A note on the interaction between corporate social responsibility and financial performance

Abstract

This paper studies the interaction between financial and social performance. The research is about the overall social strengths and concerns of firms as well as strengths and concerns with respect to firms' community involvement, employee relations, diversity, environment and product. Using a sample of 289 firms from the US covering the period 1991–2004 and employing two different test methods, namely lagged OLS and Granger causation, there appears to be preliminary evidence that the direction of the ‘causation’ predominantly runs from financial to social performance. However, the specific interaction patterns tend to vary along the different dimensions.

Introduction

Attention for corporate social responsibility (CSR) has increased significantly during the last decade. Many firms started reporting about their ethical, social and environmental conduct. And in marketing, being green and social is positioned as a relevant product and firm characteristic. In academic research, CSR has become a topic of interest too. Many studies investigate the connection between financial and social performance (see Lockett, Moon, and Visser, 2006). Numerous theoretical views on the link between financial and social performance are put forward (for an overview see Allouche and Laroche, 2006). Furthermore, a large number of empirical studies investigate the relationship between social and financial performance (see Orlitzky, Schmidt and Rynes, 2003). Not surprisingly, there are different opinions about the interaction between financial and social performance and the empirical research has not arrived at a consensus. First, the liberal view suggests a negative link as social responsibility involves costs and therefore worsens a firm's competitive position (Friedman, 1970). Related is the view that social constraints on firms and socially responsible behavior may conflict with value maximization (Brummer, 1991, Jensen, 2001). There may also be a negative link between social and financial performance when managers pursue their own objectives, which may conflict with shareholder and stakeholder objectives (Williamson, 1964, Jensen and Meckling, 1976). Sethi (1979) argues that firms will put social responsibility over financial performance in a quest for legitimacy and when they are under pressure from stakeholders. Preston and O'Bannon (1997) argue that an accrual of funds to invest in social performance can lead to poorer financial performance due to negative synergy. Alternatively, the market equilibrium might cancel out the costs of corporate socially responsible behavior (McWilliams and Siegel, 2001). The stakeholder view holds that satisfying stakeholders' interests will result in an improvement of the firm's financial and economic performance (Freeman, 1984, Porter and Van der Linde, 1995). However, McGuire, Sundgren and Schneeweiss (1988) find that a firm's prior financial performance conditions corporate social responsibility more than its subsequent financial performance. McWilliams and Siegel (2001) argue that firms invest in social activities because they want to satisfy the demands of their stakeholders. In market equilibrium, the costs and the profits of socially responsible conduct will compensate each other. This is the basis for a neutral interaction between financial and social performance. Lastly, there is the view that the links are quite complex (Bowman and Haire, 1975, Moore, 2001, Barnett and Salomon, 2002). For example, there can be an (inverted) U-shaped relationship between the two (Barnett and Salomon, 2002).

Margolis and Walsh (2001) offer an excellent overview of the numerous empirical studies after the relationship between social and financial performance. They find that corporate social performance is treated as an independent variable in most studies. This variable is used to predict or precede financial performance. Approximately 50% of the studies found a positive relationship between the two, 25% found no relationship, 20% had mixed results and 5% had a negative relationship. A minority of the studies treated corporate social performance as the dependent variable. In two thirds of these, there was a positive relationship between social and financial performance. Margolis and Walsh (2001) were very cautious about deriving conclusions from their overview. This is because there are serious methodological concerns about many of the studies. Their main criticism is with respect to the measurement of corporate social responsibility, the wide diversity of measures used to assess financial performance, and the direction and mechanisms of causation. Furthermore, Orlitzky, Schmidt and Rynes (2003) performed a meta-analysis. They found that the relationship between social and financial performance is rather positive in a wide variety of contexts and sectors. However, they also establish that the residual variance usually is quite substantial.

The connection between social and financial performance plays an important role in the analysis of socially responsible investing (SRI) too. From a portfolio perspective, SRI eliminates securities from the universe of investable assets. Consequently, SRI reduces the potential for diversification. This has been studied, among others, by Bauer, Koedijk and Otten (2005) and Bello (2005). They find that the risk and return attributes of these screened SRI portfolios do not significantly differ from their conventional counterparts. Geczy, Stambaugh and Levin (2005) find that the cost of SRI depends on the perspective of the investors. Socially responsible investors who do not believe in the ability of mutual funds to outperform the market in terms of the Capital Asset Pricing Model do not face a significant opportunity loss from investing in SRI mutual funds. However, an investor believing in a world consistent with the Fama and French (1992) three-factor model may face an opportunity loss of approximately 30 basis points per month (Gezcy et al., 2005).

This paper aims at complementing the existing literature by explicitly studying the interaction between financial and social performance. The key question it addresses is “Does financial performance precede social performance, or is it the other way round?”. To answer this question, the research will associate the timing of financial performance with that of social performance. Of course, precedence is not identical with causality, but it comes much closer to this concept than the usual regression approach where correlation is at the basis of the analysis. This study is innovative in five respects. First, it is the first study that explicitly models the timing issues of social and financial performance with well-established techniques in a systematic matter. It employs two different techniques to assess the interaction between CSR and financial performance, namely simple OLS with distributed lags and Granger causation. Second is that it uses a well-established database for this purpose. To investigate the potential for causation, it uses financial data from Datastream and the database from Kinder, Lydenberg and Domini that takes account of the multidimensional aspects of CSR (see Sharfman, 1996, for an assessment of the construction of this KLD-database). A third innovation is that the paper looks into a timeframe of more than one decade. The period investigated is very interesting as it covers a time when many firms introduced social, ethical and environmental policy programs and when the stock market was turbulent. Fourth is that both financial risk and irresponsible social behavior are included in the analysis. This is to account for the fact that financial and social performance also have a downside. Within the context of the discussion of the interaction between these two items, the current study is the first to explicitly address this issue. Fifth is that the study does not only look into composite scores for corporate social responsibility but also into subscores, i.e. strengths and weaknesses of firms' community involvement, diversity in management, employee relations, environmental conduct, and product characteristics.

The structure of the remainder of the paper is as follows. The data and the methods employed are introduced first. Then, the results are reported. Lastly, there is a discussion of the findings and a brief conclusion.