Armco, Inc.: Midwestern Steel Division

34. J. Roberts, The Modern Firm: Organizational Design for Performance and Growth (New York, Oxford University Press, 2004), p. 181; and D. L. Wenner and R. W. LeBer, “Managing for Shareholder Value From Top to Bottom,” Harvard Business Review, 67, no. 6 (November–December 1989), pp. 2–8. 35. S. Datar, S. L. Kulp and R. A. Lambert, “Balancing Performance Measures,” Journal of Accounting Research, 39, no. 1 (June 2001), pp. 75–92; G. A. Feltham and J. Xie, “Performance Measure Congruity and Diversity in Multi-Task Principal–Agent Relations,” The Accounting Review, 69, no. 3 (July 1994),

pp. 429–53; R. A. Lambert, “Contracting Theory and Accounting,” Journal of Accounting and Economics, 32, no. 1–3 (December 2001), pp. 3 –87; and G. Baker, “The Use of Performance Measures in Incentive Contracting,” American Economic Review, 90, no. 2 (2000), pp. 415–20. 36. D. Solomons, Divisional Performance: Measurement and Control (Homewood, IL: Richard D. Irwin, 1965), p. 83. 37. K. J. Arrow, “Control in Large Organizations,” in M. Schiff and A. Y. Lewin (eds), Behavioral Aspects of Accounting (Englewood Cliffs, NJ: Prentice Hall, 1974), p. 284.

Case Study

Armco, Inc.: Midwestern Steel Division
In January 1991 management of the Kansas City Works of Armco’s Midwestern Steel Division began implementing a new performance measurement system. Bob Nenni, Director of Finance for the Midwestern Steel Division, explained:
With our old system, our managers spent more time explaining why changes in costs were caused by problems with our accounting system than they did fixing the problems. The new performance measurement system is designed to give us better management focus on the things that are most important for them to worry about, earlier warning of problems, and improved commitment to achieve objectives.

In the summer of 1991 the new system was still being implemented and its design refined. But Bob Nenni believed that the new system would be successful at the Kansas City Works, and he hoped that its use would spread throughout Armco.

BACKGROUND OF ARMCO AND THE KANSAS CITY WORKS
Armco, Inc. was a producer of stainless, electrical, and carbon steels and steel products. Through joint

ventures the company also produced coated, high strength and low-carbon flat rolled steels and oil field machinery and equipment. In 1990 Armco was the sixth largest steel manufacturer in the United States with slightly over $1.7 billion in net sales, and operating profits of $77 million. Exhibit 1 shows a three-year history of Armco’s financial results. Armco’s Midwestern Steel Division generated $550 million in sales in 1990. (A division organization chart is shown in Exhibit 2.) Within the division, the Kansas City Works was by far the largest entity, accounting for approximately $250 million in sales. Like that of most of the firms in the US steel industry, business at the Kansas City Works had declined significantly in the last decade. Employment was down from 5,000 employees in 1980 to 1,000 in 1990.1 The Works had recorded significant losses in the decade of the 1980s, but it had been marginally profitable since 1988. The Kansas City Works produced two primary products: grinding media and carbon wire rod.
1

Over the same period, Armco Inc. decreased in size from 70,000 to 23,000 employees.

This case was prepared by Patrick Henry, Research Assistant, and Professor Kenneth A. Merchant.
Copyright © 1991 by Kenneth A. Merchant.

37

t

Chapter 2 · Results Controls

Grinding media were steel balls used for crushing ore in mining operations. Carbon wire rod was used to make shopping carts, bed springs, coat hangers, and other products. In 1990 the Kansas City Works sold 700,000 tons of steel: 200,000 tons of grinding media and 500,000 tons of rods. Armco was recognized as the leading supplier of grinding media products in the United States. Armco’s balls had proven themselves to be the most durable, and Armco received fewer customer complaints about its balls than did its competitors. Carbon wire rods, on the other hand, were basically a commodity product. Armco’s rod mill, which used relatively old technology, was not cost competitive, so rods were not a profitable product. But the rods did generate volume and helped cover some of the fixed costs of the plant. The Kansas City Works was not a low-cost manufacturer. Its union labor costs in Kansas City were higher than those of some of its nonunion competitors, particularly those located in the southeastern United States and non-US locations. And the Works had an inefficient plant infrastructure because the plant was designed to accommodate five times as many employees as were currently working there. Instead of being efficiently laid out, the buildings still being used were spread across a 900-acre plant site. Because of the plant’s cost disadvantage, the Work’s managers looked for ways to differentiate their products and to develop new higher-value products, and they had had some success in doing so. Each year approximately 10% of the shipments of the Kansas City Works were of new higher value, high carbon content products. All salaried employees in the Works were eligible for cash incentive awards based on a performance evaluation made by their immediate superior and, ultimately, Rob Cushman, the division president. The incentive award potentials ranged from approximately 5–30% of annual salary depending on the individual’s organization level. The performance evaluations were subjective but were based on, typically, three measures of performance applicable to the position. For example, Rob Cushman described the criteria he used for evaluating the performance of Charlie Bradshaw, the Works Manager, as being based approximately one third on plant safety, one third on hard production numbers (particularly productivity and quality), and one third on his

evaluation of Charlie’s “leadership” (i.e. “Do I hear good things and see good things going on?”).

THE MANUFACTURING PROCESS AT THE KANSAS CITY WORKS
The manufacturing process used for making both rods and grinding media included four basic steps. First, scrap steel was melted in the ladle arc furnaces. Second, the melted steel was poured into a continuous caster that produced solid bars 30 feet in length with a 7″ by 7″ cross-section. Third, the 19″ mill pressed the steel bars between two large cylindrical rollers to give them either a square or circular shape and with either 3″ or 4″ cross-sections.2 Finally, the bars were processed into finished rods or balls. The rod mill shop worked with square cross-section bars. It reduced the bars’ diameter to between 1/4″ and 5 /8″ and coiled them into 2,000-pound bundles for shipment to customers. The rods were further reduced in size (“cold drawn to wire”) at the customers’ facilities for use in their products. The grinding media shop worked with the circular cross-section bars produced by the 19″ mill. It formed them into spheres using a roll-forming machine. Finished balls ranged in size from 1″ to 5″ in diameter.

CRITICAL SUCCESS FACTORS IN THE WORKS A. The melt shop
The melt shop, which included the ladle arc furnace and the continuous caster, produced molten steel in 167-ton batches known as “heats.” The shop’s goal was to run three “turns” (shifts) a day, seven days a week, 50 weeks a year, excluding the eight hours a week used for preventive maintenance. The other two weeks of the year were used for extensive preventive maintenance and installation of new equipment. The melt shop could theoretically produce about 110 heats/week, but the best quarter it had ever achieved was an average of 99 heats/week. For a number of reasons, good performance in the melt shop was critical to the performance of the Kansas City Works as a whole. First, the melt shop was the “bottleneck” operation, so output from
2

The distance from center to center of the two pressing rolls was 19 inches. Hence, the name for the process.

38

Armco, Inc.: Midwestern Steel Division

this phase of manufacturing process determined the output of the plant as a whole. Second, the melt shop costs accounted for nearly 40% of the total steel conversion costs incurred in the plant. The largest expenditures in the melt shop were for labor, production materials of various types, and energy. Energy alone accounted for approximately 10% of the melt shop costs. Works managers were working toward computer control of energy, but in 1991 the melt shop manager still made most decisions about the heat used in the furnace, a major energy consumer. In 1988 Armco made an $8 million investment in a new ladle arc furnace that significantly changed the melting furnace technology used in the plant, and costs were declining as the melt shop managers learned how best to use the new technology. Third, the quality of raw steel produced by the melt shop was an important component in determining whether the finished products met the required specifications. Quality was affected by the grades of scrap steel and nonmetallic materials used in the process. Nonmetallic materials were consumable items added to batches to remove contaminants from the steel. Armco managers purchased a variety of grades of scrap steel and nonmetallic materials, and they used different proportions of scrap to nonmetallic materials depending on the grades of scrap and nonmetallics being used; lower grades of scrap typically contained more contaminants. Some of the production processes were standardized, with the addition of some nonmetallics done either by automated equipment or by production employees following standardized recipes. Other processes, however, required the manufacturing manager and his technical supervisors to exercise judgment.

grinding operations. Those tests had shown that Armco’s balls were more than competitive; they lasted up to 15% longer than did its closest competitor’s balls. The rolling areas were heavily capital intensive. Significant costs in the finishing areas were for labor, energy, maintenance, and yield losses.

C. Maintenance
Maintenance was also an important determinant of success in the Kansas City Works. The goal of maintenance was to maximize equipment uptime while controlling maintenance expenditures. Organizationally, the maintenance activities were divided into three groups. Teams of electrical and mechanical maintenance employees were assigned to each manufacturing cost center. A third group operated a centralized maintenance shop. The cost of maintenance was significant, as approximately 40% of the 700 hourly employees in the plant were maintenance workers.

THE OLD PERFORMANCE MEASUREMENT SYSTEM
The manufacturing areas of the Kansas City Works were divided into five responsibility centers: melting, casting, the 19″ mill, the rod mill department, and the grinding media department. Each responsibility center was comprised of one or more cost centers. Before changes were made in 1991 the performances of the cost center managers and their superiors in the plant were evaluated in terms of cost control and safety. The key cost performance measure was a summary measure called “Cost Above” which included the cost added per ton of steel at each production stage and for the entire plant. Cost Above and the items that comprised it were reported to the manufacturing managers on an Operating Statistics Report that was produced on approximately the 15th day following each month end. The Operating Statistics Reports provided a five-year history, monthly and year-to-date actuals, and monthly and year-to-date objectives and variances from objectives for each of the factors that determined total Cost Above for each cost center. Exhibit 3 shows a portion of the Operating Statistics Report for one cost center – the #2 melt shop.3
39

B. Rolling and Finishing
Personnel in the Rolling and Finishing areas were asked to make parts to specification while controlling yields and costs. Customer specifications for rods usually contained physical property requirements, such as for ductility and elasticity. One specification for balls required a two-story drop test. If the test ball cracked into parts on impact after being dropped from two stories, then the product was rejected on quality grounds. In addition, the lives of the balls were tested in Armco’s customers’ actual

Chapter 2 · Results Controls

(The entire report, printed on five computer pages, included detailed information about 46 separate expense categories.) The report also gave cost per net ton ($/NT) for many of the cost categories. The Total Cost Above/NT is shown in the next to last column of page 3 of Exhibit 3. The Operating Statistics Reports used the same accounting information that was used for financial reporting and inventory valuation purposes, so the figures included allocations of indirect manufacturing costs. For example, to provide smoother cost patterns, the charge for the two-week plant maintenance shutdown was spread over the 50 weeks of operations. These costs, which included labor and material, were shown on the Operating Statistics Report as “S-Order” costs. The operating managers had become accustomed to the Operating Statistics Report and in general they liked it. For example Gary Downey, the Melting Operations manager, said that he looked at 95% of the information presented in the report, although he acknowledged that some of the items were quite small in dollar value. Paul Phillips, the Rolling and Finishing Manager, liked having the monthly and annual trends and the information comparing actual costs with objectives. Paul felt that the Operating Statistics report was “the minimum amount of detail necessary.” He would have preferred to have the Operating Statistics Report on a weekly basis and in his hands on Monday morning because, for example, “If we see that fuel consumption is unusually high, we can go and look for the cause.” The accounting department also provided other reports showing the detail behind the figures for some of the cost elements on request. For example, one report showed the cost for nonmetallic materials broken down by the specific materials used.

keeping the old system going. On November 1, 1990 Rob Cushman was appointed as president of the Midwestern Steel Division, and Rob sponsored the implementation of a new performance measurement system. He allowed Bob Nenni to discontinue production of the Operating Statistics Report in January 1991 in order to implement the new system. Rob Cushman observed:
The old system wasn’t working. People were relying on something that was not adequate . . . Enough companies are using good performance measurements as building blocks to excellence. I don’t want to go against the grain. I want to give my managers the information they need. And I want to have good measures that tell us how we’ve done. I’m not using the performance measures as a threat. I’m trying to make it fun so that when we determine we’ve done well we can celebrate our successes . . . If this plant does everything right, we should be able to make $30 million per year. But we’re not doing it. This system is part of a spirit of change that has to happen. We will give people more responsibility . . . and more latitude to fail.

Bob Nenni added:
The cost part of our old performance measurement system was built for accountants. It was designed to produce financial statements, operating reports, and product cost reports. One system can’t do all these things well.

THE NEW PERFORMANCE MEASUREMENT SYSTEM A. The goals of the new system
Bob Nenni, the director of finance, had been working on a performance measurement system since 1989, but due to staff constraints he had been unable to design and implement the new system while
3

The #1 melt shop contained obsolete equipment and was no longer used.

The new system was designed both as a means of providing middle- and lower-level managers with a greater understanding of how their actions related to the implementation of the division’s business strategies and as an improved method for managers at all levels to assess the extent to which the desired results were being achieved. The vision and goals of the organization were to be translated into key success factors which would be disaggregated into department and individual objectives that would be compared with measures of actual results. The basic philosophy is illustrated in chart form in Exhibit 4. Rob Cushman and Bob Nenni thought that the new system promised two major improvements. First, the new system was designed so that managers would focus on the few key objectives that largely determine the success of the Kansas City Works and not get involved in the detail until a problem existed. As Bob Nenni observed:

40

Armco, Inc.: Midwestern Steel Division

When managers get too much data, they can easily get unfocused. The new system will cause them to focus on the five or six things that cause 80% of the costs, not the 40 that cause 100%.

Second, the new system was designed to provide an improved basis for evaluating operating managers and manufacturing supervisors. The system would include a balanced set of performance measures, including quality, schedule achievement, and safety, in addition to costs. And the cost reports would be improved because they would include only those costs deemed controllable by each individual operating manager. They would not be distorted by volume changes as in the old system.

B. The design of the new system
The new system design process began early in 1990. Rob Cushman, Charlie Bradshaw, Bob Nenni, Gil Smith (commercial director), and others defined 10 key performance measures for the Kansas City Works:
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. heats per week tons per man hour disabling injury index total quality index spending maintenance performance cash flow product mix inventory days on hand sales price minus cost of net metal.

Performance measure 1, heats per week, was only relevant to the melt shop. However, since the melt shop was the bottleneck operation, heats per week was a critical measure for the Works as a whole. Measures 2 through 6 were applicable to all manufacturing areas. Tons per man hour was a productivity measure. The disabling injury index was a safety measure. The total quality index was the product of three measures: physical yield, percentage of product meeting specification, and percentage on-time shipment. Spending was the accumulation of all expenses incurred by the people reporting directly to a manager. The maintenance performance measures had not yet been clearly

defined by the middle of 1991, but maintenance labor cost and material cost were being measured. Performance measures 7 through 10 were plantwide (not cost center) measures. Cash flow was measured monthly for the plant. Product mix was the percentage of high carbon products sold compared to low carbon. Inventory days on hand was tracked monthly. Accountability for inventory performance was shared among plant purchasing managers, manufacturing managers, and commercial managers. Sales price minus net metal, a measure of value added, was tracked monthly. The design group discussed the components of each performance area and the ways in which each measure could be disaggregated to guide performance at lower management levels. For example, the cascading of goals relating to total quality is illustrated in Exhibit 5. Total quality at the Works level was affected by the proportion of products meeting customer specifications, the yields, and the percentage of on-time shipments; and each of these indicators could be disaggregated further. The intent was to measure each of these areas of performance at the lowest relevant level of the organization. One of the most significant changes was the elimination of the Cost Above measure. Production managers were no longer held accountable for all costs incurred in or allocated to their respective areas so, in effect, they were no longer cost center managers. The cost detail in the new performance reports was reduced considerably. In the new system the only cost figure on which managers were evaluated was the spending by the employees in their organizations. For example, in January 1991 spending on maintenance in the melt shop was $300,000, but only $30,000 of this amount was spent by people reporting in Gary Downey’s organization. Thus, Gary’s report included only the $30,000 figure. The other $270,000 was reported to other managers, particularly those of maintenance and purchasing.

C. The implementation process
On January 1, 1991 Bob Nenni discontinued the Operating Statistics Report system. He believed that the new system would have had no chance if “the managers kept using the old data and never seriously considered improvements that could be made.” The accounting department began the process of
41

Chapter 2 · Results Controls

producing new sets of reports. As the entire task could not be accomplished immediately they focused their attention on producing some pilot reports for a subset of the measures. They focused first on heats per week, tons per man hour, physical yield (a component of the total quality index), and spending. Exhibit 6 gives an example of a report for the Melting Operations Manager. The operating managers’ initial reaction to the sample reports they were given was dissatisfaction. The early reports did not provide the line-item expense detail to which they had become accustomed. In addition, they were no longer given Cost Above information, so they asked, “Where are my spending numbers?” In early April 1991 Charlie Bradshaw told Bob Nenni, “I’ve received nothing of use from your department since you discontinued the old reports.” Another manager complained, “It almost seems like the operating managers finally understood the old report, so they decided to change it.” In late April the accounting group backed off their initial implementation plan. They started to provide spending numbers for the entire cost center in addition to the spending initiated by a manager’s direct reports. This change was made to give the operating managers a number that they could compare to their budgeted spending targets which had been prepared using the old measurement philosophy. Starting in 1992, however, they promised that the reports would reflect only the new cost performance philosophy. By then the performance targets would be set using that same philosophy. In June 1991 Bob Nenni reflected on nine months since the design meetings began. He was convinced that the company was on the right track even though some of the managers were uncomfortable with the new system. And he knew that the delays in the implementation process had frustrated both the information users and his accounting staff. He noted:

We’re trying to change the way the managers think. The new system is not yet part of their mentality. Changing mindsets is ultimately more important and more challenging than the technical job of producing the reports. But we in accounting feel we will now be more useful to the organization. We were spending 60% of our time in accounting on the non-value-added chores of inventory valuation for financial reporting purposes. We have now reduced that to 20%.

REMAINING ISSUES
In 1991 two related performance evaluation/ incentive issues arose in discussion. One was an issue about how to evaluate managers’ performances in situations where the numbers were distorted by uncontrollable factors. For example, early in 1991 the melt shop suffered two transformer failures, apparently because of fluctuations in the line voltages provided by the local utility, Kansas City Power and Light. Such failures had happened nearly every year, but shop managers had recently upgraded some of their electrical switches to try to eliminate the problem. Nonetheless, the failures occurred again, and by April Gary Downey knew that his goal to average 101 heats per week was impossible. The failure of the melt shop to achieve its plan would mean that the Kansas City Works as a whole would not be able to achieve its plans for 1991. Rob Cushman knew that at the end of the year he would have to decide whether or not to let this, and perhaps other similar occurrences, affect the evaluations of his operating managers. The second was an issue as to whether to increase the proportion of total compensation that was linked to individual performance evaluations. In other words, how much of total compensation should be provided in fixed salary, and how much should be paid only to those who were good at getting things done and done well?

42

Armco, Inc.: Midwestern Steel Division

EXHIBIT 1 Three-year financial history of Armco, Inc.1
(Dollars and shares in millions, except per share data) 1990 Sales2 Operating profit3 Net income (loss)4 Net income (loss) per share – primary Capital expenditures Number of shares of common stock outstanding Number of employees5 Total domestic retirees
1 2

1989 2,422.7 239.8 165.0 1.78 169.7 88.4 10,500 15,900

1988 3,277.3 222.3 145.4 1.57 120.2 87.8 19,500 21,400

1,735.2 76.9 (89.5) (1.10) 85.8 88.5 9,800 15,700

3 4 5

Certain amounts in the prior periods have been reclassified to conform to the 1990 presentation. Effective May 13, 1989. Armco sold certain assets and a portion of its Eastern Steel Division’s business to Kawasaki Steel Investments, Inc. in exchange for cash and a partnership interest in a joint venture. This division had annual revenues in excess of $1 billion. Includes special credits (charges) of $80.7 and $(35.0) in 1989 and 1988 respectively. Includes the cumulative effect of accounting changes in 1988 of $37.4 or 43 cents per share. Excludes discontinued operations, associated companies, and Armco Financial Services Group.

43

EXHIBIT 2 Organization chart, Midwestern Steel, March 1991

44

Armco, Inc.: Midwestern Steel Division

EXHIBIT 3 Excerpts from Operating Statistics Report, 2711–#2 Melt Shop
Tons/ Tap/Tap Hour 36.01 37.40 39.17 41.30 45.15 Non-Metallics Tons $/NT 0.04056 0.04304 0.04536 0.04776 0.04224 3.43 3.58 3.58 3.74 4.58 Salaries MH $/NT 0.0815 0.0818 0.0756 0.0822 0.0649 1.67 1.73 1.63 2.06 1.53

1 of 3
Hourly Supervision MH $/NT 0.0000 0.0000 0.0013 0.0006 0.0023 0.00 0.00 0.03 0.02 0.07

Tons 1985 1986 1987 1988 1989 706,237 737,380 741,234 800,581 870,768

Yield 93.64 93.38 93.85 92.79 94.42

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

69,920 68,106 68,240 83,797 74,507 80,190 76,513 32,489 70,475 85,128 85,992 85,945

93.88 91.41 95.74 95.30 95.50 96.18 94.58 96.32 95.54 95.21 96.47 94.15

42.71 44.81 49.58 46.33 46.96 44.44 44.37 43.84 51.03 52.66 54.48 54.48

MONTHLY STATISTICS 0.0411 6.50 0.0471 7.69 0.0512 7.75 0.0398 5.57 0.0593 7.50 0.0532 6.56 0.0409 4.95 0.0851 11.18 0.0578 5.90 0.0466 5.76 0.0546 5.85 0.0400 5.72 YEAR-TO-DATE STATISTICS 0.0441 7.09 0.0464 7.30 0.0445 6.81 0.0475 6.94 0.0486 6.88 0.0474 6.59 0.0496 6.86 0.0506 6.75 0.0501 6.63 0.0506 6.55 0.0495 6.47 OBJECTIVES 0.0000 4.18 0.0000 4.18

0.0620 0.0628 0.0625 0.0442 0.0433 0.0462 0.0486 0.1140 0.0525 0.0430 0.0426 0.0427

1.58 1.57 1.55 1.14 1.30 1.28 1.27 3.10 1.34 1.21 1.13 1.24

0.0057 0.0047 0.0038 0.0033 0.0046 0.0047 0.0032 0.0047 0.0038 0.0014 0.0037 0.0038

0.19 0.15 0.12 0.11 0.15 0.15 0.10 0.15 0.13 0.04 0.12 0.13

FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

138,026 206,268 290,065 364,572 444,761 521,274 553,763 624,238 709,366 795,358 881,303

94.91 95.72 96.10 96.32 96.57 96.55 96.63 96.70 96.73 96.86 96.79

43.72 44.00 45.48 45.65 45.88 45.66 45.59 45.38 45.99 46.63 47.30

0.0625 0.0625 0.0572 0.0543 0.0529 0.0523 0.0559 0.0554 0.0540 0.0528 0.0518

1.58 1.57 1.44 1.41 1.39 1.37 1.47 1.46 1.43 1.40 1.38

0.0052 0.0047 0.0043 0.0044 0.0044 0.0043 0.0043 0.0042 0.0039 0.0039 0.0039

0.17 0.15 0.14 0.14 0.14 0.14 0.14 0.14 0.13 0.13 0.13

DEC YEAR

80,910 942,114

92.75 92.75

48.10 48.10

0.0537 0.0600

0.83 0.85

0.0005 0.0005

0.02 0.02

DEC YTD

5,035 −60,811

0.41 1.24

5.38 −1.80

VARIANCE FROM OBJECTIVES −0.0400 −1.536 0.0111 −0.04952 −2.288 0.0083

−0.41 −0.53

−0.0033 −0.0034

−0.11 −0.11

Note: Figures are disguised.

45

Chapter 2 · Results Controls

EXHIBIT 3 continued
Repair Labor Tons 1985 1986 1987 1988 1989 706,237 737,380 741,234 800,581 870,768 MH 0.0671 0.0670 0.0749 0.0862 0.0685 $/NT 2.06 2.25 2.22 2.86 2.32 S-Order Labor MH $/NT 0.0005 0.0047 0.0000 0.0000 0.0021 0.01 0.14 0.00 0.00 0.06 S Order Matl 0.01 0.17 0.19 0.61 0.12 Maint Matl 2.41 2.60 1.57 2.53 2.47 Maint Outage 0.00 0.00 0.00 0.00 0.00

2 of 3
Electricity KWH $/NT 528.3264 494.3517 479.9061 469.7077 455.1210 20.90 19.23 18.41 18.37 18.99

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

69,920 68,106 68,240 83,797 74,507 80,190 76,513 32,489 70,475 85,128 85,992 85,945

0.0848 0.0686 0.0722 0.0563 0.0749 0.0521 0.0338 0.3352 0.0472 0.0504 0.0526 0.0347

2.97 2.38 2.48 2.01 2.71 1.92 1.25 12.40 1.76 1.84 1.96 1.40

MONTHLY STATISTICS 0.0006 0.01 0.54 0.0013 0.04 0.99 0.0020 0.06 1.06 0.0030 0.10 0.74 0.0035 0.11 0.83 0.0050 0.16 0.71 0.0061 0.20 0.72 0.0174 0.57 1.81 0.0010 0.03 1.19 0.0007 0.02 1.99 0.0003 0.01 2.07 0.0001 0.00 1.02 YEAR-TO-DATE STATISTICS 0.0009 0.03 0.76 0.0013 0.04 0.86 0.0018 0.06 0.83 0.0021 0.07 0.83 0.0026 0.08 0.81 0.0031 0.10 0.79 0.0040 0.13 0.85 0.0036 0.12 0.89 0.0033 0.11 1.02 0.0030 0.10 1.14 0.0027 0.09 0.93 OBJECTIVES 0.15 0.34 0.15 0.34

3.50 4.27 3.73 2.81 2.89 2.68 2.92 8.66 2.42 3.40 1.79 2.18

0.84 0.86 0.86 0.70 0.79 0.73 0.77 −19.91 0.83 0.69 0.68 0.68

497.9036 507.3894 522.4422 491.3568 512.6357 516.9778 511.5135 540.0536 540.4950 528.3277 533.7443 510.6142

19.20 19.82 18.34 18.01 20.43 18.51 19.89 28.03 21.54 18.66 19.51 18.95

FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

138,026 206,268 290,065 364,572 444,761 521,274 553,763 624,238 709,366 795,358 881,303

0.0768 0.0752 0.0698 0.0708 0.0674 0.0625 0.0785 0.0750 0.0720 0.0699 0.0665

2.68 2.61 2.44 2.50 2.39 2.22 2.82 2.70 2.60 2.53 2.42

3.89 3.84 3.54 3.41 3.28 3.22 3.54 3.42 3.41 3.24 3.13

0.85 0.86 0.81 0.81 0.79 0.79 −0.42 −0.28 −0.17 −0.07 0.00

456.8288 462.1852 458.1932 459.7953 461.6315 462.1277 463.7972 466.9283 468.5328 470.3372 469.7382

19.50 19.12 18.80 19.13 19.02 19.15 19.67 19.88 19.73 19.71 19.64

DEC YEAR

80,910 942,114

0.0552 0.0552

1.95 1.95

0.0048 0.0048

1.75 1.75

0.00 0.00

445.5000 445.4858

19.15 19.15

DEC YTD

5,035 −60,811

0.0205 −0.0113

0.55 −0.47

VARIANCE FROM OBJECTIVES 0.0047 0.15 1.36 − 0.43 0.0047 0.07 −0.59 −1.38

−0.68 0.00

−18.69472 −24.25236

0.20 −0.48

46

Armco, Inc.: Midwestern Steel Division

EXHIBIT 3 continued
Natural Gas MMBTU $/NT 0.0147 0.1766 0.2242 0.2408 0.2180 0.06 0.50 0.60 0.70 0.52 Gas & Diesel Fuel 0.02 0.01 0.01 0.02 0.02 Loco Cranes Lubricants 0.00 0.00 0.01 0.03 0.03 Hrs 0.0000 0.0000 0.0000 0.0003 0.0002 $/NT 0.00 0.00 0.00 0.03 0.02 Total Cost Above 76.06 79.38 76.30 79.03 79.40

3 of 3
Total Cost 168.70 164.20 173.78 216.37 211.40

Tons 1985 1986 1987 1988 1989 706,237 737,380 741,234 800,581 870,768

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

69,920 68,106 68,240 83,797 74,507 80,190 76,513 32,489 70,475 85,128 85,992 85,945

0.1530 0.1419 0.2234 0.2002 0.2033 0.2094 0.2698 0.3314 0.9326 −0.6225 0.0847 0.0888

0.31 0.33 0.81 0.54 0.55 0.57 0.72 0.74 2.06 −0.98 0.20 0.23

MONTHLY STATISTICS 0.01 0.16 0.00 0.03 0.00 0.12 0.01 0.00 0.01 0.00 0.01 0.21 0.01 −0.15 0.04 0.11 0.00 0.04 0.01 0.04 0.02 0.05 0.01 −0.02 YEAR-TO-DATE STATISTICS 0.01 0.10 0.01 0.11 0.01 0.08 0.01 0.06 0.01 0.09 0.01 0.05 0.01 0.06 0.01 0.05 0.01 0.05 0.01 0.05 0.01 0.05 OBJECTIVES 0.02 0.03 0.02 0.03

0.0002 0.0001 0.0001 0.0000 0.0000 0.0001 0.0002 0.0005 0.0003 0.0000 0.0001 0.0001

0.02 0.01 0.01 0.00 0.00 0.00 0.02 0.04 0.02 0.00 0.02 0.00

89.62 93.82 90.68 81.56 94.82 89.58 90.10 142.82 99.91 93.99 86.31 80.60

198.21 207.92 202.70 193.05 211.44 204.01 204.57 251.19 216.90 205.63 195.64 194.33

FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

138,026 206,268 290,065 364,572 444,761 521,274 553,763 624,238 709,366 795,358 881,303

0.1475 0.1726 0.1806 0.1852 0.1896 0.2014 0.2090 0.2907 0.1811 0.1707 0.1627

0.32 0.48 0.50 0.51 0.52 0.54 0.56 0.73 0.52 0.49 0.46

0.0002 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001

0.02 0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01

91.69 91.35 88.53 89.82 89.78 89.82 92.92 93.72 93.76 92.95 91.74

203.01 202.91 200.06 202.38 202.68 202.95 205.79 207.04 206.87 205.66 204.57

DEC YEAR

80,910 942,114

0.2192 0.2192

0.51 0.51

0.0002 0.0002

0.02 0.02

76.36 76.53

DEC YTD

5,035 −60,811

0.1304 0.0565

VARIANCE FROM OBJECTIVES 0.28 0.01 0.05 0.0001 0.05 0.01 −0.02 0.0001

0.01 0.01

− 4.24 −15.21

47

Chapter 2 · Results Controls

EXHIBIT 4 Vision management process

EXHIBIT 5 Cascading of total quality goals

48

EXHIBIT 6 New system pilot performance report
Actual 1990 Jan Feb Mar Apr May June Plan June Var June YTD Actual YTD Plan YTD Var

PRODUCTIVITY EAF/LAR1: Tons 587,535 12,164 38.64 79.20 70.40 75.20 78.40 61.60 77.60 40.30 37.97 39.19 39.86 37.44 39.32 42.75 81.60 1,152 982 1,133 1,149 956 1,143 1,102 58,080

Tap to tap hrs

46,624 55,532 57,281 46,706 56,188 58,832

(3,302) (53) (4.29) (5)

320,359 6,517 39.32 73.60

347,564 6,617 42.02 80.00

(34,007) 100) (3.37) (8)

Tons/hr

Heats/wk

HIT rate 572,704 6,346.40 72.19 80.82 53.78 76.99 81.75 563.20 512.00 563.20 550.40 556.80 65.74 56,901 44,655 54,201 56,246 45,751 55,272 550.40 80.34 57,361 550.40 83.38 (2,611) 0) (3.80) 313,030 3,296.00 75.98 338,876 3,308.80 81.94 (32,310) 12.80 (7.44)

CASTER:

Tons

Hours

Tons/hr

LABOR EAF/LAR: Man hours 107,307 4.38 78,110 6.01 6.81 5.86 6,683 6,089 6,724 6.45 4.72 4.17 4.34 9,837 8,929 10,236

9,806 4.03 4,964 6.86

9,806 3.81 6,938 5.27

10,000 4.66 6,724 6.58

9,514 4.94 6,567 6.98

(608) (0.56) (196) (0.51)

58,632 4.37 39,724 6.30

55,629 5.00 39,616 6.84

(3,753) (0.78) (136) (0.87)

Prod tons/manhr

CASTER:

Man hours

Prod tons/manhr

Armco, Inc.: Midwestern Steel Division

1

Electric arc furnace/ladle arc furnace.

49

50
Actual 1990 Jan Feb Mar Apr May June Plan June Var June YTD Actual YTD Plan YTD Var 95.5% 95.4% 97.5% 96.0% 95.3% 95.6% 97.6% 96.2% 98.1% 98.4% 97.5% 0.9% 95.6% 94.3% 95.0% 95.0% 95.0% 0.0% 96.4% 96.6% 97.0% 96.3% 97.9% 96.0% 96.7% 95.2% 97.7% 95.0% 97.5% 0.2% 0.2%

Chapter 2 · Results Controls

EXHIBIT 6 continued

YIELD EAF/LAR: Reported

Applied

CASTER:

1,243,625 68,427 290,937 305,132 313,962 312,456 289,136 79,751 46,168.80 136,724 92,687 1,329,800 1,499,361 1,386,807 1,421,037 115,669 325,392

SPENDING 1,443,067 116,149 351,824 1,658,740 236,102 22,982 36,255 8,325,300 539,908 1,863,450 9,870,859 693,743 1,936,292 1,545,559 153,835 72,842

EAF (electric arc furnace)

LAF (ladle arc furnace)

CASTER

TOTAL SPENDING Total$ 1,909,441 26.87 28.72 25.31 26.45 32.10 1,602,990 1,714,684 1,859,492 1,835,987

1,804,461 26.12

2,099,801 295,340 29.29 3.17

10,728,658 12,500,895 27.42 29.51

1,772,236 2.10

$/NT

ADDITIONAL MEASURES EAF: KWH/NT 3.41 4.94 0.16 28.80 0.78 0.54 0.45 20.80 23.20 25.60 0.63 0.09 0.49 0.46 5.41 5.52 4.86 342.40 353.60 344.00 331.20 5.07 0.48 27.20 0.00

353.60 5.30 0.56 28.80 1.46

342.40 4.73 0.39 25.60 0.60

332.00 4.73 0.18 28.80 0.80

(13) 0.00 (0.25) 3.20 0.20

344.00 4.11 0.40 25.60 0.59

332.00 3.78 0.19 29.60 0.80

(15) (0.51) (0.26) 4.00 0.21

Electrodes/NT:

MMBTU’s/NT

LAR:

KWH/NT

Electrodes/NT: