Week 5 Critique of Quantitative Research Article

Beyond Dichotomy: The Curvilinear Relationship Between Social Responsibility
And Financial Performance

For
MGT 361: Managerial Research Design & Analysis

In partial fulfillment of the requirements
For
The Bachelor of Business Administration Degree

Submitted By

AL King

November 20, 2010

This is a critique of the research article, Beyond Dichotomy: The Curvilinear Relationship between Social Responsibility and Financial Performance, written by Michael L. Barnett and Robert M. Salomon. Michael Barnett is a professor at the College of Business Administration, University of South Florida and Robert Salomon is a professor at the Stern School Business, New York University. There has been much debate on whether or not there is a link between financial and social performance when it comes to investing. Most research is done by comparing socially responsible investing (SRI) funds to those of funds that do not screen their holdings based on social and or environmental criteria. Barnett and Salomon’s research also tackles this question however, instead of doing the same types of comparison, they chose to compare and measure mutual funds that all practice socially responsible investing (SRI), in order to discern which funds perform better and what makes them different from the other SRI funds. Barnett and Salomon developed four (4) hypotheses to test during their research:
1. The relationship between the intensity of social screening and financial performance for SRI funds is curvilinear (U-shaped).
2. SRI funds that select firms for their portfolios based on labor relations screening criteria will earn higher financial returns than those that do not.
3. SRI funds that select firms for their portfolios based on community relations screening criteria will earn higher financial returns than those that do not.
4. SRI funds that select firms for their portfolios based on environmental screening criteria will earn higher financial returns that those do not. The data that Barnett and Salomon used to conduct their research came from several sources. Their initial sample consisted of SRI mutual funds tracked by the Social Investment Forum, which is a non-profit organization that promotes and encourages the growth of socially responsible investing. From this source Barnett and Salomon were able to identify their initial sample of 67 socially responsible funds. The article does not specify an exact amount of how many SRI funds were in existence during the research phase (population). However, as of 2010 there were 250 socially screened mutual fund products in the U.S. By contrast, there were just 55 SRI funds in 1995. (Social Investment Forum). In this research the following three different variables were used:
Dependent Variable Because the authors are testing for the effects of social screening on financial performance the dependent variable was the risk-adjusted financial performance of a given SRI fund in a month. Risk-adjusted performance (RAP) is defined as the average monthly return, measured as the percentage change in a fund’s market value from the beginning to the end of a month. (Barnett and Salomon, 2006)
Independent Variable In previous research, SRI funds were categorized simply as being a fund that screened for social responsibility versus funds that did not. Their performance was then compared to non-SRI funds. Yet, all SRI funds are not the same; some have stricter screening standards than others. Therefore the independent variables in this research are the different screening intensity used by the funds. The Social Investment Forum lists 12 screens that SRI funds may use to filter firms from their investment portfolios, based on their affiliation with the following 12 industries or issues: alcohol, tobacco, gambling, defense/weapons, animal testing, product/service quality, environment, human rights, labor relations, employment equality, community investment and community relations. Screening intensity varies from 1 to 12; if a fund’s screening intensity is a 12 then the fund used all of the 12 previously listed screens, where a screening intensity of 1 means only one screen was used by the fund.
Control Variable Because the dependent variable measures the financial performance of a fund the authors had to find a way to control random factors that could affect financial performance. These random factors were the fund’s age, size, global economic cycles and risks. In order to control these factors the authors’ created the following control variables:
Fund Age – a count of the number of months since the fund’s inception.
Total Assets – a measure of overall fund assets (measured in millions of U.S. dollars)
Global Fund – takes the value of 1 for funds with international holdings, zero if otherwise.
Percent Stock – measures the effects of pursuing different general investment strategies.
Percent Bonds – measures the effects of pursuing different general investment strategies.
Yearly Dummy Variables - control for any residual macro-economic factors that affect all funds similarly. (Barnett and Salomon, 2006) Along with the data collected from the Social Investment Forum, Barnett and Salomon also used CRSP (The Center for Research in Security Pricing) data to track each fund’s financial performance. The authors compiled monthly financial performance data from 1972 to 2000. This data was also supplemented with mutual fund information from Weisenberger and ICI (Investment Company Institute) which are mutual fund tracking services. The reason for supplementation is because most SRI funds did not exist before 1996, there has been over a 218% growth in the number of funds since then. Fortunately, none of the funds in the initial sample exited during the event window. (Barnett and Salomon, 2006) In selecting an appropriate statistical method to test their hypotheses, the authors began with the least-squares method. This type of method is used in regression analysis (one-tailed tests were used to test the hypotheses relationship). Regression analysis helps us understand how the typical value of the dependent variable (RAP of SRI funds) changes when any one of the independent variables (screening intensity) is varied, while the other control variables are held fixed. Descriptive statistics were also used such as tables that showed product moment correlations. Scatter plots were also used and showed fund performance and screening intensity over time, as well as the variance across different screening strategies and their effects on the RAP of SRI funds. The results from the statistical methods show that there is a curvilinear relationship between screening intensity and fund performance. A curvilinear relationship is a relationship between two or more variables which is shown graphically by anything other than a straight line.
Since there is a curvilinear relationship, hypothesis 1 can be accepted. Risk-adjusted performance did decline at first as screening intensity increased (reaching a minimum at 7 screens), but then increased continuously until it reached the maximum social screening intensity (12 screens). (Barnett and Salomon, 2006). Hypothesis 2 however had to be rejected due to the results of the statistical testing. Funds that excluded firms that violated norms of equal employment actually suffered a financial loss of about 0.29 percent per month. Hypothesis 3 was partially supported due to the fact that no relationship could be found between screening firms that invested in their local communities and financial performance, yet, mutual funds that developed positive relationships with their local communities did perform better. Finally, hypothesis 4 also had to be rejected. Screening on the basis of environmental criteria was negatively related with risk-adjusted financial performance. SRI funds that eliminated environmentally poor firms performed about .038% worse per month than the baseline SRI fund. Of course there were many other factors that come into play when dealing with investments, especially SRI funds that can make results from the statistical methods somewhat less than reliable, and the authors are quite aware of this. Which is why I was satisfied with their attempts to implement control variables as well as launching a sensitivity analysis. Sensitivity analysis is a technique used to determine how different values of an independent variable will impact a particular dependent variable under a given set of assumptions. In conclusion Barnett and Salomon discuss their findings and explain how they support both portfolio and stakeholder theories to varying degrees. Funds that use strict screening may eliminate underperforming firms from their portfolio; however, on the other hand SRI funds that use few social screens improve financial performance through benefits received from increased diversification. It appears that even though social screening forces a narrowing of investment choices, if implemented effectively social screening can lead to an increase in financial returns. The authors also stress that with any study of financial performance, their analysis is retrospective and must be interpreted with caution; Barnett and Salomon also encourage others to improve upon their findings with more precise and specific measures.

REFERENCES
Barnett, M. L., & Salomon, R. M. (2006, September 13). Beyond Dichotomy: The Curvilinear Relationship between Social Responsibility and Financial Performance [Electronic version]. Strategic Management Journal, 1101-1122.
Socially Responsible Investing Facts. (n.d.). In Social Investment Forum. Retrieved November 15, 2010, from http://www.socialinvest.org/resources/sriguide/srifacts.cfm