[an occasional series skating on the edge of ethics, investing and SRI (Sustainability, Responsibility, Impact)]

The winner of this year’s Moskowitz prize for excellence in quantitative SRI research is Dr. Caroline Flammer, Ivey Business School, for “Does Corporate Social Responsibility Lead to Superior Financial Performance? A Regression Discontinuity Approach”. The approach taken in this study finds a positive causal relationship between CSR and financial performance.

This paper uses a new approach to assess, quantitatively, whether CSR[1] is beneficial to companies. What is novel is the use of a regression discontinuity design that allows the author to draw sharp causal inference on the impact of CSR on financial performance.

Prior studies of the effect of CSR on performance have found, in general, a small positive correlation between CSR and financial performance.[2] A common concern with these studies is that endogenous factors, not reflected in the model or that cannot be measured, can just as easily explain the higher financial performance associated with CSR.

To address these concerns, Dr. Flammer’s study uses a different approach. It examines the impact of shareholder-sponsored CSR proposals that pass or fail by a small margin – “close-call” proposals – on stock prices. This approach addresses the above concerns in two ways:

• Shareholder-sponsored proposals are developed by external shareowner groups, and hence are independent of company management. Unlike management-sponsored proposals, they cannot be strategically withdrawn by the management if the outcome is expected to be unfavorable to the management.
• Restricting the analysis to close-call votes eliminates proposals that pass or fail by a wide margin (“wide-margin” proposals). The rationale for eliminating wide-margin proposals is that the outcomes of wide-margin proposals are anticipated and reflected in stock prices before the vote takes place. In contrast, close-call votes have a random element: their outcomes are uncertain. Close-call votes are not reflected in prices in advance and therefore are more useful for estimating the causal effect of CSR on financial performance.

Outcomes of close-call votes have a binary distribution. They are modeled using a “Regression Discontinuity Design”, a method from economics used to compare outcomes just above and just below a discontinuous threshold. The 50% majority requirement to pass is the threshold around which a minor difference in votes causes a discontinuity in outcomes – i.e., the proposal either passes or fails.

Key findings:

• Companies whose shareholder proposals pass on a “close-call” vote have an average abnormal return of +0.92% relative to close-call proposals that fail.
• Discontinuity at 50% is real. Abnormal returns are a continuous and smooth function of vote share up to the 50% threshold, at which point they see a significant positive jump.
• Abnormal returns move toward zero as the vote share increases or decreases away from 50%, an indicator that the outcome of non-close votes is anticipated by the market and reflected in stock prices prior to the vote.
• Positive abnormal returns relate to approval, not implementation of shareholder CSR proposals. Approved proposals have a 52% probability of implementation by company management. On a probability-adjusted basis, the average estimated positive abnormal return of shareholder CSR proposals that are both approved and implemented is 1.77%.
• Firms with relatively low levels of CSR prior to the vote see higher value gains; those with higher levels of CSR see lower value gains. This finding is consistent with decreasing marginal returns.
• Higher CSR gains are seen in industries with higher CSR norms (i.e. “clean” industries).
• Operating performance improvements from close-call votes persist for years following the vote. Improvement is attributed to higher labor productivity and sales growth.

Summary of results:

• 2,729 total proposals: abnormal return ~0%.
• 2.607 wide-margin votes: abnormal return ~0%
• 122 proposals, vote share +/-10% of majority: abnormal return = 1.07%, significant at 5% level
• 61 proposals, vote share +/-5% of majority: abnormal return = 1.36%, significant at 5% level.

Institutional ownership of all companies in the sample averaged 63% (range: 8.4%  – 91%). In the years preceding the vote, there is no significant difference in institutional ownership for companies that subsequently approve or reject a close-call CSR proposal.

The Moskowitz Prize[3] is sponsored by the Center for Responsible Business, Haas School of Business, University of California, Berkeley, to recognize outstanding research on socially responsible investing and the financial implications of responsible business practices. It is the only global award for excellence in quantitative SRI research (sustainable, responsible, impact investing). Studies are reviewed by an expert panel of judges from academia and the investment industry. Judging criteria include:

1.      Practical significance to practitioners of socially responsible investing

2.      Appropriateness and rigor of quantitative methods

3.      Novelty of results

The prize has been awarded annually since 1996. It is named after Milton Moskowitz, one of the first to publish quantitative research on the financial performance of screened and unscreened portfolios, and known for Fortune Magazine’s “The 100 Best Companies to Work for in America”. Past winners include Meir Statman, Denys Glushkov, Elroy Dimson, Oğuzhan Karakaş, Xi Li, Sadok El Ghoul, Omrane Guedhami, Chuck C. Y. Kwok, Dev Mishra, Rob Bauer, Daniel Hann and others.

Full paper: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2146282

About the author: http://www.ivey.uwo.ca/faculty/directory/caroline-flammer/

About the Moskowitz Prize: http://responsiblebusiness.haas.berkeley.edu/programs/moskowitzresearchprogram.html

[3]  http://www.ssrn.com/update/cgn/cgnann/ann13003.html