lack commitment. International Journal of Production Economics, 27(1), 57–68. Deming, W. E. (1986). Out of the crisis. Cambridge, MA: M.I.T. Center for Advanced Engineering. Doll, W. J. (1985). Avenues for top management involvement in successful MIS development. MIS Quarterly, 9(1), 17–35. Doll, W. J., Hendrickson, A., & Deng, X. (1998). Using Davis’ perceived usefulness and easeof-use instruments for decision making: A confirmatory and multigroup invariance analysis. Decision Sciences, 29(4), 839–870. Dreyfus, L. P., Ahire, S. L., & Ebrhimpour, M. (2004). The impact of just-in-time implementation and ISO 9000 certification on total quality management. IEEE Transactions on Engineering Management, 51(2), 125–151. Flynn, B., Schroeder, R. G., & Sakakibara, S. (1995). The impact of quality management practices on performance and competitive advantage. Decision Sciences, 26(5), 659–691. Finch, B. J. (1986). Japanese management techniques in small manufacturing companies: A strategy for implementation. Production and Inventory Management, 27(3), 30–38. Fornell, C., & Larcker, D. F. (1984). Misapplications of simulations in structural equation models: Reply to Acito and Anderson. Journal of Marketing Research, 21(1), 113–117. Foster, G., & Swenson, D. W. (1997). Measuring the success of activity-based cost management and its determinants. Journal of Management Accounting Research, 9, 107–139. Funk, J. L. (1995). Just-in-Time manufacturing and logistical complexity: A contingency model. International Journal of Operations and Production Management, 15(6), 60–71. Gale, B. T. (1994). Customer satisfaction – relative to competitors – is where it’s at. Strong evidence that superior quality drives the bottom line and shareholder value. Marketing and Research Today, 22(1), 39–53. Garvin, D. A. (1998). The processes of organization and management. Sloan Management Review, 39(4), 33–50. Gatignon, H., & Xuereb, J. M. (1997). Strategic orientation of the firm and new product performance. Journal of Marketing Research, 34, 77–90. Gujarati, D. N. (1995). Basic econometrics. New York, NY: McGraw-Hill.

252

ADAM S. MAIGA AND FRED A. JACOBS

Gunasekaran, A., Korukonda, A. R., Virtanen, I., & Olli, Y. (1994). Improving productivity and quality in manufacturing organizations. International Journal of Production Economic, (36), 169–183. Hambrick, D. C. (1983). High profit strategies in mature capital goods industries: A contingency approach. Academy of Management Journal, 26, 687–707. Hamlin, B., Reidy, M., & Stewart, J. (1997). Changing the management culture in one part of British Civil Service through visionary leadership and strategically led research-based OD interventions. Journal of Applied Management Studies, 6(2), 233–251. Harter, D. E., Krishnan, M. S., & Slaughter, S. A. (2000). Effects of process maturity on quality, cycle time, and effort in software product development. Management Science, 46(4), 451–467. Hayduk, L. A. (1987). Structural equation modeling with LISREL: Essentials and advances. Baltimore, MD: John Hopkins Press. Hayes, R. H., Wheelwright, S. C., & Clark, K. (1988). Dynamic manufacturing. New York, NY: The Free Press. Henderson, B., & Henderson, D. (1979). Corporate strategy. Cambridge, MA: Abt Books. Hughes, M. A., Price, L. R., & Marrs, D. W. (1986). Linking theory construction and theory testing: Models with multiple indicators of latent variables. Academy of Management Review, 11(1), 128–144. Innes, J., & Mitchell, F. (1995). A survey of activity-based costing: A survey of CIMA members. Management Accounting Research, 6(June), 137–153. Innes, J., Mitchell, F., & Sinclair, D. (2000). Activity-based costing in the U.K.’s largest companies: A comparison of 1994 and 1999 survey results. Management Accounting Research, 11, 349–362. Ittner, C. (1994). An examination of the indirect productivity gains from quality improvement. Productions and Operations Management, 3(3), 153–170. Ittner, C., Lanen, W. N., & Larcker, D. F. (2002). The association between activity-based costing and manufacturing performance. Journal of Accounting Research, 40(3), 711–726. Jeffery, B., & Morrison, J. (2000). ERP, one letter at a time. CIO Magazine, 13(22), 72–76. Joreskog, K. G., & Sorbom, D. (1993). LISREL 8: User’s reference guide. Chicago, IL: Scientific Software. Juran, J. M. (1988). Juran on planning for quality. New York, NY: The Free Press. Kaplan, R. S. (1990). The four-stage model of cost systems design. Management Accounting, 71(8), 22–26. Kaplan, R. S., & Norton, D. P. (1992). The balanced scorecard measures that drive performance. Harvard Business Review, 70(1), 71–79. Kaplan, R. S., & Norton, D. P. (1993). Putting the balanced scorecard to work. Harvard Business Review, 71(5), 134–147. Kaplan, R. S., & Norton, D. P. (1996). Using the balanced scorecard as a strategic management system. Harvard Business Review, 74(1), 75–85. Kaynak, H. (2003). The relationship between total quality management practices and their effects on firm performance. Journal of Operations Management, 21, 405–435. Kennedy, T., & Affleck-Graves, J. (2001). The impact of activity-based costing techniques on firm performance. Journal of Management Accounting Research, 13, 19–45. Kline, R. B. (1998). Principles and practice of structural equation modeling. New York, NY: The Guilford Press.

Activity-Based Cost Management

253

Kotha, S., & Vadlamani, B. L. (1995). Assessing generic strategies: An empirical investigation of two competing typologies in discrete manufacturing industries. Strategic Management Control, 16(1), 75–83. Krumwiede, K. R. (1998). The implementation stages of activity based costing and the impact of contextual and organizational factors. Journal of Management Accounting Research, 10, 239–277. Locke, E. A., Shaw, K. N., Saari, L. M., & Latham, G. P. (1981). Goal setting and task performance. Psychological Bulletin, 90, 125–152. Lyne, S., & Friedman, A. (1996). Activity-based techniques and the new management accountant. Management Accounting, 74, 34–35. MacDuffie, J. P. (1995). Human resource bundles and manufacturing performance: Orgnaizational logic and flexible production systems in the world auto industry. Industrial and Labor Relations Review, 48(2), 197–221. Madu, C. N., & Kuei, C. (1995). The view of quality: Middle managers’ perspectives. Industrial Management, 37(5), 20–22. Malmi, T. (1997). Towards explaining activity-based costing failure: Accounting and control in a decentralized organization. Management Accounting Research, 7, 459–480. Marsh, H. W. (1987). The factorial invariance of responses by males and females to a multidimensional self-concept instrument: Substantive and methodological issues. Multivariate Behavioral Research, 22, 457–480. Maruyana, G. M. (1998). Basics of structural equation modeling. London: Sage Publications, Inc. McGowan, A. (1994). An investigation of the behavioral implications of adopting activitybased cost management systems: An exploratory study. Dissertation, University of North Texas. McGowan, A., & Klammer, T. (1997). Satisfaction with activity-based cost management implementation. Journal of Management Accounting Research, 9, 217–237. McGowan, M. K., & Madey, G. R. (1998). Adoption and implementation of electronic data interchange. In: R. J. Larson & E. McGuire (Eds), Information systems innovation and diffusion: Issues and directions (pp. 116–140). London: Idea Group. Medsker, G. J., Williams, L. J., & Holahan, P. J. (1994). A review of current practice for evaluating causal models in organizational behavior and human resources management research. Journal of Management, 20, 439–464. Meyer, C. (1993). Fast cycle time: How to align purpose, strategy and structure for speed. New York, NY: The Free Press. Milgrom, P., & Roberts, J. (1995). Complementarities and fit: Strategy, structure, and organizational change in manufacturing. Journal of Accounting and Economics, 19(2,3), 179–208. Milligan, B. (1999). Buyers face new supply challenges. Purchasing, 127(7), 63–72. Mintzberg, H. (1979). The structuring of organizations. Englewood Cliffs, NJ: Prentice-Hall. Moenaert, R. K., Souder, W. E., De Meyer, A., & Deschoolmeester, D. (1994). R&DMarketing integration mechanisms, communication flows, and innovation success. Journal of Product Innovation Management, 11(1), 3l–45. Morrow, M., & Connolly, T. (1994). Practical problems of implementing ABC. Accountancy, (3), 76–78. Nandakumar, P., Datar, S. M., & Akella, R. (1993). Models for measuring and accounting for cost of conformance. Management Science, 39(1), 1–16.

254

ADAM S. MAIGA AND FRED A. JACOBS

Nunnally, J. C. (1967). Psychometric theory (1st ed.). New York: McGraw-Hill. Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1985). A conceptual model of service quality and its implications for future research. Journal of Marketing, 49, 41–50. Phillips, L. W., Chang, D. R., & Buzzell, R. D. (1983). Product quality, cost position and business performance: A test of some key hypotheses. Journal of Marketing, 47, 26–43. Porter, M. E. (1980). Competitive strategy: Techniques for analyzing industries and competitors. New York, NY: Free Press. Porter, M. E. (1985). Competitive advantage: Creating and sustaining superior performance. New York, NY: Free Press. Powell, C. (1995). International direct mail comes of age. Direct Marketing, 58(8), 43–44. Ramaswamy, R. (1996). Design and management of service processes: Keeping customers for life. Reading, MA: Addison Wesley. Reeve, J. M. (1996). Projects, models, and systems: Where is ABM headed?. Journal of Cost Management, 10(2), 5–15. Roth, H. P., & Borthick, A. F. (1989). Getting closer to real product costs. Management Accounting, 70(11), 28–33. Rust, R. T., Moorman, C., & Dickson, P. R. (2002). Getting return on quality: Revenue expansion, cost reduction, or both? Journal of Marketing, 66(4), 7–14. Saraph, J. V., Benson, P. G., & Schroeder, R. G. (1989). An instrument for measuring the critical factors of quality management. Decision Sciences, 20(4), 810–829. Schilling, M. A., & Hill, C. W. L. (1998). Managing the new product development process: Strategic imperatives. The Academy of Management Executive, 12(3), 67–81. Schmenner, R. (1986). Comparative factory productivity: A report prepared for the Economic Development Administration, Washington, D.C., NTS No. PB87-140877. Sharland, A., Eltantawy, R. A., & Giunipero, L. C. (2003). The impact of cycle time on supplier selection and subsequent performance outcomes. The Journal of Supply Chain Management, 39(3), 4–9. Sharman, P. A. (1993). Activity-based management: A growing practice. CMA, 67(2), 17–21. Shields, M. D. (1995). An empirical analysis of firms’ implementation experiences with activitybased costing. Journal of Management Research, 7, 148–166. Shields, M., & Young, S. M. (1994). Managing innovation costs: A study of cost conscious behavior by R&D professionals. Journal of Management Accounting Research, 6, 175–196. Shields, M. D., & Young, S. M. (1989). A behavioral model for implementing cost management systems. Journal of Cost Management, 2(Winter), 17–27. Simers, D., & Priest, J. W. (1989). The expanding role of the manufacturing supervisor. Industrial Management, 31(2), 27–29. Sirinopolis, N. (1994). Small business management: A guide to entrepreneurship (5th ed.). Boston, MA: Houghton Mifflin. Snell, S. A., & Dean, J. W., Jr. (1992). Integrated manufacturing and human resource management: A human capital perspective. Academy of Management Journal, 35, 467–504. Stalk, G., Jr. (1988). Time: The next source of competitive advantage. Harvard Business Review, 66(4), 41–51. Stalk, G., Jr., & Hout, T. (1990). Competing against time. New York, NY: The Free Press. Steiger, J. H. (1990). Structural model evaluation and modification: An interval estimation approach. Multivariate Behavioral Research, 25, 173–180.

Activity-Based Cost Management

255

Steimer, T. (1990). Activity-based costing for total quality. Management Accounting, 72(4), 39–42. Swamidas, P. M., & Kotha, S. (1998). Explaining manufacturing technology use, firm size, and performance using a multidimensional view of technology. Journal of Operations Management, 17(1), 23–39. Tait, P., & Vessey, I. (1988). The effect of user involvement on systems success: A contingency approach. MIS Quarterly, 12(1), 91–108. Tatikonda, L. U. (2003). Critical issues to address before you embark on an ABC journey. The National Public Accountant, 5. Tatikonda, M. V., & Montoya-Weiss, M. M. (2001). Integrating operations and marketing perspectives of product innovation: The influence of organizational process factors and capabilities on development performance. Management Science, 47(1), 151–172. Tichy, N., & Devanna, M. A. (1986). The transformational leader. Industrial Management, 10(9), 44–56. Voss, C., Blackmon, K., Hanson, P., & Oak, B. (1995). The competitiveness of European manufacturing: A four country study. Business Strategy Review, 6(1), 1–25. Wheaton, B., Muthen, B., Alwin, D., & Summers, G. (1977). Assessing reliability and stability in panel models. In: D. Heise (Ed.), Sociological methodology (pp. 84–136). San Francisco, CA: Jossey-Bass. Young, S. M., & Selto, F. H. (1993). Explaining cross-sectional workgroup performance difference in a JIT facility: A critical appraisal of a field-based study. Journal of Management Accounting Research, 5, 300–330.

256

APPENDIX: QUESTIONNAIRE
Part I Please provide the extent to which the following items are present for your business unit’s ABC implementation. 1 ¼ Extremely Low ABC receives strong active support from top management Upper management has provided adequate resources to the ABC implementation effort ABC has been closely tied to the competitive strategies of the business unit Adequate training was provided for designing ABC Adequate training was provided for implementing ABC 1 2 3 4 5 6 7 ¼ Extremely High
ADAM S. MAIGA AND FRED A. JACOBS

7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

1 1

2 2

3 3

4 4

5 5

6 6

7 7

Activity-Based Cost Management

Adequate training was provided for using ABC Departments outside accounting (e.g., manufacturing, marketing, etc.) have shown personal ownership for ABC’s success The ABC implementation team was (is) truly cross-functional ABC has been linked to performance evaluations of non-accounting personnel When the ABC initiative began, there was consensus about its specific objectives When the ABC initiative began, its purpose was clear and concise

1 1

2 2

3 3

4 4

5 5

6 6

7 7

1 1

2 2

3 3

4 4

5 5

6 6

7 7

1

2

3

4

5

6

7

1

2

3

4

5

6

7

257

258

APPENDIX (Continued )
Part II Please indicate the extent to which the following performance measures have improved over the last 5 years. 1 ¼ Extremely Low Improvement Finished product first pass quality yield in percentage terms Scrap and rework costs as a percentage of sales Manufacturing/operations costs Research and development costs Marketing costs Manufacturing cycle time from start of production to completion of product in hours Standard lead time from order entry to shipment in days Return on assets Return on sales 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 4 5 5 5 5 5 5 6 6 6 6 6 6 7 ¼ Extremely High Improvement
ADAM S. MAIGA AND FRED A. JACOBS

7 7 7 7 7 7

1 1 1

2 2 2

3 3 3

4 4 4

5 5 5

6 6 6

7 7 7

Activity-Based Cost Management

259

Part III The following statements are used to classify your business units into one of three ABC implementation stages. Please check mark below the stage to which your business unit belongs. A. Acceptance is achieved when ABC is used at least somewhat by nonaccounting management for decision making. B. Routinization is achieved when ABC is commonly used by nonaccounting management for decision making and considered a normal part of the information system. C. Infusion is defined as not only using ABC extensively but also integrating it with the primary financial system. A. B. C. Other(s): (please specify) ___________ ___________ ____________ __________________________________

Part IV Regarding advanced manufacturing practices, please indicate which of the following is used in your business unit (please check mark all that apply). Balanced Scorecard Benchmarking Computer Integrated Manufacturing (CIM) Computer Aided Engineering (CAE) Computer Aided Design (CAD) Flexible Manufacturing Systems (FMS) Just-in-Time Total Quality Management Other(s): (please specify) ______ ______ ______ ______ ______ ______ ______ ______ ______

260

ADAM S. MAIGA AND FRED A. JACOBS

Part V Please answer the following: How long have you implemented ABC? What is your business two-digit SIC code? What is the number of employees at your company? Please check below: o100 101–150 151–200 201–250 251–300 301–350 >350 4. Number of years at this position? 5. Number of years in management? 1. 2. 3. __________ __________ __________ ____________ ____________ ____________ ____________ ____________ ____________ ____________ ___________ ___________

TEAM PERFORMANCE MEASUREMENT: A SYSTEM TO BALANCE INNOVATION AND EMPOWERMENT WITH CONTROL
Frances Kennedy and Lydia Schleifer
ABSTRACT
A current highly competitive and rapidly changing business environment requires companies to continually innovate to survive. An increasing number of companies are using teams to leverage the knowledge and experience of their employees in order to improve quality, reduce costs and ‘delight’ the customer. The growing prevalence of teams signals the need to examine the adequacy of management accounting information and its use in performance measurement and control systems. Some research has examined the impact of team empowerment on creativity and innovation, while other research discusses the sometimeshampering role of performance measures in team environments. This paper contributes to this research, with two major goals. First, it discusses innovation and empowerment and examines how performance measurement can both encourage and hinder team performance. The second purpose is to propose a team performance measurement system using ratios based on activity-based management that seeks to encourage innovation and empowerment while maintaining a system of control.

Advances in Management Accounting, Volume 16, 261–285 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16009-3

261

262

FRANCES KENNEDY AND LYDIA SCHLEIFER

INTRODUCTION
In a competitive, ever changing world, the success and survival of organizations depends on innovation (Mumford & Licuanan, 2004, p. 163).

Changing an organization to compete in a highly volatile business environment usually requires multiple and continuous innovations (Damanpour, 1991). Innovation has been defined as new approaches to methods or technologies (Schroeder, Scudder, & Elm, 1989) that help to meet organizational objectives (Drake, Haka, & Ravenscroft, 1998). It begins as a creative idea that has been implemented within an organization (Amabile, 1996). Teams help companies to compete by improving quality, reducing costs and developing new products (Alper, Tjosvold, & Law, 2000). The premise is that through the pooling of ideas, expertise and experiences, teams will more quickly develop innovations (Scott & Tiessen, 1999) that will better position the company to compete in this challenging business environment. Achieving flexibility requires moving decision-making authority to lower levels (Simons, 1995). As companies are changing their work structures to involve teams, teams are being empowered to make decisions within the scope of the team mission. Sim and Carey (2003, p. 112) define empowerment as ‘‘y a means of giving the authority to make decisions to that level or people in the organization which, by virtue of available knowledge and closeness to the activity concerned, is most able to make a correct, quick, and effective decision.’’ It is now fairly widely recognized and accepted that empowerment can be associated with effectiveness of managers and corporations (Bennis & Nanus, 1985; McClelland, 1975), and can lead to higher commitment to a project and a greater chance to meet team goals (McDonough, 2000). Empowered teams contribute to achieving success in implementing such new management philosophies as total quality management (TQM), just in time (JIT) (Cua, McKone, & Schroeder, 2001) and world class manufacturing (WCM) (Lind, 2001). The growing prevalence of teams in organizations signals the need to examine the adequacy of management accounting information. A recent survey found that 81% of Fortune 500 companies are building, at least partially, team-based organizations and 77% use temporary project teams to perform core work (Lawler, Mohrman, & Benson, 2001). Work structures that involve teams are often more efficient and effective than individual work (Banker, Potter, & Schroeder, 1993). Birnberg (1999) argues that coalitions evolve in order to facilitate the pooling of information. This concept suggests that team structures develop to make information

Team Performance Measurement

263

more available and decisions more transparent. In their overview of management accounting research, Atkinson et al. (1997, p. 80) state that, ‘‘Changing [team] structures imply changes in the information needed, and the way information is used to measure and motivate performance.’’ Indeed, Scott and Tiessen (1999) found that teams are a response to complexity and use financial and nonfinancial measures to adequately capture performance. Simons (1995, p. 110) suggests that empowerment requires greater control: ‘‘The control systems used, however, must balance empowerment and control in such a way that empowerment does not lead to a control failure, and correspondingly, control does not lead to an empowerment failure.’’ Control systems that include performance measures have traditionally focused on results-based measures (Merchant, 1985). McNair and Carr (1994) argue that the traditional view of control uses outcome measures largely based on a pre-determined historical performance standard and is ill equipped to add value to modern organizations that increasingly emphasize achieving continuous improvement goals. Current measurement systems focus on monitoring and controlling behavior when what is needed with new work structures are systems that support process improvement – the new ‘mantra’ of the age of world class manufacturing (WCM) (Ghalayini, Noble, & Crowe, 1997). The dilemma becomes how to measure and monitor the progress of teams while providing the encouragement and environment necessary to align team behavior with the organizational goal of innovation. This paper has two goals. It first examines possible influences of performance measurement on innovation by exploring their effects on five conditions for team creativity (Amabile et al., 1996) and four dimensions of team empowerment (Kirkman & Rosen, 1999). The second goal is to propose a team measurement system that seeks to balance innovation and empowerment with control. This measurement system was developed with input from representatives of five diverse companies in both service and manufacturing industries. The goal of this development team was to define a system of measurement that assessed team performance as well as measured the financial impact of team structures on overall performance. The next section describes challenges in current performance measures with regard to team measurement. This is followed by a discussion of performance measurement effects on innovation and the enabling conditions that promote creativity, and on four dimensions of team empowerment. Then, a team performance measurement system (TPMS) is proposed. The TPMS is then critically assessed according to the models of innovation and empowerment. Finally, limitations and future research are addressed.

264

FRANCES KENNEDY AND LYDIA SCHLEIFER

TEAM PERFORMANCE MEASUREMENT CHALLENGE
Performance Measurement and Control Rapidly changing customer demands require a high degree of customization (Maskell & Baggaley, 2004) that can only be met with flexible work systems (Kalagnanam & Lindsay, 1998). This trend requires changes in organization structures – both in production processes and in measurement systems. Ansari, Bell, Klammer, and Lawrence (2004, p. 4) define management accounting as ‘‘y a system of measuring and providing operational and financial information that guides managerial action, motivates behaviors, and supports and creates the cultural values necessary to achieve an organization’s strategic objectives.’’ This broad definition not only allows management accountants to consider appropriate measurement techniques for contemporary flexible systems, but also demands that they strive to provide the right balance of information to mitigate actions contrary to organizational objectives. Control involves focusing on human behavior in an effort to influence people to strive towards meeting organizational goals. Merchant’s (1985) control framework includes results, action and personnel controls. Results controls involve rewarding people or holding them accountable for certain results. Action controls concern how work is done and involve providing periodic feedback that decision-makers can use to assess progress and alter actions. Personnel controls involve managerial actions taken to ensure having the right people through hiring, training, team assignment and motivational practices. Skills are developed through correction and reinforcement with feedback information (or action controls) (Brannick & Prince, 1997). Action controls are, therefore, critical for both changing course and for skill development. Nanni, Dixon, and Vollman (1990) assert that management accountants have tended to provide performance measures with a product-oriented rather than a process-oriented focus. McNair and Carr (1994) suggest there may be a shift from results to action controls in order to support flexibility and speed. In their review of performance measurement systems, Ghalayini et al. (1997) argue that current systems focus on monitoring and controlling when what is needed are systems that support process improvement. Process measurement is an action control that provides feedback mechanisms that support continuous improvement goals and a rapidly changing environment. Traditional measurement, on the other hand, is largely designed to

Team Performance Measurement

265

encourage boundaries and maintain the status quo through systems of accountability. Simons (1995) argues that with decision-making authority (empowerment) comes accountability for decisions, and accountability can be perceived as a great motivator (Epstein & Birchard, 2000). Tetlock’s (1985) behavioral theory of accountability maintains that simply the knowledge that one will be held accountable affects behavior. Further, when the dimensions upon which individuals will be held accountable are clearly defined, they will organize themselves around those dimensions. Therefore, establishing a system that clearly makes managers and teams accountable will make them more aware of the need to react to information cues (Birnberg, 1999). Birnberg suggests that high levels of accountability may be inconsistent with high levels of innovation. As the responsibility for decisions descends to lower levels, there is greater need for mechanisms that consider the dispersion of decision authority. Failure to make appropriate adjustments to traditional measurement can lessen team effectiveness, undermine new management practices, such as JIT or WCM (McNair & Carr, 1994), and cause members to forget team goals and revert to prior work patterns (Meyer, 1994). In addition to technical considerations of what to measure and how, it is critical to consider the behavioral effects of the measures on the teams. The next section describes models of innovation and empowerment, and discusses how performance measures can both positively and negatively influence team behavior in these areas.

Behavioral Implications of Team Measurement Previous discussion highlighted innovation and empowerment as critical factors in successful teaming organizations as well as the need for action and results controls to provide appropriate control mechanisms. A performance measurement system is an essential ingredient in a control system. The following discussion describes the potential influences of team performance measurements on the dimensions of Amabile et al.’s (1996) and Kirkman and Rosen’s (1999) models of innovation and empowerment. Innovation Schroeder et al. (1989) define innovation as new approaches or technologies that help to meet organizational objectives (Drake et al., 1998) and is essential to the future success of the organization (Brennan & Dooley, 2005).

266

FRANCES KENNEDY AND LYDIA SCHLEIFER

Argrell and Gustafson (1996) portray innovation as an interactive process. Teams intentionally attempt ‘‘to arrive at [through their own efforts] anticipated benefits for the individual, team, organization or society as opposed to top-down change’’ (Drach-Zahavy & Somech, 2001, p. 111). Managing motivation is about managing an entire system of innovation rather than about doing any one thing really well, like research and development (Bessant, Lamming, Noke, & Phillips, 2005). Motivating innovation is complex and some research supports that traditional external motivators, such as rewards, may diminish creativity. The greatest creativity emerges when employees are intrinsically motivated through enjoyment, interest or challenge (Amabile et al., 1996). A company’s ability to develop and exploit creative abilities is critical to the eventual development of innovations. A primary consideration when developing a culture that is conducive to creativity is the importance of perception. Employee or team perception of reality is more important than whether that reality actually exists (Amabile et al., 1996; Spreitzer, 1996) because perceptions shape interpretations (Thomas & Velthouse, 1990). These interpretations then become a stronger influence than actual occurrences. Developing and nurturing a culture that is perceived by employees as encouraging creativity and valuing innovation should be the goal of innovative organizations. Amabile et al. (1996) focused on identifying conditions that must exist to foster creativity and implement ideas into innovations. The result was a model of five dimensions that promote (or stifle) creativity. (1) Encouragement of creativity comes from several sources: the organization, the supervisor and the work group itself. By setting goals and providing performance evaluations, the organization can encourage or inhibit risk-taking and idea generation. If creating new ideas is explicitly included as part of their job, employees will be more likely to engage in that type of activity. The expectation of a fair performance evaluation can motivate teams; however, if the expectation is threatening, employees are less likely to take chances and try something new. Supervisors can encourage creativity by being actively involved with the team, communicating clear goals to the team and supporting the team’s ideas. These actions communicate that the team’s role is meaningful to the organization. The group itself can also provide encouragement. The members of a team can encourage each other by exchanging ideas, being openly supportive and demonstrating a shared commitment. (2) Autonomy originates from a sense of ownership over decisions and processes. When team members feel a sense of ownership, they are more likely to seek information and make decisions concerning their processes.

Team Performance Measurement

267

The team’s creativity is assumed to increase as choice increases (Amabile et al., 1996). (3) The allocation of resources to team projects signals to the team that its work is important. When the team perceives that its contribution positively impacts the organization, there is likely to be even more creativity and innovation. The perception that resources will support innovation will further enhance the team’s creativity. (4) Pressures interact with the creative processes in two potentially opposite ways. Excessive ‘workload’ pressures can suppress creativity. On the other hand, the pressures of a good ‘challenge’ can stretch ideas and promote creativity. (5) Organizational impediments include formal management structures, conflict and conservatism. These are external influences that send mixed signals and detract from idea generation and creative processes. If team members perceive that such impediments provide the extrinsic motivation, or pressure, to accomplish tasks, then there may be a decrease in the intrinsic motivation, like satisfaction, that actually inspires creativity (Amabile et al., 1996). As outlined in Exhibit 1, performance measurement can both encourage and hinder creativity. On one hand, capturing team innovations in a measurement system enables the team to be recognized and rewarded. On the other hand, performance reviews that are critical and punishing on the basis of this measure can result in the employee or team not being willing to try new ideas and suppressing the flow of innovations. As teams begin to make more decisions, they also experience a heightened accountability for those decisions. Measuring the impact of decisions can help a team to think about long-term implications as well as the scope of their decisions. However, if teams are held accountable for decisions or outcomes over which they have little or no control, their motivation to continue to participate in the creative process is hampered. Teams, particularly newly formed teams, are often challenged to prioritize projects. They look to the cues in the organization to help them get started. A performance measurement system can help them accomplish this task and relieve some of the initial confusion and pressure. In addition, having a clear vision of expectations and targets helps the team to adjust their activities and can enhance the challenge, intrinsically motivating team members. Finally, performance measurement systems that are not aligned with team goals or that are used as a ‘club’ to oversee team processes can be perceived negatively by teams and can promote behavior not conducive to the creative process. Amabile et al.’s (1996) model of conditions that foster creativity highlights the intricacies involved in maintaining an environment conducive to creative ideas and subsequent innovation. This creativity is critical for process innovation (Sim & Carey, 2003) and requires fostering an

268

FRANCES KENNEDY AND LYDIA SCHLEIFER

Exhibit 1.
Dimension

Team Performance Measurement Influence on Dimensions of Creativity and Empowerment.
Description Influence of Team Performance Measurement

Creativity (Amabile, Conti, Coon, Lazenby, & Herron, 1996  Critical reviews and Encouragement of Includes (1) fostering an consequences can repress creativity environment that promotes risk-taking and idea risk-taking.  Capturing team successes generation, (2) supporting makes the accomplishment new ideas, (3) rewarding and recognizing creativity and (4) visible and able to be exposure to ideas across the rewarded. organization.  Performance measurement Autonomy Provides sense of ownership and control over tasks. provides autonomy through accountability.  Monitoring progress signals Resources Signals to the team that its work is important. to managers when resources are needed. Pressures Includes workload pressure and  Performance measurement challenge. communicates priorities, relieving negative pressures.  Performance measurement offers targets and benchmarks and enhances challenge. Sends mixed signals, detracting  Performance measurement Organizational impediments from the creative processes; can be negatively includes formal management perceived, providing structures, conflict and extrinsic motivation that conservatism. undermines the intrinsic motivation that fosters creativity. Empowerment (Kirkman & Rosen, 1999) Potency Collective belief of the team that it can be effective.

Meaningfulness

Team members’ shared perception of how to value the task and whether it is worthwhile.

 Measurement used as feedback reinforces a team’s confidence through either confirming the team’s direction or signaling a necessary change.  Measurement places a value on an innovation.

Team Performance Measurement

269

Exhibit 1. (Continued )
Dimension Description Influence of Team Performance Measurement  Performance measurement provides autonomy through accountability.  Measurement systems communicate to the team the value of their work.

Autonomy

Degree of decision-making authority Perception of the value of their work to the organization.

Impact

environment that encourages the development and sharing of creative ideas. It implies that when teams have ownership over their processes and when they receive support and reinforcement, the probability that they will generate creative ideas is greater. Empowerment Inherent in the discussion of creative processes and innovation is the idea of empowering teams to develop new mechanisms and make decisions concerning their process. Bowen and Lawler (1992) assert that employees are empowered by providing information and resources that enable them to make appropriate decisions. Instructing teams to improve their process or solve a problem, and then not providing the encouragement or resources to do so, sends the conflicting message that the team is not really empowered at all. This will certainly reduce motivation to fully participate in the creative process. Kirkman and Rosen (1999) describe four dimensions of team empowerment: potency, meaningfulness, autonomy and impact (Exhibit 1). Potency is the collective belief of a team that it can be effective (Guzzo, Yost, Campbell, & Shea, 1993). Team potency is influenced by the team’s perceptions of available resources and the skills and abilities of its team members. Meaningfulness at the team level is a shared perception of how to value the team’s task and whether the team has a worthwhile impact on the organization. Autonomy is the degree of decision-making authority a team exercises in its task environment. Team autonomy reflects freedom and independence in choosing actions as a team. The last dimension is impact, which measures the team’s perception about the value of its contributions to the organization. Teams can assess their impact by gathering feedback from outside the team (e.g., from customers). These four dimensions of empowerment are distinct, yet they are interrelated and mutually reinforcing (Kirkman & Rosen, 1999).

270

FRANCES KENNEDY AND LYDIA SCHLEIFER

Management accounting information and performance measurement are key to providing the information base teams need to succeed (Scott & Tiessen, 1999). Performance measures have a strong influence over the four dimensions of empowerment described by Kirkman and Rosen. As the accomplishments of teams are tracked and recognized, team efforts are reinforced, giving them a stronger sense of potency or confidence that they can continue to meet their goals. The manner in which the organization measures the effects of team innovations helps the team and others in the organization to value the team’s contribution. If the team perceives the impact of their efforts, they experience an increase in their feeling of worth and, consequently, derive more satisfaction and are intrinsically motivated to continue to innovate. Conversely, if the organization makes no effort to quantify or recognize team accomplishments, the team is not able to see where their contributions fit into operations nor do they experience any increase in either intrinsic or extrinsic motivation. The prior discussion of creativity/innovation and empowerment presents closely aligned constructs that are necessary to foster innovative environments. The two models, innovation and empowerment, contain overlapping and complimentary dimensions, and both are influenced by performance measurement. This discussion underscores the importance of developing a TPMS with great care and with consideration of the motivational and behavioral impact of those measures. Brennan and Dooley (2005, p. 1392) summarized the influence of organizations on employee empowerment: ‘‘From an organizational perspective, barriers to creativity include: intolerance of differences, overly rational thinking, inappropriate incentives and excessive bureaucracy. Meanwhile, employee participation and empowerment can contribute significantly to the increase and ultimately the level of organization innovation.’’ The next section proposes a TPMS that strives to provide both process and output measures while encouraging an environment that will empower teams and be conducive to the development of innovations.

Team Performance Measurement System Current performance measurement systems use a mix of financial and operational measures to guide organizational goals. Therefore, the team measurement system should contain a mix of measures that fully reflect strategy if an organization is to achieve its long-term objectives. Along with traditional financial measures, the system should include nonfinancial metrics that are

Team Performance Measurement

271

specific to the team’s purpose (e.g., reduction in scrap rate, decrease in customer complaints, reduction in cycle time), and are critical for teams to focus their activities and monitor their progress towards organizational goals. The TPMS described below was developed with input from representatives of five diverse companies: one bank services organization, one petrochemical company, two defense contractors and a commercial product manufacturer. This team met using conference calls at regular two-week intervals with the purpose of developing a team measurement system dynamic enough to use in each of their various industries and that satisfied their information needs. Once the system was developed, each company allowed access to teams to further consider these measures from the team perspective. Subsequently, participating teams summarized their recent projects using the proposed measures. The examples used in this paper originate from these teams and their own ex post application of the TPMS. The TPMS contains four measurement categories: financial, operational, effectiveness and innovation measures. When viewed in concert, these measures provide a broad perspective of team activity, progress and successes. In addition to the range of measurement categories, the TPMS provides a mechanism for consolidation that facilitates assessment information at the facility level. Exhibit 2 illustrates the TPMS for one team (called The Wheelies Team) that completed three projects over a two-year period. This example is used to illustrate the TPMS in the subsequent discussion of each measurement category and to demonstrate how this team’s performance is consolidated as part of the facility performance. Projects and data reflect real team innovations.

Components of the Measurement System Financial Measures It usually becomes difficult for companies to quantify in financial terms the impact of team performance. When teams tackle a problem or a process, they normally utilize various tools in their analysis. Flowcharting, storyboarding, fishbone diagrams and brainstorming are some examples of activity analysis techniques used by teams. Regardless of the tools used, the process of outlining the tasks performed in an activity for the purpose of making improvements is commonly referred to as activity-based management (ABM) (Ansari et al., 2004). ABM provides the team with a common understanding of all the tasks and resources included in an activity.

272

FRANCES KENNEDY AND LYDIA SCHLEIFER

Exhibit 2.

Illustration: Project and Team Effectiveness Indexes.
PANEL A Project Effectiveness Index (PEI) Calculation The Wheelies Team (Warehouse) Project No. 1

Incremental savings Labor: (1/2 h/shift  3 shifts  30 forklifts  325 days  $20/h) Extended battery life: (30 forklifts  $5,000 ¼ $150,000; $150,000/3 years) – ($150,000/4 years) Total incremental savings Incremental costs Equipment: Purchase and installation of 30 monitors Total incremental costs Net annualized savings Project effectiveness index (PEI) ($305,000 savings/$54,000 costs) PANEL B Team Effectiveness Index The Wheelies Team Project No. 1 No. 2 No. 3 Total Annualized Savings $305,000 $300,000 $26,000 $631,000 PANEL C Grassroots Facility Multiple Teams’ Performance Summary Team The Wheelies Sweepers Shippers Number crunchers Total Number of Innovations 3 2 0 1 6 Net Project Savings $537,000 49,000 0 16,000 $602,000 Annualized Costs $54,000 $40,000 $– $94,000

$292,500 12,500 $305,000 $54,000 $54,000 $251,000 5.6

PEI 5.6 7.5 20.0 –

TEI – – – 6.7

Overall Index 6.7 9.2 0.0 20.0 7.0

Note: The TEI (Panel B) and overall index (Panel C) are recalculated using the total annualized savings and costs

Team Performance Measurement

273

This process analysis helps teams to identify unnecessary or redundant steps, thereby identifying wasted resources. Activity-based costing (ABC) techniques facilitate the assignment of dollars to tasks and activities involved in a process and helps teams quantify, or dollarize, resources saved in a process change. Exhibit 2 (Panel A) calculates the net annualized savings for The Wheelies Team’s project no. 1. The Wheelies Team is a work team of forklift drivers responsible for collecting finished product from manufacturing, storing it in the warehouse and subsequently retrieving it for shipment. Each forklift driver was responsible for keeping the battery in his lift charged. It had become customary for the forklift drivers to congregate in the battery charging area at the beginning of the shift in order to be charged up and, therefore, ready to move product during the rest of the shift. The Wheelies Team noted that drivers’ average wait time was between 20 and 40 minutes due to limited charging resources, and it was often necessary to work overtime to complete shipping requirements. After collecting data and analyzing the charging process, the team recommended that battery monitors be installed on each lift and required that drivers bring the lift in for charging only when the monitor showed the charge had dropped below a specified level. During their investigation, the team also discovered that charging the batteries prematurely had shortened the total life. Four-year batteries were being replaced in three years. It was anticipated that the new procedure of charging batteries only when necessary would extend the useful life of the batteries to at least four years. The annualized savings for the project, including savings in labor and in battery life, totaled $305,000. Incremental costs included the purchase and installation of monitors on all forklifts. Exhibit 2 (Panel A) details the calculations. The net annualized savings for this project is $251,000. Why use annualized savings and costs? At one time it would have made sense to use discounted cash flow techniques and project savings over the life of the process change. In the current competitive environment, however, change is continuous and an innovation this year becomes next year’s status quo. Shortening the horizon to one year and eliminating the need for using discounting methods increases the validity of the savings estimate and simplifies the measure enabling team members to calculate with a minimum of training. Using financial measures in a balanced team measurement system is important because it is like speaking a universal language. Everyone can determine the relevance and importance of results expressed in terms of savings, costs and net savings. Furthermore, financial measures can always

274

FRANCES KENNEDY AND LYDIA SCHLEIFER

be accumulated over time and added across entities, teams or projects. The key that enables teams to quantify the impact of their projects has a solid foundation in basic ABM concepts. Appropriate training in ABM concepts provides the resources necessary to perform the analysis. This is an opportunity for management accounting to become involved by supporting teams with education and in providing necessary information (e.g., labor rates, material prices). Operational Measures As previously discussed, teams are created for a specific purpose, such as improving receiving report accuracy, can easily track their progress with the absolute number of errors or as a percentage of reporting errors to total reports. This clearly communicates the progress made toward the team’s goals. Operational measures are an important aspect of a balanced approach to team measurement because such measures may be closest to how team members are used to formulating and thinking about their tasks. Exhibit 2 (Panel A) uses the number of labor hours saved and 33% increase in battery life to estimate their project savings. Problems can arise with regard to operational measures in that teams may have a narrow vision for themselves if they were formed for a particular purpose. Improvements made to boost one metric may be at the expense of another. However, once they have experienced success within their stated purpose, teams can be encouraged to brainstorm or think outside of the box and formulate stretch goals and new innovations or new missions for themselves. Effectiveness Indices The magnitude of project savings in the preceding financial measure is one way of quantifying contributions. However, it does not provide assurance that the project is the best use of operating dollars. Ansari et al. (2004) discuss the benefits of viewing costs from different perspectives. Another way to frame the estimated results of an innovation is to calculate a simple index. A ratio represents a measure of productivity in that it quantifies the number of outputs produced to the physical inputs consumed and can be a useful supplement to other financial and nonfinancial information (Kaplan, 1983). Ratios also have the benefit of scaling for volume and increase the opportunity for comparisons across time and units. Indeed, ratios are meaningless except in comparison across reporting units or periods (Banker, Datar, & Kaplan, 1989). Similar ratios are used in other analyses. For example, a present value factor (Horngren, Datar, & Foster, 2003) is often

Team Performance Measurement

275

used to evaluate capital investment opportunities. A project effectiveness index (PEI) is calculated by dividing annualized savings by annualized costs. Exhibit 2 (Panel B) illustrates this calculation. An index of 5.6 indicates, very simply, that every dollar expended to maintain a solution is estimated to yield $5.60. Interviews with teams and managers revealed that many teams implement continuous improvement innovations that do not require an investment in funds nor do they require any ongoing cost to maintain – perhaps merely a change in the process. This is exactly what companies desire – continuous improvement without additional investment – a company’s optimal solution! As a result, however, their innovations were not documented because they did not need to request funding and no system was in place to capture these savings. Since the PEI cannot be calculated without a denominator, a review of projects’ PEIs in the pilot study that did have a cost ran as high as 16.9. Considering these results, a default PEI of 20.0 was assigned to projects not requiring any cost to implement and maintain. This is illustrated in Exhibit 2 (Panel B) for project innovation no. 3. Teams can also use the PEI to compare alternative solutions, as well as to communicate their successes and innovations to managers. It is simple to use and understood by team members in all different types of teams. A target hurdle index could also be incorporated into goals and objectives. The PEI calculates the benefits of one project. The same method can be applied to the team. Considering that each innovation could take as little as one week or as long as 18 months, it is important to consolidate all the innovations for a team for a given time period. Consensus among participating companies during development was to use a rolling two years as an appropriate time frame for calculating the team effectiveness index (TEI). Exhibit 2 (Panel B) illustrates how the TEI is calculated for The Wheelies Team. Annualized project savings for all projects implemented during the time period is divided by the total annualized project costs, resulting in an effectiveness index of 6.7. This recalculates a TEI to reflect all team innovations. The TEI can be used to compare dissimilar teams and may contribute to reward and recognition programs. Innovation Two problems arise with respect to managing innovation. The first is that teams have varying abilities and talents that influence their capacity to innovate. For example, a team with an engineer or two has greater potential for recommending a leap innovation that will result in large savings. Another team consisting of machine operators may not have the same potential

276

FRANCES KENNEDY AND LYDIA SCHLEIFER

for large improvements, but can initiate several smaller incremental innovations. For this reason, the magnitude of savings (though certainly important) from a particular innovation is not necessarily the best metric to capture both efforts and impact. Monitoring the frequency of innovations shows a team that is meeting regularly and consistently making changes – however small – to their process. Motivating this consistent activity indicates a culture that positively encourages innovative thinking. The second problem with managing innovations concerns the team that has a major success – and then stops and basks in the acknowledgment. This team may not jump in and work on another idea very quickly, either because the team members do not know how to top themselves or they simply want to ‘hold on’ to the last one. Interviews with company employees during the research phase of the project confirm this phenomenon. A human resource manager with a petrochemical company recounted the first success of a cross-functional team. The team estimated (and subsequently documented) annual savings of $1.2 million dollars by changing a formula mix. There was no cost to implement the change – only benefit. This team received accolades and toured other facilities within their company telling their story. When asked about subsequent projects the team initiated, the human resource manager answered, ‘‘Well y that project was almost 2 years ago. And since then, the team hasn’t made any new recommendations. And, actually, I am just not really sure they are making any progress either.’’ Tracking the number and frequency of innovations (recommendations for change) will help to identify and encourage teams of all types that are consistently developing and implementing new ideas. Therefore, measuring results and/or progress with regard to innovations is a major aspect of a balanced approach to measuring team performance. Exhibit 2 (Panel C) shows that The Wheelies Team implemented three innovations over the last two years, showing consistent progress and effort.

Assessing Team Performance on a Plant-wide Basis The TPMS offers two complimentary methods to aggregate the performance of multiple teams into a facility summary. One of these is a roll-up summary of the four categories of team measures, while the other provides a plant-wide financial impact summary. Exhibit 2 (Panel C) offers the summarized results of four teams over a two-year period. It shows how the number of innovations, net project savings and the TEI can be used to monitor dissimilar teams’ performance, as well as provide an overview of

Team Performance Measurement

277

Exhibit 3. Summary of Team Financial Impact on Facility.
PANEL A Grassroots Facility Annualized Team Support Costs Type of Cost Training (internal) Meeting time (all members) Celebration costs Team system managers Conferences, dues Materials, supplies Refreshments Travel Total support costs PANEL B Grassroots Facility Team Savings Net of Support Costs Team Net Team Savings Cost $25,000 38,000 13,000 20,000 12,000 5,500 1,500 10,000 $125,000

The Wheelies Sweepers Shippers Number crunchers Total incremental savings Less: Total support costs Net team savings retained Percent team savings retained

$537,000 49,000 0 16,000 $602,000 $125,000 $477,000 79.2%

the entire facility’s team activity. From the number of innovations, a manager can quickly discern which teams appear most active (The Wheelies and Sweepers) and which may be faltering and needing support (Shippers). Comparing the results of the Sweepers team and The Wheelies illustrates why having multiple measures is critical. The Wheelies saved $537,000, while the Sweepers saved $49,000. Looking at only this information implies that

278

FRANCES KENNEDY AND LYDIA SCHLEIFER

the Wheelies performed at a higher level. Further investigation, however, shows that the Sweepers’ TEI was 9.2 compared to the Wheelies 6.7, indicating that their project results in a higher return on operating dollars. In other words, both teams are engaging in innovative behavior and no single measure tells the whole story. The second method of assessing plant-wide performance provides a financial impact summary that incorporates both team innovations and the costs of supporting the team system. Up to this point, all four measures previously described (number of innovations, operational measures, financial measures and effectiveness indices) focus on what teams control – their own improvement recommendations, including both the savings and costs of those improvements. There are, however, significant costs incurred in a team-based organization that are necessary both to implement teams and to maintain an environment conducive to their success. These costs are not controlled by teams, but by managers. Such costs include team training, meeting time, supplies, celebration costs, consultants, team system managers, refreshment, travel related to team events and information systems. Generally, these types of costs are embedded in different cost centers within the accounting system. For example, the costs of meeting time for production teams may be captured in the manufacturing cost center while the costs of meeting facilities are a small part of a larger overhead allocation. Training and celebration costs may be captured in the human resource department. Upper management is interested in knowing not just whether the teams are performing to expectations but whether continued investment in the entire team system is warranted. Basically, management needs to know: Are they realizing a return on the total team support system costs? And is this return sufficient to justify future investment? Exhibit 3 (Panel A) presents a summary of annualized team support costs totaling $125,000. The majority of these costs can be captured through appropriate account codes in the general ledger, but may also be calculated using solid assumptions (e.g., number of hours of training times the number team members trained). Panel B of Exhibit 3 summarizes the net financial impact of team innovations net of team support costs. The result is $477,000 or 79.2% of team net savings retained after consideration of support costs.

Evaluation of the Team Performance Measurement System In any new measurement system, there are questions about whether the measurements are relevant to decisions and whether they encourage the

Team Performance Measurement

279

decisions needed to accomplish the established goals. With a team measurement system, identifying relevant and effective measures depends on knowing what the team is expected to accomplish. The main roles of team members relate to identifying problems, improving processes and recommending solutions (Kennedy, 2003). As part of this aspect of decisionmaking, teams have to decide what information is needed in order to consider their options. Having a well-defined system of metrics can aid the team in identifying and gathering the necessary information. Knowing what is being measured aids in knowing what is valued and in knowing what to target. Teams generate multiple solutions and identify the best or the most feasible solutions. Metrics can aid team members in identifying what is likely to be the best solutions. Simply providing a system, such as the TPMS that is easily accessed by teams, enables teams to communicate all their successes that previously went unnoticed. Theoretical Support for the TPMS The TPMS considers the various dimensions of innovation and empowerment. It can be used to encourage creativity by setting team goals and rewarding teams for their performance. TPMS enables the team to make decisions and encourage a sense of ownership over decisions by providing feedback to help change direction as needed. In addition, implementing the TPMS will send a message that the team’s work is important enough to allocate resources to it. Proper implementation should involve enough pressure to challenge and motivate team members but not so much pressure that it overwhelms and discourages teams from doing their best work. There is no guarantee that there will not be organizational impediments to the use of the TPMS, but an awareness of possible impediments should help to mitigate their impact. The use of TPMS can engender potency by leading a team to believe that it can be effective since it will show tangible evidence of the team’s performance and success. The use of TPMS will enable teams to observe how they impact the organization and therefore be able to assess the meaningfulness of their contribution to the organization. Autonomy would be enhanced by the use of TPMS since measuring the team’s performance would enable the team members to make independent decisions and take independent actions. Exhibit 4 summarizes how the measurement components impact the models of creativity and empowerment presented earlier. According to Beyerlein and Harris (2003, p. 135), team accountability means ‘‘how to enable others to act responsibly.’’ Management has to empower teams by giving them authority in areas for which they will be held accountable. Then those team members can be rewarded for living up to

280

FRANCES KENNEDY AND LYDIA SCHLEIFER

Exhibit 4.
Measurement Goal

Assessment of TPMS Impact on Innovation and Empowerment.
Impact of TPMS Components

Creativity (Amabile, Conti, Coon, Lazenby, & Herron, 1996)  Organizational  Innovations: Tracking the number of innovations encouragement encourages the generation of multiple new ideas.  Autonomy  Effectiveness indices provide tools for teams to assess and compare alternative solutions and enable decision-making.  Resources  All four components taken together provide managers with an overview of activity and progress for teams.  Pressures  Effectiveness indices help teams prioritize solutions and provide a vehicle for communicating the successes, relieving excess work pressures. These metrics can also be used to set stretch targets to challenge teams.  Organization impediments  Having a well-defined team performance measurement system eliminates mixed signals and subjective judgments concerning team performance. Empowerment (Kirkman & Rosen, 1999)  All four components provide feedback to the team  Potency signaling necessary change and reinforcing successes.  Meaningfulness  Operational and financial measures place values on the gains from the solution.  Autonomy  Effectiveness indices provide tools for teams to assess and compare alternative solutions and enable decision-making.  Impact  Financial and effectiveness indices assesses the organizational impact of the team’s contribution.

expectations. It is useful to be able to assess when teams have responded to expectations in an accountable way. The TPMS enables that assessment and does it in a way that allows teams to demonstrate several areas of competence and success. Practical Considerations for the TPMS Teaming and collaboration depend on acceptance by the participants. Beyerlein and Harris (2003) assert that inefficiency results when not all resources are being used and that the two most critical under-utilized resources are the hearts and minds of the individual employees and the synergies that emerge from effective collaboration. Of all the characteristics

Team Performance Measurement

281

needed in a good measurement system, the ones that probably do the most to win the hearts and minds of employees are the attributes of being understandable and easy to use. In the TPMS, tracking the number and frequency of innovations is a simple activity, and team members would be able to understand the value of such a measurement. Operational measures, like number of man-hours, can or should be more understandable the more closely they relate to an activity. Financial measures might be a little out of the realm of what many employees normally use, but with training and the use of straightforward analytical tools that show how time and resources are utilized, employees will begin to understand and use such financial measures. The PEI is simplicity itself in that it focuses on annual costs and savings without the need to discount future cash flows. In addition, Hinrichs and Ricke (2003, p. 295) point out that ‘‘employees need to gain a holistic view of their business.’’ Their work with a manufacturing company demonstrated to them that employees performed better when they understood the whole system, the business processes, the language of their business and the data about their business. According to Estrin and Kanter (1998), subjectivity can introduce bias that serves political purposes and leads to a perceived lack of fairness. The number and frequency of innovations is a measure for which any team member or manager would come up with the same information as any other member (Estrin & Kanter, 1998). Operational measures, depending on the particular ones chosen (e.g., number of errors in inventory reports), will also be more or less objective and verifiable. Financial measures and the PEI are informed estimates calculated with information from historical records and accounting information. It is a truism that more timely information is probably better than less timely information. According to Estrin and Kanter (1998), historical information may be acceptable for strategic decision making, but it is less likely to be useful for managerial control decisions and operational decisions. The information produced by the TPMS can be produced on a realtime basis. The technology is available and the environment of JIT and lean manufacturing has created a more spontaneous and flexible work culture.

LIMITATIONS
There are limitations to the use of TPMS. The system is designed to measure incremental improvements over current performance and may not be appropriate for all types of teams, such as new product development teams and

282

FRANCES KENNEDY AND LYDIA SCHLEIFER

management teams. Management teams are typically evaluated on the overall financial improvements of a facility and measured with traditional financial benchmarks, such as return on assets (ROA) and return on investment (ROI). Also, the use of TPMS may not be appropriate for certain processes, like major initiative implementations. A system like TPMS will be useful for teams that can establish a baseline of performance for a process or project. However, until there is enough information to establish a baseline, teams using TPMS may experience some uncertainty regarding whether they are implementing it correctly or gaining useful information from it. The implementation of a performance measurement system, like the one described here, could face the same challenges and pitfalls experienced in previous implementations of a balanced scorecard approach or, in general, any combination of financial and nonfinancial measures. A multidimensional team measure seems intuitively preferable. Caution should be exercised when implementing the TPMS as it may be appropriate for some types of teams, but may not be appropriate for other types of teams (e.g., product development teams). In the same way, the system described here may be appropriate for some contexts (like decision-making, training, developing strategic plans or evaluating the achievement of organizational objectives) and may not be appropriate for other contexts (like compensating managers and employees).

CONCLUSION
Strategic performance measures align strategic goals with operations through well-defined metrics. The primary goals of teams include developing and implementing innovative solutions. To achieve these goals, team members must be empowered to seek information and make decisions, as well as encouraged to try new ideas. Simons (1995) suggests that empowerment requires greater control. This implies that measures are needed to ensure that decisions are made to further organizational goals without creating unacceptable risk. A single metric, therefore, cannot adequately capture the efforts and results of teams. Current strategic performance measurement systems include a mix of relevant financial and nonfinancial measures. It follows that a comprehensive set of team measures should also include well-chosen metrics that are informative and help guide team and manager actions. The four categories of metrics in the TPMS framework together provide a broad view of team performance. Measuring the number of innovations

Team Performance Measurement

283

over time communicates the activity level of the team, while the impact of the changes is represented in the operational and financial measures. Operational (or nonfinancial) measures align team goals with strategy and effectively monitor progress towards those goals. Financial measures, represented as annualized net savings, estimate the magnitude of the process change. Finally, the effective use of operating dollars is expressed using effectiveness indices. The PEI helps teams to manage their solutions and compare alternatives, as well as give assurance to managers that additional operating dollars are used optimally. The TEI provides a mechanism to summarize all of a team’s activities, and can be used as a means of recognizing high performing teams. Used together, the TPMS offers a well-rounded perspective of team activity and innovation and provides a simple method of tying team recommendations to bottom-line impact.

REFERENCES
Alper, S., Tjosvold, D., & Law, K. S. (2000). Conflict management, efficacy, and performance in organizational teams. Personnel Psychology, 53(3), 625–642. Amabile, T. M., Conti, R., Coon, H., Lazenby, J., & Herron, M. (1996). Assessing the work environment for creativity. Academy of Management Journal, 39(5), 1154–1184. Ansari, S., Bell, J., Klammer, T. P., & Lawrence, C. (2004). Strategy and management accounting. In: S. Ansari (Ed.), Management accounting: A strategic focus (pp. 4–23). Boston, MA: Houghton Mifflin Company. Argrell, A., & Gustafson, R. (1996). Innovation and creativity in workgroups. In: M. A. West (Ed.), Handbook of workgroup psychology (pp. 317–334). Chichester, England: Wiley. Atkinson, A. A., Balakrishnan, R., Booth, P., Cote, J. M., Groot, T., & Malmi, T. (1997). New directions in management accounting research. Journal of Management Accounting Research, 9, 79–108. Banker, R. D., Datar, S. M., & Kaplan, R. S. (1989). Productivity measurement and management accounting. Journal of Accounting Auditing and Finance, 4(4), 528–554. Banker, R. D., Potter, G., & Schroeder, R. D. (1993). Reporting manufacturing performance measures to workers: An empirical study. Journal of Management Accounting Research, 5, 33–55. Bennis, W. G., & Nanus, B. (1985). Leaders: The strategy for taking charge. New York: Harper and Row. Bessant, J., Lamming, R., Noke, H., & Phillips, W. (2005). Managing innovation beyond the steady state. Technovation, 25, 1366–1376. Beyerlein, M., & Harris, C. (2003). Guiding the journey to collaborative work systems: A strategic design workbook. San Francisco, CA: Jossey-Bass/Pfeiffer. Birnberg, J. G. (1999). Management accounting practice and research as we end the twentieth century. Advances in Management Accounting, 8, 1–26.

284

FRANCES KENNEDY AND LYDIA SCHLEIFER

Bowen, D. E., & Lawler, E. E. (1992). The empowerment of service workers: What, why, how and when. Sloan Management Review, 33(3), 31–40. Brannick, M. T., & Prince, C. (1997). An overview of team performance measurement. In: M. T. Brannick, E. Salas & C. Prince (Eds), Team performance assessment and measurement: Theory, methods, and applications (pp. 3–16). Mahwah, NJ: Lawrence Erlbaum Associates. Brennan, A., & Dooley, L. (2005). Networked creativity: A structured management framework for stimulating innovation. Technovation, 25, 1388–1399. Cua, K. O., McKone, K. E., & Schroeder, R. G. (2001). Relationships between implementation of TQM, JIT, and TPM and manufacturing performance. Journal of Operations Management, 19(6), 675–694. Damanpour, F. (1991). Organizational innovation: A meta-analysis of effects of determinants and moderators. Academy of Management Journal, 34(3), 555–590. Drach-Zahavy, A., & Somech, A. (2001). Understanding team innovation: The role of team processes and structures. Group Dynamics: Theory, Research, and Practice, 5(2), 111–123. Drake, A. R., Haka, S. F., & Ravenscroft, S. P. (1998). Incentive effects on innovation, interaction and productivity in group environments. Advance in Management Accounting, 6, 93–112. Epstein, M. J., & Birchard, B. (2000). Counting what counts: Turning corporate accountability to competitive advantage. New York: Perseus Books. Estrin, T. L., & Kanter, J. (1998). Accounting for throughput time. Advances in Management Accounting, 6, 55–74. Ghalayini, A. M., Noble, J. S., & Crowe, T. J. (1997). An integrated dynamic performance measurement system for improving manufacturing competitiveness. International Journal of Production Economics, 48, 207–225. Guzzo, R., Yost, P., Campbell, R., & Shea, G. (1993). Potency in groups: Articulating a construct. British Journal of Social Psychology, 32, 87–106. Hinrichs, G., & Ricke, K. (2003). Gaining commitment to high performance work systems: John Deere case study. In: M. M. Beyerlein (Ed.), The collaborative work systems fieldbook: Strategies, tools, and techniques (pp. 287–306). San Francisco, CA: Jossey-Bass. Horngren, C. T., Datar, S. M., & Foster, G. (2003). Cost accounting: A managerial emphasis. Upper Saddle River, NJ: Prentice Hall. Kalagnanam, S. S., & Lindsay, R. M. (1998). The use of organic models of control in JIT firms: Generalizing Woodward’s findings to modern manufacturing practices. Accounting, Organizations and Society, 24(1), 1–30. Kaplan, R. S. (1983). Measuring manufacturing performance: A new challenge for managerial accounting research. The Accounting Review, 58(4), 686–705. Kennedy, F. (2003). Managing a team-based organization: A proposed strategic model. In Team-Based Organizing, 9, 91–111: Elsevier Science Ltd. Kirkman, B., & Rosen, B. (1999). Beyond self-management: Antecedents and consequences of team empowerment. Academy of Management Journal, 42(1), 58–74. Lawler, E. E., Mohrman, S. A., & Benson, G. (2001). Organizing for high performance. San Francisco, CA: Jossey-Bass. Lind, J. (2001). Control in world class manufacturing: A longitudinal case study. Management Accounting Research, 12, 41–74. Maskell, B., & Baggaley, B. (2004). Practical lean accounting. New York: Productivity Press.

Team Performance Measurement

285

McClelland, D. C. (1975). Power: The inner experience. New York: Irvington Publishers. McDonough, E. F. (2000). Investigation of factors contributing to the success of crossfunctional teams. Journal of Product Innovation Management, 17(3), 221–235. McNair, C. J., & Carr, L. P. (1994). Responsibility redefined: Changing concepts of accountingbased control. Advances in Management Accounting, 3, 85–117. Merchant, K. A. (1985). Control in business organizations. Cambridge, MA: Ballinger Publishing Company. Meyer, C. (1994). How the right measures help teams excel. Harvard Business Review, May–June, 95–103. Nanni, A. J., Dixon, J. R., & Vollman, T. E. (1990). Strategic control and performance measurement. Journal of Cost Management, Summer, 33–42. Schroeder, R., Scudder, G. D., & Elm, D. R. (1989). Innovation in manufacturing. Journal of Operations Management, 8(1), 1–15. Scott, T. W., & Tiessen, P. (1999). Performance measurement and managerial teams. Accounting, Organizations and Society, 24(3), 263–285. Sim, K. L., & Carey, J. A. (2003). Organizational control and work team empowerment: An empirical analysis. Advances in Management Accounting, 11, 109–141. Simons, R. (1995). Levers of control. Boston, MA: Harvard Business School Press. Spreitzer, G. M. (1996). Social structural characteristics of psychological empowerment. Academy of Management Journal, 39(2), 483–504. Tetlock, P. E. (1985). Accountability: The neglected social context of judgment and choice. Research in Organizational Behavior, 7, 297–332. Thomas, K. W., & Velthouse, B. A. (1990). Cognitive elements of empowerment: An ‘‘interpretive’’ model of intrinsic task motivation. Academy of Management Review, 15(4), 666–682.

AN EXPERIMENT OF GROUP ASSOCIATION, FIRM PERFORMANCE, AND DECISION DISSEMINATION INFLUENCES ON COMPENSATION
Arron Scott Fleming and Reza Barkhi
ABSTRACT
Reports citing excessive CEO compensation continue to make the news with evidence of peer relationships between the CEO and the compensation committee often the center of debate. The compensation committee of the board of directors determines CEO pay and is often comprises CEOs from other companies as well as non-CEOs such as academic, exgovernment, and professional individuals. This study examines the influence of the psychological factor of social comparison over accounting performance measures in a compensation experiment with 176 subjects. The results of this study are consistent with social comparison theory in that CEO director-subjects award greater pay and shield the compensation of the CEO when firm accounting performance is below average. Additionally, we find shielding is mitigated when subjects are informed that the decision of the amount of compensation awarded will be revealed to the public.

Advances in Management Accounting, Volume 16, 287–309 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16010-X

287

288

ARRON SCOTT FLEMING AND REZA BARKHI

INTRODUCTION
The board of directors provides guidance and oversight to management and acts as the key body for representing shareholders and investors. Proper oversight or governance by a board is a cornerstone element to our capital markets. This governance is often conducted within sub-committees of the board, such as the compensation committee. The task of determining the compensation of the CEO falls to this committee and it represents a significant fiduciary duty to the board, shareholders, and investors alike. One of the nation’s largest pension funds expressed the significance of this committee and the compensation process as a critically important and highly visible responsibility of the board of directors of a corporation. In a real sense, it represents a window through which the effectiveness of the board may be viewed (TIAA-CREF, 2002). Our primary research interest is in compensation committee ineffectiveness, where the decision-making outcome may not be in the best interest of the board or shareholders. We build from associative findings of compensation shielding by the compensation committee for unfavorably performing CEOs (Dechow, Huson, & Sloan, 1994; Gaver & Gaver, 1998; Duru, Iyengar, & Thevaranjan, 2002; Adut, Cready, & Lopez, 2003). Shielding occurs when the compensation committee minimizes reductions in executive compensation in the face of reduced firm performance. The compensation committee effectively limits the downward exposure of compensation to the executive in times of reduced performance. In this study, we test whether subjects role-playing as CEO directors on the compensation committee shield or protect the pay level of the chief executive officer when firm performance is below the industry average. Further, we test to see if potential publicity of CEO director’s decisions mitigate this shielding effect. Our experimental findings provide evidence for shielding by showing CEO director-subjects award greater compensation than non-CEO director-subjects when firm performance is below the industry average. Additionally, we find potential publicity surrounding individual decisions by CEO director-subjects reduces the shielding effect. In conducting this research, we expand the causal understanding of the influence of connections to peer groups within individual decision making. This extends the current body of literature in that it examines individual decision factors found within the executive compensation setting process through an experimental methodology. Motivation for continued compensation research stems from the relative importance of the topic in the business, investment, and political community. Disclosures regarding executive pay, such as the NYSE chief

CEO Compensation

289

executive’s pay package of roughly $140 million and subsequent resignation in 2003, highlight the repercussions and agency costs of governance process breakdowns. While the primary media focus was on the magnitude of compensation, much less attention was applied to the board of directors and the make-up of the compensation committee that awarded such a package. In the NYSE case, most of the committee members have titles of president, CEO, or chairman. Given the excessive CEO pay package and a lack of linkage to pay and firm performance, political and social pressures appear to have forced a change in the governance and compensation setting structure of the NYSE. The direct result is a change in board of director and compensation committee membership, and a possible return of excess-awarded compensation.1 The potential cause of the high pay package may be attributed to the nature and composition of compensation committee within the board of directors where alliances and interactions may compromise rational decisionmaking (Perel, 2003). The board of directors and the compensation committee is often comprises CEOs of other companies, academicians, retired military or government officials, and professional directors. It is the coterie of CEO directors within the compensation committee that may affect the compensation setting process, thus representing an agency problem in managerial incentives between the owner’s of the firm and those in control of the firm (Fama, 1980). Our research attempts to experimentally determine if CEO directors look out for their own, particularly when firm performance is below average. Agency problems, where management elevates their personal interests over the interests of the shareholders (Fama, 1980), result in various form of agency costs within an organization. CEO compensation setting processes are no exception, and mechanisms or structures that unnecessarily elevate CEO compensation are agency costs. In this area of concern, researchers have examined the board using inside or outside director categorization (O’Reilly, Main, & Crystal, 1988; Daily, Johnson, Ellstrand, & Dalton, 1998; Newman & Mozes, 1999; Bhagat & Black, 2002). Additionally, though, a contributing factor relating to CEO compensation may be the number of outside directors who are also CEOs. Nell Minow, editor of the Corporate Library, indicates that the best predictor of CEO overpay is the number of CEOs on a compensation committee (Burns, 2003, p. R6). While the boardroom is comprised of inside management such as the domicile CEO and outside directors, it is the outside director who is also a CEO that identifies most with the domicile CEO. This identification or social comparison to another individual or group (Festinger, 1954) forms the basis for

290

ARRON SCOTT FLEMING AND REZA BARKHI

agency costs within the compensation setting process. The objective of this study is to experimentally examine the aspect of social comparison as an agency cost. This is accomplished by studying subjects role-playing as CEO directors in the compensation setting process. Ideally, the compensation committee considers firm performance when setting CEO compensation, but this may not always be the case. While the Wall Street Journal/Mercer Human Resource Consulting (2003) notes a pronounced positive relationship between CEO annual pay and performance for 2001 and 2002, the Economic Research Institute found executive compensation grew faster than firm revenues in 2002.2 This occurred during a time when stock prices continued to decline, suggesting that compensation and performance does not always run in parallel. As the information and details of compensation packages become public, the compensation to performance incongruity has led shareholders to more closely scrutinize the CEO compensation award decision of the boards and file resolutions with the SEC. According to the Investor Responsibility Research Center, shareholder resolutions filed with the SEC in 2003 aimed at curbing CEO compensation have risen 200% over the previous year – General Electric’s CEO Jeff Immelt was subjected to 26 compensation-related resolutions (Ulick, 2003b). Even CEOs who meet or exceed expectations, such as Jeff Immelt’s predecessor Jack Welch, face investor criticism when pay and retirement packages become public. Welch returned significant portions of his post-employment compensation perquisites ($2.5 million/ year) when details of the retirement package became public during divorce proceedings in 2002 (Naughton, 2002). Publicity regarding excessive compensation carries negative consequences for both CEOs and directors. In the case of NYSE, not only was excess compensation ordered returned, but also the CEO was ousted along with certain board members supporting the pay package. Taken together, this suggests that public scrutiny of pay packages may by increasing and may have an impact on compensation awards. We conjecture that CEO peers on the compensation committee positively affect CEO pay. Further, we conjecture that publicity of excessive pay negatively affects CEO pay. It is the interplay of the number of CEO members on the compensation committee and the publicity of their decision of the pay package that is the focus of this study. In this paper we report the results of a 2 Â 2 Â 2 between-subjects experimental study. The three factors as illustrated in Fig. 1 are: group association of director type (CEO director-subjects versus non-CEO director-subjects), firm performance (above or below industry average), and compensation decision dissemination (public or private).

CEO Compensation
Model of Factor Association To Compensation

291

Association
+ -

(CEO, non-CEO Director)

Performance
+ -

Compensation

(above, below average)

Decision Dissemination
+

(pubic, private)

Fig. 1.

Model of Factor Association to Compensation.

The remaining paper is organized as follows: Section 2 provides a brief literature review, develops the hypotheses, and explains the model; Section 3 explains the methodology; Section 4 presents the results; and Section 5 discusses the implications, limitations, and direction for future research.

HYPOTHESES DEVELOPMENT
As a proxy for shareholders and acting on their behalf, the board of directors monitors, hires, fires, and guides the direction of the professional managers within the firm. The compensation committee, a sub-group to the board, determines the compensation of the CEO. The significance of executive compensation is emphasized in the following statement: The governance of the executive compensation process is a critically important and highly visible responsibility of the board of directors of a corporation. In a real sense, it represents a window through which the effectiveness of the board may be viewed (TIAA-CREF, 2002). Although the board is purported to represent the shareholders, agency problems with the CEO can become an issue. Top management may elect expropriation of wealth as opposed to competition once having gained

292

ARRON SCOTT FLEMING AND REZA BARKHI

control of a board (Fama, 1980). Overt agency problems are manifest through financial fraud, large perquisites, and excessive compensation, but we cannot discount the possibility of agency problems through a more subtle control by the CEO over the board or compensation committee via peer ascendancy. While the board of directors represents the shareholders, CEO directors more closely resemble the CEO from a social and professional standpoint. This peer association can create an effect of control or influence consistent with the representatives in the group. Research on groups has given attention to the natural formation of groups, uniformity within groups, and a normalization of behavior (Greenberg, Solomon, & Pyszczynski, 1997; Baumeister & Leary, 1995; Festinger, 1950). Similar observations appear in business contexts. Corporations are hierarchical while the board of directors is a collegial group working toward consensus (Bainbridge, 2002). In choosing an outside successor, the board tends to pick someone demographically similar to their own profiles (Zajac & Westphal, 1996), and the compensation committee is influenced by the demographic similarities to the CEO (Young & Buchholtz, 2002). Social capital (social status and network ties) of the CEO is associated to higher compensation (Belliveau, O’Reilly, & Wade, 1996) as is the compensation level of the outside director on the compensation committee (O’Reilly et al., 1988). These results can be explained by Social Comparison Theory. The theory suggests that individuals make comparisons to those they perceive as similar and associate with those having similar characteristics (Festinger, 1954). Examples of associations include status, position, and wealth. Hence, CEO directors associate more with the CEO, thus their compensation award decision is likely to be biased and positively influenced. We hypothesize that subjects role-playing as a CEO director will award greater pay than subjects roleplaying as non-CEO directors when evaluating a CEO and awarding compensation. We suggest the following hypothesis: H1. CEO director-subjects will award greater compensation than nonCEO director-subjects. From prior research we expect compensation to be positively associated to firm performance3 (Sloan, 1993; Natarajan, 1996; Gaver & Gaver, 1998; Duru & Iyengar, 1999; Tosi, Werner, Katz, & Gomez-Mejia, 2000; Sheikholeslami, 2001; Lambert & Larcker, 1987). In the absence of all other factors, we expect performance to be positively associated with pay. Hence, we present the following hypothesis:

CEO Compensation

293

H2. Director-subjects will award greater compensation when firm performance is above industry average as compared to below industry average. Previous archival research has also shown that the makeup of the compensation committee mediates the pay to performance ratio for underperforming firms. That is, CEOs of firms that are poor or unfavorable performers may have their compensation levels or package protected by the compensation committee (Dechow et al., 1994; Gaver & Gaver, 1998; Duru et al., 2002; Adut et al., 2003). Further, director type affects the extent of such compensation shielding (Newman & Mozes, 1999). We conjecture the CEO director, a more closely associated member of the coterie, exacerbates this protection or shielding of CEO compensation. Through this social comparison or group association, the CEO director-subject will award greater compensation than the non-CEO director-subject when performance is below average. We propose the following hypothesis: H3. CEO director-subjects will award greater compensation than nonCEO director-subjects when performance is below the industry average. In addition to group association and firm performance, we study the impact of individual decision dissemination. Research has shown decisions of groups involve greater levels of risk-taking than individuals and can exacerbate or escalate decision trends (e.g., Stoner, 1961; Argote, Seabright, & Dyer, 1986; Whyte, 1993). Given the CEO director-subject is a member of a group or coterie within the compensation committee, the publication of the decision makes salient the individuality of the subject and breaks the mental association to the group. Without individual decision publicity, individual decision makers may be prone to the more risk-taking attitude of a group. If the individual decision is public, though, then the dynamics of the group association and decision escalation is less likely to materialize. Therefore, the publicity of the individual decision can mitigate the compensation shielding effects of the group and lessen the agency costs. Specifically, we hypothesize that the CEO director-subject will award lower levels of compensation when the individual subject’s decision is noted to be made public as compared to being kept private. Hence, we present the following hypothesis: H4. When performance is below the industry average, CEO directorsubjects will award lower compensation when the individual decision is noted to be made public as compared to being kept private.

294

ARRON SCOTT FLEMING AND REZA BARKHI

Fig. 1 summarizes the main components of the model that describe the study reported in this paper. The association of the director type, the performance of the firm, and the decision dissemination potential are individually and jointly affecting compensation levels determined by the compensation committee member.

METHOD
Sample We conducted this research at a large American university. We used subjects enrolled in the second of two principle accounting courses. A total of 115 men and 61 women participated in the study. Subjects were on average 20.4 years old (SD ¼ 1.3) with an average of 0.7 years (SD ¼ 1.3) of full-time work experience (see Table 1). The subjects were primarily first- and secondyear undergraduate students enrolled in the college of business. Although the use of student subjects in behavioral accounting research is not unusual, we acknowledge that it is not ideal but may be a practical solution given limited accessibility to CEO and board of director subjects. Following Sedor and Kadous (2004), student subject are appropriate since this study employs theories centered on characteristics not dependent on the professional population (Peecher & Solomon, 2001; Libby, Bloomfield, & Nelson, 2002). Evidence in student surrogate studies examining attitudes show there is a divergence between students and other subjects, while in studies examining decision making, considerable similarities exist4 (Ashton & Kramer, 1980). Since our experiment is centered on decision making and the subjects were immersed in their roles,5 we believe the results obtained from using student subjects provide strong internal validity and reasonable external validity as applied to decision makers composing boards of directors in general. While it may be argued that undergraduate subjects are unlikely surrogates for Table 1. Subject Descriptives.
Mean Age Full-time work experience 20.4 0.7 Males Gender 115 Standard Deviation 1.3 1.3 Females 61

CEO Compensation

295

CEO directors, it may also be argued that given the pragmatic distance of reality from such subject to such population, any evidence obtained in such a weak manipulation indicates the presence of a stronger bona fide effect. Variables The dependent measure in our study is compensation. To reduce potential subject anchoring confounds and biases we elected not to use dollars but rather a non-bounded artificial currency we labeled as ‘‘Qwert’’. This follows from accounting and economic literature where researchers in lieu of directly employing dollars use points (e.g., Kachelmeier & Shehata, 1997) or other artificial denominations (e.g., Friedman, 1967; Forsythe, Palfrey, & Plott, 1982; Plott & Sunder, 1982; Forsythe & Lundholm, 1990). Independent variables include director type, performance, and decision dissemination. Within each vignette subjects were assigned to the role as a CEO director or non-CEO director on the compensation committee. The performance of the subject firm was either above or below the industry average. This was indicated primarily in two ways: (1) it was shown numerically as a comparative growth rate and through earnings per share data, and (2) through a verbal statement stating the company’s operating margins and net income levels were above or below the industry average.6 Lastly, within each vignette, the compensation decision for each director was noted as either a private and confidential decision or one that would be made public. Procedures Student subjects were given a one-page overview on corporate governance (Appendix A). The subjects were asked to participate in an in-class experiment for the following week. Participation was voluntary and those who chose to participate received either extra-credit or a waiver of one homework grade, equal to 3 points out of 550 total points for the class. Subjects were given a pre-numbered cover sheet and demographic questionnaire (Appendix B) to complete. After signing the cover sheet we collected and gave them to the instructor for credit purposes. Subjects at this point were tracked only via the pre-numbered forms. The pre-numbered demographic forms were collected and the subjects were introduced to an individual who played the role of the CEO. The subjects were given an overview of the experiment and told that they were role-playing as compensation committee members of the board of directors and would determine the compensation of the CEO who was being evaluated. The subjects were told that roughly half were role-playing as CEO directors from other companies, one-quarter were role-playing as retired

296

ARRON SCOTT FLEMING AND REZA BARKHI

public servants, and one-quarter were role-playing as academicians. Subjects were then given name tags with their title as either CEOs of fictitious companies (randomized three-letter abbreviated names), or the titles of either retired senator or business school dean. A pre-numbered vignette (see sample in Appendix C) was given to each subject. This is a 2 Â 2 Â 2 study with subjects assigned to one of eight variations. Within each vignette is information describing the compensation committee and their role, their compensation in their own profession, the accounting performance of the fictitious CEO’s company as compared to the industry average, and the industry average compensation level. Additionally, each subject was informed within the vignette whether or not their compensation decision is to be kept private and confidential or made public. From this information the subject determined the compensation of the fictitious CEO. On completion of the task the vignettes were collected and pre-numbered post-experimental surveys were distributed (Appendix D), completed, and collected.

RESULTS
A 2  2  2 (director  performance  decision) ANOVA is presented in Table 2 with cell descriptives in Table 3. Overall results indicate significant main effects for director type and performance (decision was not significant), no significant two-way interactions, but a significant three-way interaction. Hypothesis H1, CEO director-subjects will award greater compensation than non-CEO director-subjects, is supported. The mean award for a CEO director-subject is 71.1886 and the mean for a non-CEO director-subject is 68.9494 (p ¼ 0.007). Hypothesis H2, director-subjects will award greater compensation when firm performance is above average as compared to below industry average, is also supported. The above industry average performance compensation mean is 74.3250 versus the below industry average of 67.1049 (po0.001). Hence, the experimental results suggest that both director type and performance have significant influence on CEO compensation. Results are represented graphically in Figs. 2 and 3 for hypotheses H1 and H2, respectively. Fig. 4 illustrates the combined results to help visually compare CEO award as a result of performance (above or below average) and as decided by board member type (CEO or non-CEO). The lack of two-way interactions is not surprising given that we expect director differences only when performance is below industry average.

CEO Compensation

297

Table 2.
Source

2 Â 2 Â 2 ANOVA Table: Tests of Between-Subjects Effects.
Type III Sum of Squares 2865.52Ã 796452.61 373.90 2112.18 1.94 40.16 0.79 6.93 238.24 8307.53 870522.04 11173.04 df Mean Square 409.36 796452.61 373.90 2112.18 1.94 40.16 0.79 6.93 238.24 49.75 F Significance

Corrected model Intercept Directora Performb Decisionc Director  perform Director  decision Perform  decision Director  perform  decision Error Total Corrected total a 7 1 1 1 1 1 1 1 1 167 175 174

8.23 16010.49 7.52 42.46 0.04 0.81 0.02 0.14 4.79

0.000 0.000 0.007 0.000 0.844 0.370 0.900 0.709 0.030

à R squared ¼ 0.256 (adjusted R squared ¼ 0.225).

‘‘Director’’ is CEO director/non-CEO director categorization. ‘‘Perform’’ is firm performance above/below the industry average. c ‘‘Decision’’ is public/private individual decision dissemination. b Therefore, to test hypothesis H3 we conducted a 2  2 (director type  decision) ANOVA restricted by below average performance (Table 4). Results indicate significant main effects for director type (CEO or non-CEO) but not for decision (public or private). Therefore, hypothesis H3, CEO director-subjects will award greater compensation (mean ¼ 68.88) than non-CEO director-subjects (mean ¼ 65.07) when performance is below the industry average, is supported (F ¼ 9.480, p ¼ 0.003). This finding indicates a significant interaction of director type and performance in the negative performance domain.7 This result further supports the compensationshielding phenomenon. The significant three-way interaction (F ¼ 4.789, p ¼ 0.030) leads to the investigation of hypothesis H4, that when performance is below the industry average the CEO director-subject will award lower compensation when the individual decision is noted to be made public as compared to being kept private. Table 5 presents the results of a one-way ANOVA to testing hypothesis H4. When we restrict the data to the below industry average performance and CEO directors, the results moderately support H4. The mean compensation awarded by CEO director-subjects when performance is below the industry average and the decision is private is 70.0280 versus 67.9233 when the decision is made public (F ¼ 2.023, one-tail p-value ¼ 0.081). This indicates that CEO directors no longer shield the chief executive’s

298

ARRON SCOTT FLEMING AND REZA BARKHI

Table 3. Cell Descriptives.
Value Label Director 0 1 Perform 0 1 Decision 0 1 Private Public 92 83 Below average growth Above average growth 103 72 CEO Non-CEO 88 87 N

Director CEO

Perform Below average growth

Decision Private Public Total Private Public Total Private Public Total Private Public Total Private Public Total Private Public Total Private Public Total Private Public Total Private Public Total

Mean 70.03 67.92 68.88 74.27 76.22 75.04 71.91 70.43 71.19 63.67 66.16 65.07 74.80 71.58 73.72 69.82 67.92 68.95 67.12 67.09 67.10 74.57 73.90 74.32 70.85 69.22 70.08

Standard Deviation 3.08 6.84 5.52 3.39 3.67 3.58 3.83 7.14 5.71 7.64 7.87 7.79 4.45 16.00 9.81 8.21 11.25 9.72 6.42 7.33 6.91 3.99 11.62 7.60 6.50 9.38 8.01

N 25 30 55 20 13 33 45 43 88 21 27 48 26 13 39 47 40 87 46 57 103 46 26 72 92 83 175

Above average growth

Total

Non-CEO

Below average growth

Above average growth

Total

Total

Below average growth

Above average growth

Total

Variables are defined in Table 2.

CEO Compensation
72

299

70

71.19

68.94

p = 0.007

68 CEO Subject Type non-CEO

Fig. 2.

Compensation Award by Subject Type.

72

70

74.32

67.10

p < 0.001

68

Above
Firm Performance

Below

Fig. 3. Compensation Award by Performance Realm.

76 72 68 64

75.04

73.72 Above 68.88

“Above” avg. = 74.32

Above

Below

“Below” avg. = 67.10 65.07 Below

CEO

non-CEO Subject Type

Fig. 4.

Compensation Award by Subject Type and Performance Realm.

300

ARRON SCOTT FLEMING AND REZA BARKHI

Table 4.
Source

2 Â 2 ANOVA Table: Tests of Between-Subjects Effects.
Type III Sum of Squares 505.91Ã 453869.92 417.51 0.97 133.99 4360.13 468681.38 4866.05 df 3 1 1 1 1 99 103 102 Mean Square 168.64 453869.92 417.51 0.97 133.99 44.04 F 3.83 10305.45 9.48 0.02 3.04 Significance 0.012 0.000 0.003 0.882 0.084

Corrected model Intercept Director Decision Director  decision Error Total Corrected total

Variables are defined in Table 2. Ã R squared ¼ 0.104 (adjusted R squared ¼ 0.077).

Table 5. One-Way ANOVA: Public versus Private Comparison for CEO Directors in the Below Industry Average Domain.
Sum of Squares Between groups Within groups Total 60.40 1582.50 1642.91 df 1 53 54 Mean Square 60.40 29.86 F 2.02 Significanceà 0.081

à Reported p-value is one-tail given the directional nature of H4.

71

69

70.03

67.92

p = 0.081

6
Fig. 5.

Private Decision Type

Public

CEO Director-Subject Compensation Award by Decision Type.

CEO Compensation

301

compensation when the decision is noted as public. A graphical representation is shown in Fig. 5.

CONCLUSION
Agency problems manifest themselves in various forms within an organization, and the executive compensation setting process is no exception. Our study experimentally tests the influence of three factors: (1) director-subjects (CEO and non-CEO), (2) accounting performance (below average or above average), and (3) decision dissemination (public or private). A contribution of our study is that it shows how these three factors elevate awarded compensation. We find results consistent with previous compensation shielding literature. A limitation to this study is the subject pool. While our convenient sample provided internal validity, these subjects are not perfect substitutes for the business leaders, and thus this potentially limits our external validity. Future studies should build on this research to address the limitations of this study and examine the anchoring effects and other environmental factors that have been empirically shown to influence CEO compensation. Our experimental results indicate that director type influences the compensation setting process, particularly when firm results are below the industry average. CEO director-subjects award greater compensation in general and award significantly greater compensation as compared to nonCEO director-subjects when performance is below average. A further influencing factor presents itself when the individual decision of the director is noted as being kept either private or made public. In our study we find evidence of further shielding by CEO director-subjects when performance is below average and when the decisions are private, as compared to when the decisions are public. Thus, while director type mediates the influence of performance on pay, decision dissemination also mitigates the relation between performance and pay.

NOTES
1. Kelly, Craig, and Dugan (2003). Further, on October 19, 2006 the New York state Supreme Court in Manhattan ordered Mr. Grasso to forfeit a portion of his pay package (Lucchetti & Lublin, 2006). 2. Cash compensation increases of 5.9% versus revenue increases of 0.89% in 2002 (Ulick, 2003a).

302

ARRON SCOTT FLEMING AND REZA BARKHI

3. Iyengar (2003) finds higher compensation levels for perennially lossmaking firms with negative retained earnings. Given the uniqueness of their sample, we feel the results do not apply to the general population in regards to performance to pay association. Firms within the sample may represent companies willing to hire and compensate executives at higher levels for potential turnaround performance. 4. In a later, yet unpublished experiment (Fleming) that utilized 71 undergraduate, 63 graduate accounting, and 95 executive MBA subjects in the determination of compensation, results shows that the subjects were qualitatively similar in their decision outcomes (F ¼ 0.975, p ¼ 0.379). 5. We believe the students were fully immersed in their roles and confirm this through post-experimental questions on scenario role and group association. All subjects correctly answered the question of role (CEO director versus non-CEO director). Additionally, from post-experimental survey questions asking subjects to rate their association to a particular group (CEO, non-CEO, and board of director groups) where 1 is weak and 10 is strong, we find that CEO director-subjects significantly associated to the CEO group (mean ¼ 8.75; F ¼ 158.969; po0.001); nonCEO director-subjects significantly associated to the non-CEO group (mean ¼ 8.20; F ¼ 149.765; po0.001); and both CEO and non-CEO director-subjects associated similarly to the board of directors (CEO mean ¼ 7.69; non-CEO mean ¼ 7.29; F ¼ 1.495; p ¼ 0.223). 6. Executive compensation is often based on targets and goals set forth by the board of directors via a specific employment contract. In this research we do not make available explicit targets or goals but rather we provide the subjects with performance measures that indicate the subjects firm’s performance to a benchmark, such as previous year’s performance or performance to the industry. 7. As a further analysis, the same test was performed in the above industry average domain without significant results. The CEO director mean of 75.0364 versus the non-CEO director mean of 73.7231 proved non-significant (F ¼ 1.020, p ¼ 0.277).

ACKNOWLEDGMENTS
The authors wish to acknowledge the assistance of Richard Brooks, John Brozovsky, William Kerler III, John Maher, and Christian Schaupp. In addition, we wish to acknowledge the constructive comments of John Lee and two anonymous reviewers.

REFERENCES
Adut, D., Cready, W. H., & Lopez, T. J. (2003). Restructuring charges and CEO cash compensation: A reexamination. The Accounting Review, 78(1), 169–192.

CEO Compensation

303

Argote, L., Seabright, M. A., & Dyer, L. (1986). Individual versus group use of base-rate and individuating information. Organizational Behavior & Human Decision Processes, 38, 65–75. Ashton, R. H., & Kramer, S. S. (1980). Students as surrogates in behavioral accounting research: Some evidence. Journal of Accounting Research, 18(1), 1–15. Bainbridge, S. M. (2002). Why a board? Group decisionmaking in corporate governance. Vanderbilt Law Review, 55(1), 1–55. Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497–529. Belliveau, M. A., O’Reilly, C. A., & Wade, J. B. (1996). Social capital at the top: Effects of social similarity and status on CEO compensation. Academy of Management Journal, 39(6), 1568–1593. Bhagat, S., & Black, B. (2002). The non-correlation between board independence and long-term firm performance. Journal of Corporation Law, 27(2), 231–273. Burns, J. (2003). Everything you wanted to know about corporate governanceybut didn’t know to ask. Wall Street Journal, 10-27-2003, R6. Daily, C. M., Johnson, J. L., Ellstrand, A. E., & Dalton, D. R. (1998). Compensation Committee Composition as a determinant of CEO compensation. Academy of Management Journal, 41(2), 209–220. Dechow, P. M., Huson, M. R., & Sloan, R. G. (1994). The effect of restructuring charges on executives’ cash compensation. The Accounting Review, 69(1), 138–156. Duru, A., & Iyengar, R. J. (1999). Linking CEO pay to firm performance: Empirical evidence from the Electric Utility Industry. Managerial Finance, 25(9), 21–33. Duru, A., Iyengar, R. J., & Thevaranjan, A. (2002). The shielding of CEO compensation from the effects of strategic expenditures. Contemporary Accounting Research, 19(2), 175–193. Fama, E. F. (1980). Agency problems and the theory of the firm. Journal of Political Economy, 88(2), 288. Festinger, L. (1950). Laboratory experiments: The role of group belongingness. Miller, James Grier. Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140. Forsythe, R., & Lundholm, R. (1990). Information aggregation in an experimental market. Econometrica, 58(2), 309–347. Forsythe, R., Palfrey, T. R., & Plott, C. R. (1982). Asset valuation in an experimental market. Econometrica, 50(3), 537–568. Friedman, J. W. (1967). An experimental study of cooperative duopoly. Econometrica, 35, 379–397. Gaver, J. J., & Gaver, K. M. (1998). The relation between nonrecurring accounting transactions and CEO cash compensation. The Accounting Review, 73(2), 235–253. Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror management theory of self-esteem and cultural worldviews: Empirical assessments and conceptual refinements. Advances in Experimental Social Psychology, 29, 61–139. Iyengar, R. J. (2003). Executive compensation in perennially loss-making firms. Finance India, 17(1), 199–214. Kachelmeier, S. J., & Shehata, M. (1997). Internal auditing and voluntary cooperation in firms: A cross-cultural experiment. The Accounting Review, 72(3), 407–431.

304

ARRON SCOTT FLEMING AND REZA BARKHI

Kelly, K., Craig, S., & Dugan, L. J. (2003, November 19). Big board members appear to embrace governance changes. The Wall Street Journal, C1. Lambert, R. A., & Larcker, D. F. (1987). An analysis of the use of accounting and market measures of performance in executive compensation contracts. Journal of Accounting Research, 25(Suppl.), 85–125. Libby, R., Bloomfield, R., & Nelson, M. W. (2002). Experimental research in financial accounting. Accounting, Organizations and Society, 27(8), 775–810. Lucchetti, A., & Lublin, J. S. (2006). Grasso is ordered to repay millions in compensation; Judge faluts ex-NYSE chief over disclosure to board; Ruling will be appealed. The Wall Street Journal, A1, October 20. Naughton, K. (2002). The perk wars. Newsweek.MSNBC.com, September 23, 2. Natarajan, R. (1996). Stewardship value of earnings components: Additional evidence on the determinants of executive compensation. Accounting Review, 71(1), 1–22. Newman, H. A., & Mozes, H. A. (1999). Does the composition of the compensation committee influence CEO compensation practices? Financial Management, 28(3), 41–53. O’Reilly, C. A., Main, B. G., & Crystal, G. S. (1988). CEO compensation as tournament and social comparison: A tale of two theories. Administrative Science Quarterly, 33(2), 257–274. Peecher, M. E., & Solomon, I. (2001). Theory and experimentation in studies of audit judgments and decisions: Avoiding common research traps. International Journal of Auditing, 5(3), 193–203. Perel, M. (2003). An ethical perspective on CEO compensation. Journal of Business Ethics, 48(4), 381–391. Plott, C. R., & Sunder, S. (1982). Efficiency of experimental security markets with insider information: An application of rational expectations models. Journal of Political Economy, 90, 663–698. Sedor, L. M., & Kadous, K. (2004). The efficacy of third-party consultation in preventing managerial escalation of commitment: The role of mental representations. Contemporary Accounting Research, 21(1), 55–82. Sheikholeslami, M. (2001). EVA, MVA, and CEO compensation. American Business Review, 19(1), 13–17. Sloan, R. G. (1993). Accounting earnings and top executive compensation. Journal of Accounting and Economics, 16(1–3), 55–100. Stoner, J. A. F. (1961). A comparison of individual and group decisions involving risk. Unpublished Master’s Thesis, Massachusetts Institute of Technology. TIAA-CREF. (2002). TIAA-CREF Policy Statement on Corporate Governance. Retrieved 3 December 2002. Tosi, H. L., Werner, S., Katz, J. P., & Gomez-Mejia, L. R. (2000). How much does performance matter? A meta-analysis of CEO pay studies. Journal of Management, 26(2), 301–339. Ulick, J. (2003a, March 25). CEO salaries, bonues keep rising. CNN/Money, 1–4. Ulick, J. (2003b, April 22). Anger rising over CEO pay. CNN/Money, 1–3. Wall Street Journal/Mercer Human Resource Consulting (2003). 2002 CEO Compensation Survey and Trends, 1–28. Whyte, G. (1993). Escalating commitment in individual and group decision making: A prospect theory approach. Organizational Behavior & Human Decision Processes, 54(3), 430–455.

CEO Compensation

305

Young, M. N., & Buchholtz, A. K. (2002). Firm performance and CEO pay: Relational demography as a moderator. Journal of Managerial Issues, 14(3), 296–313. Zajac, E. J., & Westphal, J. D. (1996). Who shall succeed? How CEO/Board preferences and power affect the choice of new CEOs. Academy of Management Journal, 39(1), 64–90.

APPENDIX A. CORPORATE GOVERNANCE OVERVIEW
Corporate Governance A Very Short Overview What is Corporate Governance?
‘‘Corporate governance is a hefty-sounding phrase that really just means oversight of a company’s management – making sure the business is run well and investors are treated fairly’’ (Burns, 2003).

Publicly traded companies are those whose stock is traded in a public forum, usually over the New York Stock Exchange (NYSE), the American Exchange (AMEX), National Association of Securities Dealers Automated Quotation System (NASDAQ), or other regional exchanges such as Philadelphia or San Francisco. As such, any company can literally have thousands of ‘‘owners’’. It is difficult for a company to be managed simultaneously by potentially thousands of different owners; therefore, the owners or stockholders elect a board of directors as their representatives. The board sizes vary with an average of 9–11 members. The board of directors hires management, such as the chief executive officer (CEO), chief financial officer (CFO), and other vice-presidents to run the company – the board oversees their activities. This oversight is often conducted within a sub-committee of the board, such as the audit committee, the compensation committee, or the nominating committee. As an example, selected members of the board may be on the compensation committee – their job is to determine the compensation for the CEO and is a significant fiduciary duty as a board member.
‘‘The board’s most important job is hiring, firing, and setting compensation for a company’s chief executive, who runs the company day-to-day’’ (Burns, 2003).

306

ARRON SCOTT FLEMING AND REZA BARKHI

The membership of the board is often comprised of the CEO or other firm insiders, CEO’s from other industries, bankers, retired politicians, academicians, and professional directors (often representing mutual or retirement funds). Note: Although often on the board of directors, the CEO cannot be a member of his own compensation committee. Board member provide not only oversight but also expertise and advice, often meeting three to five times a year in addition to the (usual) legally required once a year meeting.

APPENDIX B. DEMOGRAPHIC QUESTIONNAIRE
Name: ___________________________________________ When you turn in the survey to your instructor, remove this first page. It will be used to record your participation. Instructions: Please read the following and the attached. The board of directors is the governing body for a publicly held corporation. The board represents the shareholders, decides the major investment and social policies for a company, and hires and determines the compensation of the executive management. In this case you serve on the board of directors of PUTT Company – this is not your full-time employment – please read the details of the attached for a description of your occupation. One of your duties while serving on the board of directors is to serve on the compensation committee. TIAA-CREF, a major retirement pension fund in the United States, describes the importance of this function in their 2002 policy statement as such: ‘‘The governance of the executive compensation process is a critically important and highly visible responsibility of the board of directors of a corporation. In a real sense, it represents a window through which the effectiveness of the board may be viewed’’ TIAA-CREF (2002). Please answer all the questions on the following page and the question at the bottom of the case. Thank you for your time and assistance. Demographics______________

CEO Compensation

307

This section captures basic information related to you, the survey participant. 1. Age in years ______________ 2. Gender a. Male b. Female 3. Are you currently a student? a. Yes b. No 4. Number of years of full-time employment (excludes time as a student) ______________ 5. Number of Accounting courses completed (no ranges please) ______________ 6. Number of Finance courses completed (no ranges please) ______________ 7. Number of Management courses completed (no ranges please) ______________

APPENDIX C. SAMPLE VIGNETTE
You are the Chief Executive Officer (CEO) of FHN Corporation. You are also on the board of directors of PUTT Company, an industrial company that manufactures golf equipment. Within the board of directors, one committee for which you serve is the compensation committee. Your company, FHN Corporation, does not perform any services for PUTT, nor does it anticipate doing so. You serve on the compensation committee of the board of directors for PUTT as an independent director. Serving with you on the compensation committee are five other members: Three are CEO’s of other companies, one is a dean of a business school, and one is a retired U.S. senator. Your compensation as CEO of FHN Corporation is in QWERTs, a nondenominational monetary unit. You currently make 70 Qwerts as CEO. The industry of PUTT has a CEO average compensation of 70 QWERTs.

308

ARRON SCOTT FLEMING AND REZA BARKHI

The average compensation of all CEOs in all industries is 70 QWERTs.  The golf equipment industry grew 10% this past year.  PUTT Company grew at a 6% pace.  Last year’s earnings per share for PUTT was $1.00. This year’s earnings per share for PUTT is $1.06.  PUTT’s closest competitor’s earnings per share numbers are $1.10 for the current year. The size of PUTT is comparable to the industry average, as is the total sales volume, and the number of shares of common stock outstanding.  PUTT’s operating margins and net income levels are below industry averages. The compensation committee of the board of PUTT Company performs an annual compensation review of the chief executive officer. Your task as a member of the compensation committee is to set the compensation level of the CEO in QWERTS. The compensation level you decide will be kept private and confidential. Based on the information provided, what compensation in QWERTs will you award to the CEO of PUTT Company? ___________________Qwerts

CEO Compensation

309

APPENDIX D. POST-EXPERIMENTAL SURVEY
Post Case Questions 1. Describe your role in this case a. A chief executive officer (CEO) serving on a board of directors b. A dean of a business school serving on a board of directors c. A retired senator serving on a board of directors 2. On a scale of 1 to 7 rate PUTT Company’s performance 1 2 below 3. 3 4 5 average 6 7 above

In this case, is your compensation decision a. private and confidential b. public and disclosed

Based on your role in the case, rate your association or connection to the following group(s) by circling a number. 4. Chief Executive Officers (CEOs) 1 2 3 4 5 weak 6 7 8 9 10 strong

5.

Non-CEOs (e.g., retired senators or business school deans) 1 2 3 4 5 6 7 8 9 weak Board of Directors 1 2 3 weak

10 strong

6.

4

5

6

7

8

9

10 strong

7.

Please rate the difficulty in determining the compensation level. 1 2 3 4 5 6 7 difficult easy Please rate the difficulty in assessing the information provided 1 2 3 4 5 6 7 difficult easy

8.

A NOTE ON THE READABILITY OF PROFESSIONAL MATERIALS FOR MANAGEMENT ACCOUNTANTS
Thomas J. Phillips Jr., Cynthia M. Daily and Michael S. Luehlfing
ABSTRACT
Recent changes in professional examinations have generated much debate concerning various issues. One specific debate relates to the consistency of readability levels before and after the changes. While no significant differences in examination readability were found with respect to consistency across the entire time horizon of the study, comparisons with respect to the readability of other professional materials generate questions on whether the exam is testing at an appropriate level and whether other materials such as those produced for continuing education are written at a level commensurate to practice.

Intuitively, the readability level at which professional materials for management accountants are written should reflect the readability level required daily in the business world. From time to time it seems appropriate to question whether such materials adequately prepare management accountants for the expectations placed on professionals. For example, ‘‘friendly’’ debates have emanated regarding the appropriateness of recent changes in
Advances in Management Accounting, Volume 16, 311–318 Copyright r 2007 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1474-7871/doi:10.1016/S1474-7871(07)16011-1

311

312

THOMAS J. PHILLIPS JR. ET AL.

professional examinations and if readability levels have remained consistent before and after these changes. The results of this study suggest that such a debate might rather focus on whether professional examinations test at appropriate levels of readability. Moreover, the broader issue might concern the extent that other professional materials assist management accountants in business.

BACKGROUND
There have been numerous studies on the readability of accounting literature including textbooks, authoritative pronouncements, audit reports, and financial statements (Courtis, 1986; Adelberg & Razek, 1984; Raabe, Stevens, & Stevens, 1984; Adelberg, 1982; James, Lewis, & Wallschutzky, 1981; Lewis & James, 1981; Pound, 1981). These studies used the cloze procedure, the Flesch Reading Ease Formula and the Gunning’s Fog Index in evaluating readability. The cloze procedure was one of the first methods developed to analyze readability (Taylor, 1953). In this procedure, words are deleted from selected passages of writing and the subject is asked to fill in the blanks. Replacement with the exact word implies that the reader comprehends the content of the text. Although the cloze procedure is widely used, it can be difficult to administer and is often criticized as being a better assessment of the reader’s abilities than readability. The Flesch Reading Ease Formula (Flesch, 1948) uses the average sentence length and the average number of syllables per word to calculate a score. The score can range from 0, for extremely difficult reading, to 100, for very easy reading. Gunning’s Fog Index (Gunning, 1952) is similar to Flesch’s Reading Ease formula. However, instead of counting syllables Gunning requires counting words of three or more syllables, referred to as ‘‘hard words.’’ The formula then relies on two calculations, average sentence length and percentage of words having three or more syllables, to find the grade level of the passage. To evaluate the readability level of the various exams, textbooks, professional pronouncements, rules and regulations included in the analysis, we choose to use three standard readability tests – the Flesch Reading Ease, Gunnings Fog Index (both described above), and the Flesch–Kincaid Grade Level formula. The Flesch–Kincaid Grade Level formula converts the Flesch Reading Ease score to a grade-level. Both the Flesch Reading Ease score and the Flesch–Kincaid Grade Level can be calculated using Microsoft Word. These tests were chosen because they are easy to use, objective, and their validity has been proven in earlier studies.

A Note on the Readability of Professional Materials

313

Criticism of the tests used here revolves around the fact that these procedures sometimes tend to inflate the calculated readability level when dealing with highly technical information. However, this is not a problem when comparing the readability of technically comparable passages, as in this case. While some argue that the cloze procedure more accurately reflects grade level, cloze is not feasible to use when determining readability involving large amounts of material. Before reporting the initial results of our analysis, we will first overview the exam format changes.

FORMAT CHANGES
The IMA instituted an exam format change in December 1990 (referred to hereinafter as the first change) and again in July 2004 (referred to hereinafter as the second change). While the number of exam parts was reduced from five to four with the first change in the exam, and the second change in the exam resulted in the same number of parts, albeit different, the body of knowledge to be tested essentially has not changed. For example, dropping ethics and taxes from the section titles did not eliminate those subjects from the exam. In essence, the first changes were intended primarily to be a rearrangement of topic coverage. However, the most recent change, while primarily reorganization of the content, does include some additional material. According to the IMA, Strategic Marketing has been added, as well as Strategic Planning with more emphasis on manufacturing paradigms and business process performance. Given that the content of the exam remained primarily intact after the first change, there seems to be no reason why these changes should instigate any change in the readability level of the exam. Although the exam remains primarily unchanged, it should be noted that there have been changes in the requirements for the exam. In the first change, the educational requirements were amended slightly, allowing college students to take the Certified Management Accountant (CMA) exam in their senior year. However, successful candidates are still required to possess a baccalaureate degree before they can use the CMA designation. The most recent set of exam changes included a new requirement before completion of the computerized exam itself. Candidates must complete the first three parts of the exam: Business Analysis, Management Accounting and Reporting, and Strategic Management (receiving immediate performance feedback on those sections) before attempting the fourth part, Business Applications. The subject matter in the first three parts is tested in the objective format, while Business Applications tests the same subject matter, but includes

314

THOMAS J. PHILLIPS JR. ET AL.

Table 1.
Exam before December 1990

Comparison of Exam Contents.
Exam from December 1990 to 2004 Economics, finance, and management Financial accounting and reporting Management reporting, analysis, and behavioral issues Decision analysis and information systems Exam after 2004

Managerial economics and business finance Organization and behavior, including ethical considerations Public reporting standards, auditing, and taxes Periodic reporting for internal and external purposes Decision analysis and information systems

Business analysis Management accounting and reporting Strategic management

Business applications

problems and essays. The CMA Board requires that the first three parts be successfully completed in order to demonstrate a degree of competence before allowing the candidate to apply their knowledge in the businessoriented situations of the fourth part (Table 1).

RESULTS
To evaluate the readability level of the various CMA exams included in the analysis, we selected all of the written problems and essay questions (nonobjective format) from each section of each CMA exam during the period June 1988–June 1992. This included five CMA exams before and five CMA exams after the 1990 exam format changes. The latest exam format change has resulted in a computer-based exam, with each candidate receiving a different version of the exam. Because the exam is not disclosed and candidates are required to keep the questions and answers confidential, actual exam questions are not available for comparison. Each section was then analyzed using the three previously addressed readability tests. The results of our analysis suggest that, while the format changed, the level of readability of the selected questions remained consistent (at the college graduate level), on a statistically significant basis, for all sections of all CMA exams for the entire time horizon of the analysis of the first change. While these results are laudable, the results do not speak to the appropriateness of the readability level of the CMA exam. Accordingly, we searched for benchmarks to gain insights regarding this issue. In this regard,

A Note on the Readability of Professional Materials

315

we noted that the CMA exam is written at a college graduate level versus the Wall Street Journal which is written at a college student level (Hart, 1993; Colossi, 1991). While this benchmark provides insight with respect to the business community, this benchmark is rather general in nature. To enhance comparability, the level of readability of the Certified Public Accountants (CPA) exam was analyzed for the same time horizon. To evaluate the readability level of the various CPA exams included in the analysis, we selected questions from each section of each CPA exam for the period May 1988–May 1992, using the same selection procedure as with the CMA exam. The CPA exam has also been revised since 1992 and is now a computer-based exam, with each candidate receiving a different version of the exam. Because the exam is not disclosed and candidates are required to keep the questions and answers confidential, exam questions are not available for comparison. Similarly to the analysis of the CMA exam, each section of the CPA exam (from the period May 1988 to May 1992) was analyzed using the three readability tests. The results disclose that the level of readability was statistically higher for the CMA exam than for the CPA exam (i.e., the readability level of the CPA exam was primarily at the college student level) (Table 2). To assess the significance of these results, we searched for a common benchmark for comparison. Since both exams draw heavily from the professional literature regarding financial accounting, Financial Accounting Standards Board (FASB) statements were selected as such a benchmark. We analyzed the readability level of the first 122 FASB statements, each one in its entirety, using the same three readability tests. The analysis was performed on FASB Statement Nos. 1–122. FASB Statement No. 1 was issued well before June 1988 and FASB Statement No. 122 was issued well after December 1992. For convenience, we discontinued our analysis after FASB Statement No. 122. We have no reason to believe that the results would have changed on a statistically significant basis had our analysis been extended to the most recent FASB statement. Table 2. Readability Differences between CMA and CPA Exams : Mann–Whitney U Probabilities.
Flesch–Kincaid Flesch Reading Ease Gunnings Fog Index Note: Significant probabilities in bold. .0000 .0000 .0000

316

THOMAS J. PHILLIPS JR. ET AL.

There were two significant results of this analysis. First, the results indicate that the level of readability is consistent (on a statistically significant basis) across all FASB statements before, during, and after the time horizon of the analysis. Second, the results show that the readability level of the FASB statements is at the doctoral level. Accordingly, the readability level of the FASB statements is considerably higher than the CPA exam, but only somewhat higher than the CMA exam.

DISCUSSION
This article presents the preliminary results of a long-term project. Since the underlying analysis relates to a very specific time horizon (i.e., five exams before and five exams after the date of the December 1990 CMA exam format change), this limits the generalizability of the results reported in this article to exams administered outside of this time horizon. However, we have no reason to believe that significant differences exist between the readability level of the exams included in our analysis and the exams excluded in our analysis. Since the readability analysis was centered on only the written problems and essay questions of the CMA exam, the results of the analysis may not be generalizable to the multiple choice questions. However, we have no reason to believe that significant differences exist between the readability level of the written problems/essay questions and the multiple choice questions. The results regarding the consistent readability level of the CMA exam are comforting. In contrast, the results of the remainder of the analysis are, at worst, less comforting, and at best, thought provoking. Given the practical nature of both the CMA exam and the CPA exam, we would have expected a somewhat lower readability level when compared with the technically oriented FASB statements. However, the question must be asked – how much lower? Financial statement footnotes, for instance, have been found previously to be written at the college graduate level. We have included a reference chart comparing readability grade levels of the CMA Examination with other materials. Readability levels in Table 3 are shown using Gunning’s Fog Index. The CMA exam, CPA exam, and FASB Statement scores were computed in this study. Other Fog Index approximations came from the referenced articles. The Fog scores for Sales Training Manuals (Kaminski & Clark, 1987) were derived using manuals in four industries, including insurance service, office products, steel products, and industrial cutting tools. The scores shown are

A Note on the Readability of Professional Materials

317

Table 3.
Benchmark

Readability Grade Levels using Gunning’s Fog Index.
Level 22 21 18 17 16 15 14 13 Source This study Colossi (1991) Worthington (1978) This study This study Colossi (1991) Colossi (1991) Kaminski and Clark (1987)

FASB Statements New York Times Financial Statement Footnotes CMA Examination CPA Examination Wall Street Journal MS Windows Manual Sales Training Manual

all above the high school level, with the lowest 13 (college freshman) and the highest 22 (Ph.D. level). While some might argue that language should not be a barrier to performance, others might argue that professional exams should be written at the level at which people function. Although the CMA exam’s readability is at a lower level than FASB statements, CMA questions are at a higher readability level than CPA questions. Accounting professionals may not spend time examining FASB statements as part of their daily routine, but a certain amount of time must be spent reading technical information. Furthermore, hardly anyone would argue with the accounting profession’s emphasis on the need to improve writing and other communication skills of graduating students to meet the complexity of today’s business environment. One testing expert argues that professional certification examinations should consider items such as written material and memos appropriate to expected job performance, state statutes and regulations, standard reference works required in practice, and other materials typical to the work environment (Plake, 1988). While the results of our analysis do not directly address these arguments, some might suggest that the results do provide support for the old adage that ‘‘professional exams evaluate the minimum level of acceptable competence.’’ In this regard, future research could focus on post-professional examination activities such as continuing education materials and examinations as well as specialty certification examinations and related materials.

ACKNOWLEDGEMENT
The authors would like to acknowledge their appreciation for the valuable assistance they received from the Institute of Certified Management

318

THOMAS J. PHILLIPS JR. ET AL.

Accountants, the American Institute of Certified Public Accountants, and the Financial Accounting Standards Board.

REFERENCES
Adelberg, A. H. (1982). An empirical evaluation of the communication of authoritative pronouncements in accounting. Accounting and Finance, 22(November), 73–94. Adelberg, A. H., & Razek, J. R. (1984). The cloze procedure: A methodology for determining the understandability of accounting textbooks. The Accounting Review, 59, 109–122. Colossi, D. (1991). Grade level reading. PC Sources, August, 70. Courtis, J. K. (1986). An investigation into annual report readability and corporate risk-return relationships. Accounting and Business Research, 16, 285–294. Flesch, R. (1948). A new readability yardstick. Journal of Applied Psychology, 32, 221–233. Gunning, R. (1952). The technique of clear writing. New York: McGraw-Hill. Hart, J. (1993). Writing to be heard. Editor & Publisher, November 6, 5–7. James, S., Lewis, A., & Wallschutzky, I. (1981). Fiscal fog: A comparison of the comprehensibility of tax literature in Australia and the United Kingdom. AustralianTax Review, 11(March), 26–36. Kaminski, P. F., & Clark, G. L. (1987). The readability of sales training manuals. Marketing Management, 16, 179–184. Lewis, A., & James, S. (1981). Understanding tax forms. Certified Accountant, 73(February), 48–52. Plake, B. S. (1988). Application of readability indices to multiple-choice items on certification and licensure examinations. Educational and Psychological Measurement, 48, 543–551. Pound, G. D. (1981). A note on audit report readability. Accounting and Finance, 21(May), 45–55. Raabe, W. A., Stevens, K. C., & Stevens, W. P. (1984). Tax textbook readability: An application of the cloze method. The Journal of the American Taxation Association, 6(December), 66–73. Taylor, W. L. (1953). Cloze procedure: A new tool for measuring readability. Journalism Quarterly, 30, 415–433. Worthington, J. S. (1978). Footnotes: Readability or liability. CPA Journal, 48, 27–32.