Baron, David P., Maretno A. Harjoto, and Hoje Jo. "The Economics and Politics of Corporate Social Performance." Working Paper, Stanford Graduate School of Business, April 2009.

This ambitious study seeks to place corporate social performance (CSP) in its proper context by postulating that firms are "operating in three markets: a product market, a capital market, and a market for social pressure as generated by government, NGOs, and social activists." The relationships among these markets are estimated using a three equation structural model.

(Author Baron draws a distinction between CSP and CSR, "where the latter involves a moral duty to undertake social activities. In contrast, CSP need not arise from moral responsibilities. CSP as considered here pertains to social activities that satisfy two conditions. First, the activities are beyond the law and regulation. Second, the activities involve the private provision of public goods or private redeistribution.")

The empirical portion of the study covers the 1996-2004 time period. It uses KLD data as the basis for estimating CSP performance (the KLD "strengths") and social pressure (the KLD "concerns"). IRRC corporate governance data are used "to control for corporate governance characteristics and capital market monitoring."

Corporate financial performance is measured using Tobin's Q, using Fama & French's 48 industry dummy variables as controls. An industry competitiveness index is also used as a control, as are a variety of financial characteristics (e.g., R&D intensity, debt ratio, etc.).

Finds that "for the full dataset, CFP and CSP are found to be largely unrelated," but that "greater social pressure is associated with worse CFP, which could reflect the effects of pressure on firms' reputations, brand equities, or productivity." Also finds that consumer companies have a higher propensity to better CSP, while industrial companies are less likely to be strong social performers [this finding accords with the experience of practitioners in their construction of social indexes - lk].

A breakdown between the first and second four years of the sample (which map to presidential administrations) finds a stronger relationship between CSP and CFP in the Bush era than in the Clinton era.

The authors further note that "the absence of an empirical relation between financial performance and social performance or the presence of a positive relation for consumer industries and a negative relation for industrial industries does not mean that there is no causal relation for an individual firm...the challenge for empirical research is to determine whether a causal relation exists, and this will have to be done at the level of individual firms."