The Association Between Corporate Social Responsibility and Financial Performance: The Paradox of Social Cost

Abstract

It is generally assumed that common stock investors are exclusively interested in earning the highest level of future cash-flow for a given amount of risk. This view suggests that investors select a well-diversified portfolio of securities to achieve this goal. Accordingly, it is often assumed that investors are unwilling to pay a premium for corporate behavior which can be described as “socially-responsible”.

¹ Having spelled out what we believe in an unbiased view of Friedman’s writings, it should be pointed out that even his “unequivocal” argument is ambiguous enough to provide some sanction for corporate management to engage in what they might view as socially-responsible actions. For example, in describing the proper role for corporate executives, Friedman has written that their responsibility is to conduct the business in accordance with the desires of stockholders, “which generally will be to make as much money as possible while conforming to the basic rules of the society, both those embodied in law and those embodied in ethical custom.” (1970: p. 33, emphasis added) Although he certainly does not accept the term “social-responsibility”, even Friedman recognizes the existence of corporate obligations beyond mere legal requirements. Even corporate managers of the Friedman-type need to make moral decisions about “ethical custom”, and can not escape formulating an answer to Friedman’s rhetorical question: “If businessmen do have social responsibility other than maximizing profits for stockholders, how are they to know what it is?” (1962: p. 133) Is this question different in kind to the following: If businessmen need to conform to the basic rules of society, which include those embodied in ethical custom, how are they to know what they are?

In his book (co-authored with Rose Friedman 1980) the author further elaborated:

Narrow preoccupation with the economic market has led to a narrow interpretation of self-interest as myopic selfishness, as exclusive concern with immediate rewards. Economics has been berated for allegedly drawing far-reaching conclusions from a wholly unrealistic “economic man” who is little more than a calculating machine, responding only to monetary stimuli. That is a great mistake. Self-interest is not myopic selfishness. It is whatever it is that interests the participants, whatever they value, whatever goals they pursue. (p. 18)

² The eight key attributes (of corporate reputation) listed by Fortune magazine (February 8, 1993) were the following: quality of management, financial soundness, quality of products or services, use of corporate assets, value as long-term investment, innovativeness, ability to attract, develop, and keep talented people, and community and environmental responsibility.

³ Our point here is that CSR may cause better long run financial performance. We also recognize, however that firms experiencing extreme financial distress may cut back first on CSR programs. In this special case, a deteriorating financial performance may directly lead to fewer CSR activities. This is true because there are fewer legal requirements associated with CSR commitments (implicit claims) than other more traditional corporate activities (explicit claims). It may be less costly to break CSR commitments than other more formal contractual agreements.

⁴ In selecting these firms the CEP “drew both on the holdings listed in the prospectuses of the socially responsible mutual funds and on lists provided by the SIF.” (p. 19) Additional information was drawn from reports prepared by Franklin Research and Development and Clean Yield.

⁵ The Altman’s Z-Score has been found useful in predicting bankruptcy. It is actually a combination of five additional financial ratios. For a full discussion see Stickney (1990) or Altman (1968).

⁶ Nevertheless, the contract may be extremely relevant from a legal perspective. See Martin Luther King’s, “A Letter From a Birmingham jail” (in Newton 1989).

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

It is generally assumed that common stock investors are exclusively interested in earning the highest level of future case-flow for a given amount of risk. This view suggests that investors select a well-diversified portfolio of ecurities to achieve this goal. Accordingly, it is often assumed that investors are unwilling to pay a premium for corporate behavior which can be described as “socially-responsible”. Recently, this view has been under increasing attack. According to the Social Investment Forum, at least 538 institutional investors now allocate funds using social screens or criteria. In addition, Alice Tepper Marlin, president of the New York-based Council on Economic Priorities has recently estimated that about $600 billion of invested funds are socially-screened (1992).

While the notion of socially-responsible investing is often a vague and ill-defined concept and therefore extremely difficult to quantify, there are nevertheless, a cluster of core issues which describe the practice. Among the most common issues are the following: environmental concerns, community relations, military contracts, nuclear energy, product quality, consumer relations, employee relations, philanthropy, and South African investments. There are many other issues which individual investors might use in classifying corporations as socially-responsible. One important example is the issue of the economic boycott of Israel. Because of the absence of this issue and many others, it is important to recognize that while socially-responsible investing represents an economic philosophy, in practice, it also tends to correlate with a political world-view, as well. Screening on the basis of social-responsibility refers merely to those rules which current practitioners employ in selecting corporate investments, in addition to the traditional economic screens. In this sense, it is a descriptive term only.

Although the practice of using both traditional economic criteria and social-responsibility screens to allocate funds is becoming more common, and while the legal ­constraints associated with the practice are apparently being removed, the implications are still not well understood. What are its financial benefits and costs? Is there a measurable financial impact? Therefore, the objectives of this study are four-fold. First, in the next section, we will review over twenty empirical studies which have attempted to measure both the direction and the degree of association between CSR and financial performance. Second, because of limitations inherent in the previous work, we further explore the association between corporate social-responsibility and traditional financial performance. In this study we examine the long-term financial performance of a group of 53 firms which have been identified by the Council on Economic Priorities (CEP) as being socially-responsible, and compare the financial performance of this group to a control sample matched by size and industry (Council on Economic Priorities et al. 1991). The rationale for basing our study on the CEP firms will be discussed in Sect. 2, where we describe the methodology and results of the study. Third, we hope that by further studying the statistical association between CSR and financial performance to shed additional light on both the benefits and the costs associated with socially-responsible actions, and in this way to formulate a better understanding of the nature and limitations of CSR. Finally, we conclude our study with implications for academics, investors, and corporate executives.

Section 1

In an attempt to understand the relationship between CSR and financial performance, there have been numerous studies which have measured the statistical association between perceived corporate social-responsibility and traditional financial performance. We have identified and reviewed 21 empirical studies which explicitly addressed this question as the major research objective. Our investigation reveals an important, and (we believe) unappreciated, empirical regularity. It can be succinctly stated as follows:

Nearly all empirical studies to date have concluded that firms which are perceived as having met social-responsibility criteria have either outperformed or performed as well as other firms which are not (necessarily) socially-responsible.

This surprising empirical regularity, which we label the “paradox of social cost”, demands an explanation. To the extent that social activities are costly to the firm (even while creating positive externalities), one would expect a negative relationship between social performance and financial performance at the individual firm level.

The Traditionalists’ View of the Corporation

Milton Friedman is most closely associated with the traditional view of the corporation (See Friedman 1962, 1970; Friedman and Friedman 1980). His position can be summarized as follows: Business managers have a responsibility to shareholders – the owners of the corporation – to maximize firm value. Managers, acting as agents of the shareholders, have no mandate to embark on socially-responsible projects that do not enhance the income generating ability of the firm. In addition, managers should not refrain from profitable investments which satisfy all legal constraints but do not conform to managers’ own personal social agenda. Rather, as Friedman put it, “The social responsibility of business is to increase profits.” He further emphasized, “Few trends would so thoroughly undermine the very foundations of our free society as the acceptance by corporate officials of a social responsibility other than to make as much money for their stockholders as they possibly can. This is a fundamentally subversive doctrine.” (1962: p. 133)

Friedman’s primary assumption which leads to his conclusion that CSR is a “subversive doctrine” is his belief that the term social-responsibility as applied to the corporate context, if it means anything at all, implies that the business manager “must act in some way that is not in the interest of his employers.” (1970: p. 33) Thus managers who act out of a sense of social-responsibility are engaging in a form of taxation without representation.

Further, Friedman believes that business managers have no comparative advantage when it comes to implementing social programs. Managers are experts in producing products, selling them, or financing them. Management has no necessary expertise in fighting social-ills.

We believe that Friedman’s argument is both rigorous and somewhat convincing. His voice, although the loudest, clearest, and least apologetic, is by no means solitary.¹ Numerous economists, accountants, corporate executives, and social critics either explicitly or implicitly accept a similar view of the corporation.

Empirical studies often assume the traditionalists’ view as a starting point. For example, Baldwin et al. (1986), in investigating the relationship between CSR and financial performance, wrote that the purpose of their study was to produce quantitative estimates of the penalty, as non-market risk, that investors would have to bear as a result of not being able to invest in various equity securities. The implicit assumption is that there must be a cost. The only relevant question remaining according to these authors is: Is the cost material?

Not unexpectedly corporate executives often explain and defend their economic decisions along similar lines. One example should suffice. An important national newspaper has been criticized for publishing advertisements from the Kingdom of Saudi Arabia. The paid advertisements have described a large and varied number of available professional and technical positions. Critics point out that Saudi Arabia has never concealed the fact that its laws forbid women to work with men, and indeed women are barred from most occupations (a practice which is prohibited in the United States). In response to a suggestion that the newspaper require all advertisers to declare themselves as equal opportunity employers, an executive at the newspaper replied that it would not be acceptable to refuse advertisements just because “we might disagree with the policies pursued by the countries in which advertisements are located.” He went on to state (echoing an extreme interpretation of Friedman’s position) that the policy of the newspaper was that it is “improper to use an economic sanction: the declination of advertising in furtherance of our editorial view.” (as quoted in Boycott Report 1992)

Although, Friedman’s critics often attempt to paint him as a lone-wolf, his views are arguably “mainstream”. In the introduction to his book, New Challenges to the Roles of Profit, Benjamin Friedman suggested that the traditional view is still dominant. Accordingly, he wrote that “The standard textbook view is that firms seek to make as much profit as possible within the constraints imposed by production technology (supply factors) and market conditions (demand factors). The great preponderance of scholarly research in economics … either implicitly or explicitly accepts this proposition, in order to provide a guide for determining firms’ behavior.” (1978: p. 3)

We conclude this section with what, for the purposes of our study, is the most important implication of the traditionalists’ view. It can be stated in the form of a testable hypothesis as follows:

The Traditionalists’ Hypothesis

Firms which are screened on the basis of social-responsibility will be characterized as inferior investments using traditional financial statement analysis criteria.

This hypothesis follows directly from Friedman’s observation that social responsibility, if it means anything at all, implies that the business manager “must act in some way that is not in the interest of his employers.” As stated above, however, a review of the literature is inconsistent with this hypothesis. And, therefore, although Friedman’s view may be both rigorous and somewhat convincing, it is apparently not descriptive in the sense that it is not supported by the available empirical evidence. We now turn to a discussion of these studies.

The Paradox of Social Cost: Empirical Evidence

Each of the studies discussed in this section explicitly examined the statistical relationship between perceived corporate social-responsibility and traditional financial performance. Appendix A to this report briefly describes the important characteristics of these 21 studies. In the appendix we disclose the social responsibility and financial performance criteria used in each study. In addition, we briefly summarize the main results and some additional comments.

The studies which we reviewed were published over a 20 year time span. The first study was published in 1972, and the last in 1992. The studies used a wide variety of methodologies and variables to test for an association. For example, Belkaoui (1976) compared risk- adjusted market returns of companies that disclosed pollution control information in their annual reports versus a control sample of non-disclosing firms during a 4 month period following annual report disclosure. By extreme contrast, Sturdivant and Ginter (1977) examined the difference in 10 year earnings per share growth between firms which scored high on a CSR reputational index versus low scoring firms. Both of these studies were interested in answering the same basic research question: To what degree is CSR related to traditional financial performance? However, both studies chose different ways to measure CSR (annual report disclosure versus reputational index), different ways to measure financial performance (market returns versus a financial accounting measure), and different time horizons (4 months versus 10 years).

In addition to using alternative methodologies and variables, there also exists a wide diversity in terms of industries examined. Among the industries were: chemical, electric power, food processing, iron and steel, pulp and paper, and others. Further, some of the reported studies went beyond individual industry analysis and examined inter-industry effects (see for example Cotrill 1990).

Finally, underscoring the inter-disciplinary nature of the research question, studies have been published in numerous academic journals. Five were published in the Academy of Management Journal, three in Accounting Review, two in Accounting, Organizations, and Society. The 11 remaining studies were each published in different journals ranging from Journal of Economic Studies to Journal of Business Ethics. The alternative approaches and assumptions adopted in each of the studies reflects the unique contributions of each of the academic disciplines that have participated in this research including: accounting, business ethics, economics, finance, and management. This diversity should mitigate problems associated with experimental deficiencies which might result from any one approach. The most important observations which suggest themselves from our literature review are listed below:

• 1. Our single most important observation is that of the 21 studies, 12 reported a positive association between CSR and financial performance, 1 reported a negative association, and 8 reported no measurable association. As opposed to Ullmann (1985), we conclude that there is a consistent pattern in terms of this association. While we agree with Ullmann, when he wrote that “conflicting results were reported even in cases based on the same sample of firms” (p. 543), we strongly disagree with his interpretation that “no clear tendency can be found.” Table 30.1 compares the results of our literature review to Ullmann’s. Notice that even according to Ullmann’s accounting, of the 13 studies which he identified as examining the relationship between CSR and financial performance, 8 found positive correlations, 1 found negative correlations, and the remaining 4 studies reported no correlations. While it is evidently true that not all studies report that CSR firms perform better than non-CSR firms, the overwhelming preponderance of the evidence indicates that CSR firms perform at least as well as other firms. We believe that this finding directly contradicts the traditionalists’ view of the corporation. This surprising empirical regularity constitutes prima facie evidence for the existence of the “paradox of social cost”. To the extent that social activities are costly to the firm (even while creating positive externalities), one would predict a negative relationship between social performance and financial performance at the individual firm level. In the next section, we will discuss five plausible explanations to these findings.

  Table 30.1 CSR and traditional financial performance: summary of 21 empirical studies – principal findings

• 2. In examining CSR performance, numerous surrogates have been employed. As per Table 30.2, of the 21 studies, 9 used a measure of environmental performance, 6 used reputational indexes, 2 each used disclosure and South African related criteria. Of the 12 studies that reported some positive association there is no predominance of any one variable. Four of the 9 studies that employed environmental performance as the CSR surrogate, and 4 of the 6 studies that used a reputational index, reported a positive association.

  Table 30.2 CSR and traditional financial performance: summary of 21 empirical studies – CSR criteria used

• 3. In addition to using alternative measures for CSR, the studies have also employed a wide variety of measures for financial performance. Table 30.3 provides additional details. Note that 6 of the 21 studies focused solely on financial accounting returns, 7 based their results on market based returns, and still others used multiple criteria. The 12 studies which reported positive associations are not driven by any one variable. For example, 3 of the studies which reported positive associations were based on accounting data alone, and 4 each were based on either market data alone or multiple criteria.

  Table 30.3 CSR and traditional financial performance: summary of 21 empirical studies – financial performance criteria used

• 4. The observation that researchers employed many different methodologies is corroborated by Table 30.4. Six of the studies examined CSR and financial performance of more than 61 firms, 7 of the studies included less than 20 in their sample. There are advantages and disadvantages to both large and small sample studies. For example, small sample sizes may result in better estimates of CSR. Large sample sizes will minimize sample bias. We simply observe, as above, that methodological diversity should mitigate problems associated with experimental deficiencies which might result from any one approach.

  Table 30.4 CSR and traditional financial performance: summary of 21 empirical studies – sample size

To conclude this part of our discussion, we note two possible limitations in interpreting and generalizing the results. First, most of the studies were relatively short term in nature. Only 5 of the 21 studies examined more than 5 years of data. The implications of this deficiency will be discussed in the next section. Second, only 6 studies examined data after 1975 (even though 13 studies were published after 1979). There exists, therefore, a need to update some of these earlier studies. Nevertheless, the body of work reviewed here represents a strong case against the traditionalists’ conception of CSR. Friedman observed that CSR, if it means anything at all, implies that the business manager “must act in some way that is not in the interest of his employers.” We suggested that this view implies the following formal hypothesis:

The Traditionalists’ Hypothesis

Firms which are screened on the basis of social-responsibility will be characterized as inferior investments using traditional financial statement analysis criteria.

Most studies to date have presented evidence which is inconsistent with this hypothesis. We next turn to an extensive discussion of five plausible explanations to this seeming paradox.

The Paradox of Social Cost: Five Explanations

We discuss five possible explanations to the empirical results presented in the previous section. The explanations should not be viewed as mutually exclusive. In fact, each explanation provides additional insight into the nature of CSR, and thus provides a more realistic understanding of a complex phenomenon.

Explanation 1

Explanation 1 – Socially-responsible firms are identical to non-socially-responsible firms.

As we emphasize throughout this study, the notion of socially-responsible investing is often a vague and ill-defined term. It is almost impossible to provide a precise definition. Further, social-responsibility is always a function of perception. Even if there is a growing consensus on a number of issues like environmental concern or employee relations, there is still enough disagreement that all general observations about the degree of CSR will be met with some opposition. It is therefore tempting to suggest that because of the uncertainty surrounding definitions of CSR there is no such thing as CSR, and therefore firms which may have been identified as socially-responsible are, in fact, no different from other, non-socially-responsible firms. If this proposition holds, then the paradox described in the previous section disappears.

There is ample anecdotal evidence which is consistent with this explanation. For example, the New York Times recently reported (February 11, 1993) that the Sun Oil Company of Philadelphia, the 12th-largest oil company in the United States became the first Fortune 500 company to endorse the Valdez Principles (or at least a watered-down version in which some of the original principles were negotiated). These principles are a code of corporate environmental conduct which were devised following the 1989 Alaskan oil disaster. According to Robert H. Campbell, Sun’s chairman and chief executive, there is a tremendous “philosophical congruence” (as quoted by the N. Y. Times) between what Sun already does and the environmental principles. In fact, Campbell emphasized, at the signing ceremony, that he did not foresee any major changes in company operations. To the extent that Sun Oil’s observations are generalizable to other oil companies and other industrial corporations, one should not anticipate any negative financial repercussions following the signing of the Valdez Principles.

That there may be minimal direct costs associated with CSR (relative to the size of the corporation) is suggested in a recent comment in Prudential’s annual report. Robert Winters, the CEO, wrote that the firm considers social-responsibility “critical to our success.” He further disclosed that, “The Prudential Foundation gave more than $16 million to various worthy causes.” The discussion about CSR in the CEO’s letter consumes about 10% of the total disclosures (in terms of paragraphs), whereas the $16 million charitable contribution consumes less than one tenth of 1% of reported net income.

Support for the idea that there is no difference between socially-responsible firms and other firms is the possibility that all major U. S. corporations who abide by the law are by definition socially-responsible. Regulatory requirements, and the constant threat of increased regulatory actions, coupled with an increasingly hostile tort system, may provide ample incentive for U. S. corporations to engage in socially-responsible behavior. Any attempt therefore to distinguish between socially-responsible firms and other firms is essentially arbitrary. At least in the area of pollution control there is some evidence to support this conjecture. Shane and Spicer (1983), in studying pollution ratings produced by the Council on Economic Priorities, stated that the Council’s criteria “to rate the overall efficacy of companies’ pollution-control systems correspond fairly closely to legislative requirements promulgated under the Clean Air Act Amendments … and the Federal Water Pollution Control Act.” (p. 524)

There is no doubt that in some instances CSR is nothing more than self-advertising. On the other hand, there is also no doubt that this explanation is not complete. There are often substantial costs associated with CSR behavior. For example, Freedman and Jaggi (1982) reported that in highly polluting industries as much as 20% of the total amount of capital expenditures have been devoted to pollution abatement. Belkaoui (1976) suggested that in the steel industry the percentage may reach as high as 25%.

While it is true that some of the surrogates that have been used to measure CSR are not precise, it is extremely unlikely that there are no differences between firms that are perceived as having met CSR criteria and others. It is unlikely that all, or even most, of the attempts to distinguish between socially-responsible firms and non-socially-responsible firms have been meaningless. The ability to obtain information about socially-responsible actions has become less difficult. Rockness and Williams (1988) surveyed managers of socially-responsible mutual funds about sources of “social information.” Among the most important sources of information were the companies themselves and government agencies. In addition, private social responsibility organizations like Franklin Research, and Investor Responsibility Research Center were also mentioned. In total, the authors listed 39 different sources of social information which were cited by at least one fund manager.

At the same time CSR information has become easier to obtain, mutual funds which advertise themselves as socially-responsible have begun to define the practice with more and more exactness. Table 30.5 summarizes both the positive and negative screens used by nine of the most important and influential socially-responsible mutual funds. Issues like environmental concern, South Africa, weapons production, and employee relations were cited by almost all of the mutual funds examined.

Table 30.5 Social responsibility screens used by nine mutual funds

Therefore, if there exists real differences between socially-responsible firms and other firms, the original question remains. We now turn to a second plausible explanation.

Explanation 2

Explanation 2 – The experiments to test the association between CSR and traditional financial performance have not been carefully designed or controlled.

According to this explanation, socially-responsible firms may not be identical to non-socially-responsible firms. However, the experiments to test the association between CSR and financial performance have not documented an inferior performance for CSR firms because the tests have not been well-designed.

For example, Vance (1975) argued that earlier association studies had not been “validated.” His main concern was the that earlier studies adopted an extremely short window to measure financial performance. Cochran and Wood (1984) further noted that earlier studies (including Vance) lacked methodological rigor in the sense that they failed to measure “risk-adjusted” returns. Ullmann (1985) concluded that “studies of the relationship between social performance and economic performance are highly questionable when social disclosure is used as a proxy for social performance.” (p. 545)

Cochran and Wood suggested that reputational indexes used to measure CSR are “highly subjective and thus may vary significantly from one observer to another.” (p. 43) Further, and perhaps a more severe criticism, is whether or not the reputational indexes are even purporting to measure CSR. For example, at least two studies have used Fortune magazine’s annual survey of “corporate reputations” as the surrogate for CSR (McGuire et al. 1988; Cotrill 1990). The appropriateness of this measure can be questioned given that of the eight key attributes respondents were queried about to determine corporate reputations, arguably, only two were directly related to issues of CSR.²

In addition, it has been suggested that reported results may be a function of “spurious correlations.” Chen and Metcalf (1980) criticized an earlier study which documented a positive association between pollution control records and financial performance by stating that the earlier “evidence rests on spurious relationships created through one or more intervening variables. The reported significant associations might not have been observed had the effect of intervening variables been controlled (or adjusted).” (p. 168) Chen and Metcalf showed that when they controlled for size, the positive association between CSR performance and financial performance is eliminated. In their words, “The results indicate that the conclusion of a moderate to strong association between pollution control record and financial indicators is not justified.” (p. 174)

Roberts (1992) further suggested that, in general, many of the studies in this area are merely “ad hoc” attempts to relate corporate social responsibility actions to selected corporate characteristics. Roberts suggested that the earlier work lacked a “theoretical foundation.” (p. 610) Ullmann (1985) made a similar point when he explained that, “The generally ambiguous nature of the results of the studies surveyed in the previous sections suggests that the models may be incompletely specified.” (p. 551)

In spite of these important criticisms, the possibility of methodological limitations is by no means a complete explanation. As stated in the previous section, the overwhelming preponderance of the evidence indicates that CSR firms perform at least as well as other firms. Examining the observations delineated above, there is no reason to believe that a systematic bias has been introduced. We therefore turn to a third possible explanation.

Explanation 3

Explanation 3 – A conscious pursuit of corporate social-responsibility goals causes better financial performance.

This third possibility represents an alternative view to the traditional conception of the business enterprise. The prediction that social-responsibility might lead to better firm performance cuts across the ideological spectrum. Variants of this positions have been boldly articulated by conservative thinkers (including George Gilder, Michael Novak, and Irving Kistol), centrists (including Arthur Okun and Clarence Walton), and by radical writers (including the sociologist Severyn Bruyn, and economists like Samuel Bowles and Herbert Gintis). This idea has also been periodically suggested by empirical researchers and corporate executives, as well.

In stark contrast to Adam Smith’s view, George Gilder celebrated the role of the entrepreneur (1984).

Even if we do not ask economists to perform as moral philosophers, we should demand that they accurately observe the world. Observing the world, one can see scarce factual foundation for the prevailing view of entrepreneurial activity. The capitalist is not merely a dependent of capital, labor, and land; he defines and creates capital, lends value to land, and offers his own labor while giving effect to the otherwise amorphous labor of others. He is not chiefly a tool of markets but a maker of markets; not a scout of opportunities but an inventor of them; not a respondent to existing demands but an innovator who evokes demand; not chiefly a user of technology but a producer of it. He does not operate within a limited sphere of market disequilibria, marginal options, and incremental advances. For small changes, entrepreneurs are unnecessary; even a lawyer or bureaucrat would do. (p. 17)

He concluded this discussion by emphasizing, “It is the entrepreneurs who know the rules of the world and the laws of God. Thus they sustain the world. In their careers, there is little of optimizing calculation, nothing of delicate balance of markets … They are the heroes of economic life.” (p. 19)

Michael Novak (1984) has also criticized the traditional views. In reviewing the theories about democratic capitalism inherited from Adam Smith, Jeremy Benthan, Ludwig von Mises, Frederik von Hayek, and Milton Friedman, he wrote,

The typical mistake of classic thinkers on this subject is to have laid too small a foundation to support the lived world of democratic capitalist society as we have experienced it. They have too chastely considered the economic system in abstraction from the real world, in which the political system and the moral-cultural system also shape the texture of daily life. (p.36)

Accordingly, he described a central element of democratic capitalism, “virtuous self-interest”, as follows:

The laws of free economic markets are such that the real interest of individuals are best served in the long run by a systematic refusal to take short-term advantage. Apart from internal restraints, the system itself places restraints upon greed and narrowly constructed self-interest. Greed and selfishness, when they occur, are made to have their costs. A firm aware of its long-term fiduciary responsibilities must protect its investments for future generations. It must change with the time. It must maintain a reputation for reliability, integrity, and fairness … Thus a firm committed to greed unleashes social forces that will sooner or later destroy it. Spasms of greed will disturb its own inner disciplines, corrupt its executives, anger its patrons, injure the morale of its workers, antagonize its suppliers and purchasers, embolden its competitors, and attract public retribution. In a free society, such spasms must be expected; they must also be opposed. (p. 93)

Among the so-called neo-conservatives, Irving Kristol has also voiced concern over the traditional view of the corporation, especially as advocated by Friedman. (See Two Cheers For Capitalism 1978, pp. 63–64.) In discussing the rationale for corporate philanthropy, Kristol recognized that the only justification for corporate charity (as distinct from individual charity which “refines and elevates the soul of the giver” p. 134) is that it must “serve the longer-term interests of the corporations.” He continued, “Corporate philanthropy should not be, and cannot be, disinterested.” (p. 134) Kristol’s view is consistent with the possibility that a conscious pursuit of corporate social-responsibility goals (Kristol himself used the term “social responsibility” to describe controllable philanthropic expenditures) may cause better financial performance, especially in the long run.

Arthur Okun concluded his book, Equality and Efficiency, by stating, “the market needs a place, and the market needs to be kept in its place.” (1975: p. 119) Even while recognizing the limitations of a market-based system, Okun justified the profit motive along the lines we are discussing here. In defending his belief that a reliance on self-interest is not offensive as an organizing principle for the economy, he wrote that “self-interest is consistent with an enlightened selfishness that creates loyalties to family, community, and country, as institutions that benefit the individual and extend his range of interests.” (p. 49)

Clarence Walton, one of the earliest proponents of CSR similarly noted that “Corporations will be around a long time and durable organizations exist by doing things right – right in the fullest sense of the word.” (1992: p. 60)

At the other end of the ideological spectrum, more radical theorists have, from time to time, also entertained the possibility that social-responsibility may lead to better financial performance. Bowles and Gintis (1987) suggested that democratically controlled firms may be more efficient than the traditional corporate form of organization. This prediction is suggested by the possibility that “the change in the locus of command” that would be necessarily a part of a democratically controlled firm “may be expected to reduce the wage and surveillance costs of generating a given level of labor performed.” (p. 78)

Severyn Bruyn has also predicted a positive link between social performance and economic performance. Unlike the traditional perspective, he dismissed the notion that there must be a tradeoff between them, rather the relationship between CSR and financial performance is a synergistic one. Bruyn (1987) wrote:

In reality, social considerations in the investment process can actually enhance the possibilities of economic return. The fact is that the two values are not necessarily exclusive. Social and economic values can be maximized together, this creative synergism is the practical direction taken by social investors today. (p. 12)

The possibility that the association between CSR and financial performance may be the result of a causal relationship, as discussed here, has also been periodically suggested by empirical researchers, as well. In presenting evidence that CSR firms in the food-processing industry outperformed non-CSR firms, Bowman and Haire (1975) explained that while there is not a one to one relationship between CSR and financial performance, nevertheless CSR is “a signal of the presence of a style of management that extends broadly across the entire business function and leads to more profitable operation.” (p. 54) The authors continued that “it is exactly this ability to sense, adapt, negotiate with, and cope with these forces that is … the sign of managerial excellence and hence profitability.” (p. 54)

Sturdivant and Ginter (1977) provided evidence that socially-responsible firms (as measured by a reputational index) outperformed a control sample in terms of 10 year earnings per share growth. They elaborated:

It would appear that a case can be made for an association between responsiveness to social issues and the ability to respond effectively to traditional business challenges . . . A company management group which reflects rather narrow and rigid views of social change and rising expectations might also be expected to respond less creatively and effectively in the traditional but also dynamic arenas in which business functions. Hence there is the stronger economic performance . . . (p. 38)

Kahneman et al. (1986) have provided survey evidence which supports the causal link between CSR and financial performance. They argued that a realistic description of our economic system must include the fact that consumers, suppliers, and employees care about being treated fairly and treating others fairly. In addition, they are willing to resist unfair firms even at a positive cost to themselves. Satisfying the “fairness constraint” may lead to better long-run financial performance.

Executives have attempted to describe the connection between CSR and financial performance through the vehicle of the annual report. For example, the president of Ben and Jerry’s Homemade Inc. recently defended his commitment to a social agenda in his president’s letter to shareholders as follows:

We have a two-part bottom line. This Annual Report presents both our financial progress and our progress in contributing to the quality of life in our communities … We believe that if we focus on the quality of everything we do, the traditional business measures will fall into place. We are master ice cream and frozen dessert makers. We want to be a force for progressive social change. And our staff is perpetually enthusiastic about our future. If we can continue to grow these values as fast as we grow the company we’ll be fine.

In summary, the view discussed here is a powerful countervailing paradigm to the traditional view of the corporation. Further, it is apparently more consistent with the available empirical evidence than the alternative view. Nevertheless there are major limitations. First, it is highly doubtful whether the variables which have been used as surrogates for CSR in the empirical studies are always closely related to the notions of “social-responsibility” which have been emphasized by Gilder, Novak, Okun, and, even Bowles and Ginter. We observe here that while there is a clear overlap between CSR as it has developed in practice over the last 20 years and the notions of responsibility as discussed in this section, the overlap is not exact. Therefore the explanation offered here may not be entirely appropriate for the empirical findings previously reported.

Second, intuitively, the explanation is not completely compelling. Simply put, if doing good is always costless, why isn’t everyone good? By the logic offered here, even a scoundrel would eventually notice that it is in his or her best interest to choose CSR. We therefore need a view which can explain the persistence of scoundrels, as well as saints.

Finally, the explanation as stated here is too general. In Explanation 5, below, its scope is limited. First, however, we discuss the following alternative explanation.

Explanation 4

Explanation 4 – Only firms which perform better in terms of financial criteria can afford a conscious pursuit of corporate social-responsibility goals.

Social-responsibility does not cause enhanced financial performance, but rather, financial performance allows for the performance of discretionary social actions. Anecdotal evidence supports this view For example, in response to poor financial performance, firms with “no layoff” policies have been forced to shrink their employee base. What was once viewed as a permanent part of corporate strategy to meet corporate social-responsibility goals is no longer economically viable.

According to this view, especially as it has been articulated by Ullmann (1985), the motivation for engaging in socially-responsible actions is external to the corporation. Ullmann suggested that social performance should be viewed as a result of a “strategy for dealing with stakeholder demands.” (p. 552) He continued, “When stake-holders control resources critical to the organization, the company is likely to respond in a way that satisfies the demands of the stakeholders.”

A central component to Ullmann’s “stake-holder model” is the link between financial performance and social-responsibility. In Ullmann’s view, economic performance is posited as an independent variable. Therefore, economic performance explains CSR, and not vice versa. “Economic performance determines the relative weight of a social demand and the attention it receives from top decision makers. In periods of low profitability and in situations of high debt, economic demands will have priority over social demands . . . Economic performance influences the financial capability to undertake costly programs related to social demands.” (p. 553)

McGuire et al. (1988), following Ullmann, concluded their empirical study by noting that “Firms with high performance and low risk may be better able to afford to act in a socially responsible manner.” (p. 869) Echoing Ullmann, they continued, “In essence, it may be more fruitful to consider financial performance as a variable influencing social responsibility than the reverse.”

Chen and Metcalf (1984) examining the relationship between pollution control and financial performance similarly suggested that “economically, a firm with high earnings is more likely to incur pollution abatement costs than one with low earnings.” (p. 173)

More recently, Roberts (1992) in presenting empirical evidence which is consistent with Ullmann’s stakeholder model, concurred that it is economic performance which leads to higher levels of CSR and not the other way around.

The importance placed on meeting social responsibility goals may be secondary to meeting the economic demands that impact directly on a company’s continued viability. Economic performance directly affects the financial capability to institute social responsibility programs. Therefore, given certain levels of stakeholder power and strategic posture, the better the economic performance of a company the greater its social responsibility activity and disclosures. (p. 599)

Ullmann’s stakeholder model is consistent with the traditional view of the corporation in the sense that both view social responsibility as a net cost to the corporation. In addition, a benefit of the stakeholder model is that it is compatible with much of the empirical evidence which was reviewed above.

We believe that this approach represents an important development in understanding the nature of CSR. Effective managers need to satisfy all important stakeholders, not simply the demands of shareholders. Further, it is plausible to assume that meeting the needs of consumer groups, environmental activists, labor unions, the government, and other stakeholders is becoming more important to corporate managers. Nevertheless, it may not be accurate to suggest that the demands for social-responsibility are always external to the corporation, as the stakeholder model (as developed by Ullmann) implies.

Further, an important and unappreciated implication to the stakeholder model is that if there is a net cost to CSR, in the long run it should be detected. In other words, firms which start out with a financial advantage and can therefore afford to engage in socially-responsible actions, should over time (assuming they continue to engage in CSR) forfeit their financial advantage. That CSR activities represent a material cost is directly suggested in Ullmann’s observation quoted above that economic performance influences the financial capability to undertake costly programs related to social demands. This testable implication of the stake-holder model has never been formally examined. Therefore, one of the main goals of the current research is to examine the long run financial performance of a group of socially-responsible firms.

Explanation 5

Explanation 5 – Sometimes, a conscious pursuit of corporate social-responsibility goals causes better financial performance.

Explanation 3, as suggested above, it too extreme. Explanation 5 limits its applicability. According to this last explanation, there are two types of socially-responsible actions. Some social actions have no net costs, and in fact may benefit the firm in the long run, while other socially-responsible actions (even while creating positive externalities) are costly to the firm. This explanation suggests that the traditional view (and Explanation 4 above) is wrong in assuming that social actions do not benefit the firm. The position adopted here proposes that Friedman’s statement that the very term “social responsibility” must imply behavior that is not in the interest of the corporation is needlessly provocative. (See Friedman 1970: p. 33.) Our disagreement with Friedman is a definitional one. Friedman’s view is that any action which benefits the firm is, by definition, not “socially-responsible.” Alternatively, we suggest that whether or not an action benefits the firm (in terms of increased financial performance) is irrelevant to its classification as “socially-responsible.”

If Explanation 5 is to help unravel the paradox of social cost, we must add the plausible assumption that the major corporations which have been studied in the empirical literature, and which are perceived as being socially-responsible, are pursuing corporate goals which are consistent with financial performance goals. Corporate management, on average, rejects those activities which are not congruent with shareholder demands. Under this assumption, we do not anticipate a negative association between CSR and financial performance.

The possible existence of two types of social actions, although intuitively appealing, has received little attention. The important advantages of this explanation are that:

• 1. it is consistent with the empirical studies examined above,

• 2. it does not assume that the motivation for CSR is always external to the firm (as in Explanation 4), and

• 3. it is consistent with the views of corporate executives and board members.

The explanation offered here is based, in part, on Peter Drucker’s definition of corporate social-responsibility (1989). In his book The New Realities, Drucker wrote:

We know in rough outline the social responsibility of the pluralist institutions of society. We know that their first social responsibility is to do their job. We know secondly that they have responsibility for their impacts – on people, on the community, on society in general. And finally we know that they act irresponsibly if they go beyond the impacts necessary for them to do their own job, whether it is taking care of the sick, producing goods, or advancing learning. (p. 86)

In describing the responsibility for organizational impacts Drucker amplified:

It has to exercise considerable control over the people who work for it; otherwise, it cannot do its job. It has considerable impact on people who are customers whether they buy a company’s goods or are patients in a hospital. And it has impacts on bystanders. The factory that closes at four-thirty in the afternoon creates a traffic jam for everyone in the community. Responsibility for one’s impacts is the oldest principle of the law. It does not matter whether the institution is at fault or is negligent. The Roman lawyers who first formulated this principle called it the “doctrine of the wild animal.” If the lion gets out of its cage, its keeper is responsible. Whether the lion’s keeper was careless and left open the door of the cage, or whether an earthquake released the lock, is irrelevant. (pp. 87–88)

Drucker’s “doctrine of the wild animal” thus insures that the “institution has a duty – but also a self-interest – to limit its impact to what is actually needed for the discharge of its social function.” (p. 88)

Consistent with Explanation 5 above, Drucker also underscored the existence of two types of socially-responsible actions. He pointed out that social-responsibility is effective only under stringent conditions. It must fit the organization’s value system. “It must be an extension of what it is doing rather than a diversion.” (p. 92) In a recent article, Drucker (1992) continued on the theme of social responsibility. He wrote that “we had better be watchful because good intentions are not always socially responsible. It is irresponsible for an organization to accept – let alone pursue – responsibilities that would impede its capacity to perform its main task and mission or to act where it has no competence.” (p. 99)

In addition to Drucker, a number of attempts have been made to distinguish between socially-responsible actions which lead to better financial performance and those that do not. It is often assumed that there may be a link between pollution control and financial performance. For example, in discussing the compatibility between high levels of pollution control and high profit levels, Bragdon and Marlin (1972) suggested that the poor performance of the domestic steel industry must be viewed as a consequence of poor management. They believed that “good managements are likely both to earn higher profits and to be more careful in protecting the environment.” (p. 10) According to their view, while Japanese and European firms were investing in new equipment with lower pollution levels, American steel companies refused to change over to the new technology. That foreign companies have outperformed domestic steel producers is in part a “reflection of lower costs associated with better pollution control.” (p. 9)

Coffey and Fryxell (1991) in suggesting that corporate social-responsibility involves taking actions pursuant to obligations beyond the economic and legal sphere, isolated four components of CSR that may lead to better firm performance. “Evidence of corporate social responsiveness may be related to a broad range of issues including: pollution abatement, product safety, advertising messages, the role of women and minorities in the firm. That the capability to change with social climate is important for long-term economic performance is a basic tenet of strategic management.” (p. 439)

A major limitation to Explanation 4 above is its insistence that the motivation for CSR is always external to the organization. This observation is plainly seen in Ullmann’s prediction that firms with poor economic performance, low stakeholder power, and a passive strategic posture are not likely to engage in CSR. (p. 553) Explanation 5, however, predicts that even in a period of poor economic performance, a corporation may find it in its own interest to pursue CSR objectives. The motivation for CSR can thus be an internal decision to increase long term financial performance, while simultaneously meeting responsibilities for corporate impacts.

Cornell and Shapiro (1987) further explored this possibility. What are the advantages – to the shareholder – of honoring product warranties beyond legal requirements? Cornell and Shapiro suggested that what motivates corporate executives to honor implicit contracts (with no legal ramifications) is that executives’ believe that the long term value of the firm is a direct function of its ability to sell (not only explicit claims) but also implicit claims. In the authors words, the market value of a corporation includes “organizational capital which equals market value of all future implicit claims the firm expects to sell.” (p. 10)

To clarify the distinction between implicit and explicit claims the authors used the following example:

The price at which IBM’s PC_jr, included both the price of the hardware and the prices of the implicit claims for future support, software, product enhancements, and the like. As it became clear that PC_jr’s success in the market was limited, IBM faced a difficult decision. If the company chose to discontinue the product line it would clearly lessen the organization liabilities connected with PC_jr. On the other hand, discontinuing the product reduces the payout on implicit claims previously issued by the company, which in turn reduced the firm’s organizational capital by causing the prices of future implicit claims to fall. (p. 9)

The problem that IBM and other corporations face is that if they fail to honor implicit claims for one product, stakeholders will rationally assume that they are less likely to honor implicit claims for other products, including items yet to be marketed. “For firms such as IBM that choose to identify all their products with the company name, the spillover effect is likely to be particularly strong.” (p. 9)

In this example, IBM chose what we might label the socially-responsible solution. The company chose to discontinue the production of PC_jr but it also undertook a major advertising campaign to let PC_jr owners and other stake-holders know that “If you own a PC_jr you can be sure it is still a well-cared for member of the IBM PC family.” (p. 10) They chose this solution, not out of a sense of altruism, but because of concern with their long term financial performance.

By contrast, when Exxon phased out its office systems division, Exxon “provided minimal support for customers and other stakeholders of that division.” (p. 9) Presumably, Exxon executives perceived little spillover effect as a result of this decision, as the office systems division was incidental to their main line of business.

Cornell and Shapiro extended their analysis beyond customer warranties. They wrote:

When a firm hires a new employee, he or she frequently receives promises about the work environment, the evaluation process and the opportunity for advancement, as well as an explicit employment contract … In a similar fashion, implicit claims are sold to stakeholders, such as suppliers and independent firms that provide repair services and manufacture supporting products. (pp. 6–7)

To conclude this section, we note what for our purposes is the most important implication of Explanation 5. Explanation 5 states: SOMETIMES, a conscious pursuit of corporate social-responsibility goals causes better financial performance. If this explanation holds, and if we add the plausible assumption that the major corporations choose, on average, to pursue those CSR goals consistent with financial goals, in the long run, socially-responsible firms may actually outperform non-socially responsible firms in terms of traditional financial performance. Firms identified as socially-responsible, should maintain, or even increase, their relative financial advantage over non-socially-responsible firms.³ This implication, is in direct opposition to the implication of Explanation 4 above. According to Explanation 4: Only firms which perform better in terms of financial criteria can afford a conscious pursuit of corporate social-responsibility goals. As we pointed out above, this implies that firms which start out with a financial advantage and can therefore afford to engage in socially-responsible actions, should over time (assuming they continue to engage in CSR) forfeit their financial advantage.

Section 2

A major goal of this study is to explore the association between CSR and traditional financial performance. In this way, we can begin to distinguish between Explanations 4 and 5 discussed in the previous section.

Creating the Sample

In particular, we examine the long-term financial performance of a group of 53 firms which have been identified by the Council on Economic Priorities (CEP) as being socially–responsible (GROUP 1), and compare their performance to a control sample matched by both industry and size (GROUP 2). In addition, to test for changes over time, we compare the relative performance of the GROUP 1 and 2 firms in two time periods (1985–1987 and 1989–1991). A listing of the 106 firms selected for our study is included in Appendix B.

The CEP described the companies in GROUP 1 as “ethical” portfolio companies. The advantages of choosing the CEP firms for our study are as follows⁴:

• 1. The CEP is highly regarded as a credible source of information on CSR. Numerous published studies have used previous CEP studies as the basis for forming measures of CSR. For example, of the 21 studies we reviewed in the previous section, 5 used CEP studies. We concur with Shane and Spicer (1983) who concluded that “The most detailed, consistent, and comparable data bearing on corporate social performance has been published by the CEP. It appears to be the most active external producer of information in this area.” (p. 522)

• 2. The CEP ratings are not unique. The firms included in GROUP 1 tend to be rated high in terms of CSR by numerous external groups. Table 30.6 summarizes some characteristics of the GROUP 1 and GROUP 2 firms, and provides additional support to the CEP ratings. There is significant overlap between the GROUP 1 firms, as identified by the CEP, and firms included in the Domini 400 Social Index. Only 12 of the GROUP 2 firms were included in the Domini Index. About half the GROUP 1 firms (24 firms) were rated among the “100 Best Companies to Work For”, while only 2 of the GROUP 2 firms were included on this list. Further, 12 of the GROUP 1 firms were among the “75 Best Companies for Working Mothers”, and none of the GROUP 2 firms were identified among the “75 Best Companies for Working Mothers”.

  Table 30.6 Characteristics of Groups 1 and 2 firms

  Table 30.6 also indicates that few of the GROUP 1firms are listed among the “Top 100 Defense Department Contractors”, or among the “Top 50 Manufacturers Releasing Toxic Chemicals”. Finally, and not surprisingly, none of the GROUP 1 firms were included among the “Top 100 Nuclear Weapons Contractors,” or were identified as “Tobacco Companies”.

• 3. To achieve the goals of this study we needed an aggregate measure of CSR, as opposed to a measure of one or more of the components of CSR. The CEP ratings, based on an assessment of 12 specific CSR components, provided a convenient and well respected third party assessment. Further, we believe that the CEP ratings provide a more precise measure of CSR, per se, than those obtained from the next best competitor, Fortune magazine’s annual survey of “corporate reputations”. As discussed in the previous section, the appropriateness of the survey results can be questioned given that of the eight key attributes respondents were queried about to determine corporate reputations, arguably, only two were directly related to issues of CSR.

• 4. The GROUP 1 firms were selected from diverse industries, thus enhancing the generalizability of the results. Table 30.7 reveals that 21 industries are represented among the 53 GROUP 1 firms. Nine firms were selected from both Food and Kindred Products (SIC codes 2000–2099) and Chemicals and Allied Products (SIC codes 2800–2899). Eleven industries had just one member among the GROUP 1 firms. The relatively large proportion of firms in Food and Kindred Products and Chemical categories might be considered prima facie evidence of an industry effect. The possibility that there exists an association between perceived social-responsibility and industry has been documented by Cotrill (1990), Bowman and Haire (1975).

  Table 30.7 Industry classifications for socially-responsible firms

To conclude this discussion, many of the 53 firms in GROUP 1 have been described as socially-responsible by a wide variety of outside evaluators. The CEP is one of the most highly regarded external producers of social-responsibility information. The 53 firms represent a diverse sample of companies. The sample thus provides an important, and inherently interesting, point of departure.

Financial Performance Criteria

We compared firm characteristics between GROUPS 1 and 2 over a broad range of traditional financial variables. The variables fall into one of four major categories. Specifically, we examined:

• A. Market-based Measures of Performance including market return, price to earning ratio, and market value to book value,

• B. Accounting-based Measures of Performance including return on assets, return on equity, and earnings per share,

• C. Measures of Risk including current ratio, quick ratio, debt to equity ratio, interest coverage, Altman’s Z-score,⁵ and market beta,

• D. Other Firm-specific Characteristics including capital investment intensity, size, number of lines of business, and dividend-payout ratio.

In all we examined and report results for 16 traditional financial statement variables. Each of the variables is constructed from data available on COMPUSTAT. (COMPUSTAT is a machine-readable data base with historical financial information for over 1,500 publicly traded corporations.) Individual year mean and median results are displayed in Tables 30.8, 30.9, 30.10, and 30.11.

Table 30.8 Market-based measures of performance Group 1 (G1) versus Group 2 (G2) 1985–1991

Table 30.9 Accounting-based measures of performance Group 1 (G1) versus Group 2 (G2) 1985–1991

Table 30.10 Measures of risk Group 1 (G1) versus Group 2 (G2) 1985–1991

Table 30.11 Other firms-specific characteristics Group 1 (G1) versus Group 2 (G2) 1985-1991

In general, our results indicate that there is little evidence that the GROUP 1 firms, that is firms screened on the basis of CSR criteria, can be characterized as inferior investments relative to the GROUP 2 firms. This finding, once again, contradicts what we called in the previous section, the traditionalists’ hypothesis. In addition, some evidence exists which supports the stronger proposition that the GROUP 1 firms can be characterized as superior investments relative to GROUP 2 firms.

A – Market-Based Measures of Performance

According to Table 30.8, the market returns for GROUP 1 were slightly better than the market returns for GROUP 2. The overall means for the 7 year period were 13.57% and 11.54%, respectively. Further, in 4 of the 7 years the GROUP 1 firms had higher returns than GROUP 2 firms. In 1986 GROUP 1 outperformed GROUP 2 firms at the 10% level of significance, and in 1988 this relationship was reversed.

There was almost no difference between the price to earning ratios for the two groups. The overall mean for GROUP 1 firms was 19.91 for the 7 year period, and 20.51 for the GROUP 2 firms. The only year in which there was a statistically significant difference at the 10% level was 1990 when the price to earning ratios were 19.91 and 15.68, favoring the GROUP 1 firms.

Among the market-based measures, the most consistent results were related to the market value to book value ratios. This ratio relates the market capitalization of the firm to the accounting valuations. The overall means for the 7 year period was 3.16 versus 2.96 for GROUPS 1 and 2, respectively. From 1987 through 1991, GROUP 1firms had a higher ratio in each year. In 1989 and 1991, the differences were significant at the 5% level.

B – Accounting-Based Measures of Performance

Table 30.9 presents the accounting-based measures of performance. These results are similar to the market-based results in indicating either no difference, or a slight advantage to the socially-responsible firms.

The first variable presented in Table 30.9 is return on assets. It has been suggested that return on assets “takes the particular set of environ-mental factors and strategic choices made by a firm as given and focuses on the profitability of operations relative to the investments (assets) in place.” (Stickney 1990: p. 161) An important characteristic of this accounting measure is that it separates financing activities from both operating and investing activities. The overall means for the 7 year period were 7.42% and 6.37%, for GROUPS 1 and 2, respectively. In 2 of the 7 years, 1987 and 1989, the GROUP 1 firms had significantly higher return on assets. In 1987, the mean for the GROUP 1 firms was 8.55% versus 6.67% for the GROUP 2 firms. Similarly, in 1989, the mean for the GROUP 1 firms was 7.98% versus 6.07% for the GROUP 2 firms. Further, in only one year, 1986, did the GROUP 2 firms outperform the GROUP 1 firms, and this difference was not statistically significant. We conclude from these results that the GROUP 1 firms were certainly no less efficient in generating income from assets in place than the GROUP 2 firms, and, in fact, were slightly more efficient.

Although return on common equity is usually correlated with return on assets, it is useful to report this variable as an additional measure of financial performance. It has been argued that return on common equity, which relates income available to common shareholders to average amount of common equity in use during a period, should be emphasized as the appropriate tool for assessing the profitability “from the view-point of an investor in a firm’s common stock.” (Stickney 1990: p. 219) Not surprisingly, the results here also indicate a slight advantage to the GROUP 1 firms. Although the overall means for the 7 year period were slightly higher for the GROUP 2 firms, this result was primarily driven by the 1990 results, which must be interpreted with care. Notice that in 1990, although the return on common equity for GROUP 2 was apparently much higher than the GROUP 1 results, the difference is not significantly different, and, in fact, GROUP 1 had a higher median. The difference in the reported means between GROUPS 1 and 2 is thus the result of statistical outliers in GROUP 2. The only significant difference was 1987 in which the GROUP 1 firms had a mean of 19.45% versus GROUP 2’s mean of 15.50%.

The last variable included in Table 30.9 is earnings per share. The overall means for the 7 year period were $2.37/share versus $2.31/share for GROUPs 1 and 2, respectively. Although the GROUP 1 firms outperformed the GROUP 2 firms in 4 of the 7 years, in none of the years were the results significant at even the 10% level.

C – Measures of Risk

Table 30.10 presents results related to traditional measures of risk. The first two variables presented in the table, the current ratio and the quick ratio, provide an assessment of the corporations’ ability to meet its short term obligations as they come due. These measures are often labelled short-term liquidity ratios. For both the current ratio and the quick ratio, the overall means for the 7 year period were nearly identical. For the current ratio the GROUP 1 mean was 1.92 and the GROUP 2 mean was 1.91. Similarly, for the quick ratio the GROUP 1 mean was 1.13 and the GROUP 2 mean was 1.10. The only significant difference (at the 10% level) was for the current ratio in 1990. The GROUP 1 mean was 1.87, which was higher, and thus slightly less risky, than GROUP 2’s mean of 1.62.

In addition to examining short-term liquidity ratios, Table 30.10 summarizes results for three long-term solvency measures: interest coverage, debt to equity ratio, and Altman’s Z-score. Each of these measures indicate the firms’ ability to meet interest payments and principal payments as they come due.

First, for the interest coverage variable, the overall means for the 7 year period were slightly higher (less risky) for the socially-responsible firms. For the GROUP 1 firms the mean was 7.77 and for the GROUP 2 firms the mean was 6.77. (These numbers show that for ever $1 of interest expense there was, on average, $7.77 and $6.77, respectively, of income before interest expense and income taxes.) In 1987, the difference between the two groups was significant at the 5% level; GROUP 1 was 9.19 and GROUP 2 was 6.40. Further, in every year, from 1987 through 1991 GROUP 1 had higher interest coverage than GROUP 2 in terms of both means and medians.

Second, with respect to the debt to equity ratio, which measures the amount of long-term debt financing in a firms’ capital structure, although there is some indication that the socially-responsible firms may be more risky, these results should not be overstated. Although the overall mean for the 7 year period is higher for the GROUP 1 firms than for the GROUP 2 firms, this result is, in part, a function statistical outliers among the GROUP 1 firms. In fact, the mean of the median results for the entire 7 year period, which are unaffected by the outliers, and arguably more relevant for our purposes, are nearly identical between GROUPS 1 and 2. The mean of the median results for 7 year period were 45.28% and 44.06%, respectively for GROUPS 1 and 2. Further, in none of the individual years were the differences between the two groups statistically significant.

Altman’s Z-Score, a weighted average of five financial statement ratios, has been found useful in predicting bankruptcy. It thus captures a different dimension of corporate risk. In interpreting the Z-Score, the lower the outcome, the greater the probability of bankruptcy. The results summarized in Table 30.10 indicate that in 5 of the 7 years the GROUP 1 scored higher than GROUP 2. The overall mean results for the 7 year period was 14.81 and 12.46, respectively.

The last variable examined in Table 30.10 is market beta. This variable compares the variability of stock returns for a given company with the variability of the stock market as a whole. Higher levels of beta more stock market variability in relation to the market. It is the only one of the 16 variables examined which consistently favored the GROUP 2 firms. Once again, however, the differences should not be over-stated. The overall mean results indicate that the GROUP 1 betas (overall mean 1.16) are about 9% higher than the betas for GROUP 2 (overall mean 1.06). Focusing on the individual year results, in each year the GROUP 1 firms had higher betas than the GROUP 2 firms. In 3 years, 1985, 1987, and 1991, these differences were significant at the 10% level.

D – Other Firm-Specific Characteristics

In addition to examining the performance and risk measures discussed above, Table 30.11 reports comparative statistics for four additional variables: capital investment intensity, total assets, number of lines of business, and dividend-payout ratio.

The capital investment intensity variable was created by deflating new capital investments each year by total assets. The results show that the GROUP 1 firms had higher investment ratios in all but 1 of the 7 years. In fact, in 1985, 1987, and 1988, the capital investment intensity variable is significantly higher (at the 5% level) for the GROUP 1 firms than it is for the GROUP 2 firms. The ratio for the GROUP 1 was 11%, 10%, and 10% respectively for 1985, 1987, and 1988, compared to 9%, 7%, and 8% for GROUP 2. The overall means for the 7 year periods are consistent with these findings.

In addition to greater investment activity, the GROUP 1 firms are also larger than the GROUP 2 firms, in terms of total assets. These results are interesting and important. By 1991, as per Table 30.11, the mean asset size of GROUP 1 firms is over $6 billion, compared to a mean asset size of over $4 billion for GROUP 2 firms. Although the disparity is mitigated somewhat in focusing on median results rather than mean results, nevertheless the difference is statistically significant at the 10% level.

The differences in terms of size between the two groups is somewhat surprising. This is especially true since our strategy in creating the control sample (GROUP 2) was to select firms “as close as possible” in terms of asset size. In some cases, however, this strategy did not result in extremely close matches. Table 30.12 illustrates the difficulty. According to the table, about 40% of the firms in Group 1 were the biggest firms in their respective industries (ranked on the basis of total assets). For example, Group 1 includes K- Mart and Johnson and Johnson. These are the number one firms in the Variety Stores industry and the Pharmaceutical Preparation industry, respectively. Because the socially-responsible firms were so big, many of the control firms, by construction, had to be smaller than their socially-responsible counterparts.

Table 30.12 Distribution of industry rankings of socially-screened firms

This size effect documented here confirms results of previous research. For example, Trotman and Bradley (1981) concluded that companies which provide social responsibility information are, on average, larger in size than companies which do not disclose this information. Arlow and Gannon (1982), in reviewing the literature, suggested that social responsiveness might be linked to such factors as industry and organizational size. Finally, McGuire et al. (1988) using Fortune Magazine’s annual survey of corporate reputations concluded that total assets were positively linked to social-responsibility reputations.

Returning to Table 30.11, it is interesting to note that even given the substantial size differences between GROUPS 1 and 2, there are no statistically significant differences between the groups in terms of number of lines of business. This is true, even though for 5 of the 6 years (data were not available for 1985), GROUP 1 had a higher mean than GROUP 2.

Similarly, there were no significant differences for the last variable examined, the dividend payout ratio. Again, this is true, even though GROUP 1 had higher means from 1988 through 1991. The overall 7 year means were 44.72% and 42.29%, for GROUPS 1 and 2, respectively.

An Examination of Time Trends

In addition to comparing the performance of the GROUP 1 and 2 firms over the entire 7 year period as above, we also test for changes over time. Specifically, we compare the relative performance of the GROUP 1 and 2 firms in two time periods (1985–1987 and 1989–1991).

To assess the relative performance, we first divided the sample into two time periods, an early and late period. Next, for the early period, we compared the 3 year means for each of the 16 variables between GROUPS 1 and 2. We repeated the identical procedure for the later period. Table 30.13 lists those variables in which there were statistically significant differences in one time period, but not the other.

Table 30.13 Trend analysis early period (1985–1987) versus late period (1989–1991)

Our analysis shows that in terms of the market-based measures of performance, risk measures, and other firm-specific characteristics there is no evidence that the GROUP 2 firms performed better relative to the GROUP 1 firms in the later period than in the earlier period. The single piece of evidence supporting the enhanced performance of the GROUP 2 firms in the later period is one of the three accounting-based measure of performance, return on common equity. There was no relative improvement for either return on assets or earnings per share. Thus, most of our data is inconsistent with Explanation 4, of the previous section, which posits a net cost associated with social-responsibility actions.

By contrast, the preponderance of evidence is that GROUP 1 firms performed relatively better than GROUP 2 firms in the later period. This is especially true for the market-based measures of performance. As indicated in Table 30.13, the market returns for the GROUP 1 firms were nearly twice as high in the later period than the market returns for the GROUP 2 firms. Similarly, the mean of the market value to book value in the later period is significantly higher for the GROUP 1 firms. Notice also, that the in early period, the price to earnings ratio favored GROUP 2, but the effect reverses in the later period. All three of these findings indicate a relatively enhanced performance of the socially-responsible firms in the later period.

In addition to improvements in market-based measures of performance, our results also indicate a relative improvement in one of the risk measures. The interest coverage ratio, which is not significantly different between the two groups in the early period, is significantly higher (less risky) for GROUP 1in the later period. None of the other risk measures indicate any changes over time.

Finally, examining other firm-specific characteristics, Table 30.13 shows that, although there is no significant difference in terms of size in the early period between the two groups, in the later period, the socially-responsible firms are significantly larger than the control sample. The only other variable which showed changes over time was the capital investment intensity variable. In the early period, GROUP 1 had significantly higher investment ratios than GROUP 2.

To summarize the results of this section, most of the variables, although not every variable, showed either no change between the two time periods, or indicated an improved performance over time for the socially-responsible firms relative to the control sample. It is certainly pre-socially-responsible firms perform better over time.

The issues involved in assessing socially-responsible actions, and measuring financial performance are too complex and nuanced to expect definitive answers. However, based on our results to date, and to the extent they are corroborated by additional studies using alternative samples, and even longer testing periods, Explanation 5 above, becomes more plausible. Recall, Explanation 5 suggested that, at least, some CSR activities might cause better financial performance. A relative improvement in the performance of socially-responsible firms over time is consistent with this hypothesis.

The tone of our discussion and the formulation of the conclusions to our empirical work are purposely tentative. This underscores the exploratory nature of the research project. Nevertheless, the consistency of the results reported here, and the persistent finding, across numerous studies, that socially-responsible firms certainly perform no worse, and perhaps, perform better than non-socially responsible firms, is an important and intriguing finding which demands additional attention. Although our understanding of the relationship between corporate social-responsibility and traditional financial performance is not complete, in the next section we conclude our study with a general discussion of six propositions about corporate social-responsibility.

Section 3

The specific purpose of our study has been to explore the association between corporate social-responsibility and traditional financial performance. In this study we examined the long-term financial performance of a group of 53 firms which have been identified by the Council on Economic Priorities (CEP) as being socially- responsible, and compared the financial performance of this group to a control sample matched by both size and industry (Council on Economic Priorities et al. 1991). Many of the 53 firms in GROUP 1 have been described as socially-responsible by a wide variety of outside evaluators. The CEP is one of the most highly regarded external producers of social-responsibility information. The 53 firms represent a diverse sample of companies. Thus the sample provides an important, and inherently interesting, point of departure.

Our analysis of the data suggested that there is almost no evidence that firms which are screened on the basis of social-responsibility criteria performed worse than other firms. By contrast, there is some evidence to suggest a positive association between social-responsibility and traditional financial performance. Further, there was little to suggest that the control sample performed relatively better in the later period compared to the social-responsibility group. In fact, once again, most of the evidence suggested that the socially-responsible firms performed relatively stronger in the later period. This was particularly true for the market-based measures of performance, but also held for one risk measure. In terms of other firm characteristics, the evidence showed that the socially-responsible firms were significantly larger than the control group in the later period.

This concluding section extends our discussion of CSR. We examine six formal propositions. Although there will continue to be constructive debates about many of the specific issues concerning the relationship between social activities and financial performance, the following general propositions are warranted.

1

1 – Managers, board members, and investors are increasingly confronted by business decisions with social and therefore ethical implications.

As our economy begins to spill over domestic borders, as corporations continue to expand in size, as technological impacts multiply, society’s well-being becomes more tightly linked with corporate decision-making. Simply put, as corporate power increases, the ramifications of its actions multiply. Many of the most important ethical decisions individual face are with-in the corporate context. The executive decisions which ultimately lead to the Exxon Valdez disaster were surely not only economic decisions (even if that is how they were framed by the principal actors) but also involved an ethical component, as well. The question of whether or not a beer distillery should specifically target urban areas for a high-alcohol malt liquor must be answered both with economic and ethical criteria. The decision to continue marketing or to withdraw a record album advocating the killing of police officers, regardless of its solution, demands recourse not only to profit considerations, but also requires a formulation, and at least, an implicit understanding of corporate obligations to society.

We emphasize this point, although it would seem obvious, because it is apparently not universally accepted. Milton Friedman, for example, has explicitly stated that “The really important ethical problems are those that face an individual in a free society.” We, of course, do not disagree that individuals face importance ethical problems, but we believe that more and more ethical problems are faced by individuals within the corporate context.

2.

2. – A difference of opinion regarding social and ethical obligations does not prove that CSR is unnecessary or perhaps (as some might suggest) meaningless.

Arguments against managers adopting CSR criteria often take the following form: Since it is obviously true that well-intentioned individuals disagree about CSR issue X, managers must, therefore, disregard issue X in formulating business policy Y. Once again, Friedman provides the clearest and most unequivocal formulation of this position. Friedman has written, “If businessmen do have social responsibility other than maximizing profits for stockholders, how are they to know what it is?” (1962: p. 133) Friedman’s succinct formulation captures one of the most difficult aspects related to CSR.

Advocates of this argument point out that managers are hired exclusively to maximize profits. Further, they have no special expertise in evaluating ethical considerations. Managers must therefore avoid arbitrarily usurping corporate funds in pursuit of subjective personal goals. An executive who pursues issue X is there-fore in violation of his or her contract with employers.

On proposition 2, we offer three observations: First, as has often been observed by philosophers, an awareness and recognition of diversity of opinion and practice with regard to ethical issues does not imply ethical relativism. For example, the philosopher Robert Nozick (1981) has shown that, although it is not the intention of philosophy to produce uniformity of belief, nevertheless good reasons can be put forth to show how there can be objective values and ethical truths.

If the traditionalists’ point is merely that it is difficult to precisely specify the contours of social-responsibility, it is obviously true. If traditionalists’ are only pointing out that we have not reached a consensus on many of the issues surrounding the ethical obligations of the business corporation, again, we would have no quarrel. If, however, the traditionalists’ view suggests that these reasons necessarily imply that social-responsibility is an untenable option (as Friedman’s position would seem to imply), it is unwarranted. The notion of CSR is difficult to implement in practice. This does not imply that it is impossible.

Second, even accepting the strong assumption that executives and investors explicitly agree that the sole legitimate corporate goal is profit maximization, it certainly does not follow that the ethical world thus evaporates. It is, at best unclear, why if both parties to a transaction agree to disregard an ethical obligation, their joint obligation thus disappears. It may very well be the case that each of the parties may have an a priori, and higher level obligation, to pursue issue X. Any contract therefore which supersedes X may not be binding from an ethical perspective.⁶

Finally, and perhaps most importantly, there is little evidence to suggest that investors and managers, in fact, agree to remove ethical and social-responsibility constraints from executive decision-making. The assumption that corporate management can not use ethical criteria in making good business decisions is at the core of the traditionalists’ view. It appears, on the face of it, that this is a strange suggestion.

Friedman has written that if social-responsibility means anything at all it must mean that managers act in some way that is not in the interest of employers. It maybe, however, that investors’ have a preference for social-responsibility. At minimum, the core assumption of the traditionalists’ argument should be subject to empirical investigation. Mulligan (1986) has summarized the counter-argument as follows:

There is no good reason why this remarkable claim must be true. The exercise of social responsibility in business suffers no diminishment in meaning or merit if the executive and his employers both understand their mutual interest to include a proactive social role and cooperate in undertaking that role. (p. 266)

We articulate this alternative view as a separate proposition.

3.

3. – Some shareholders will willingly forfeit profits for enhanced CSR performance.

Epstein and Pava (1992) have presented survey evidence consistent with this possibility. Though the stereotype is that investors are worried only about profits, when individual investors were explicitly asked to rank their preferences as to how corporate funds should be allocated, pollution control and product safety were rated significantly higher than increased dividends.

4.

4. – Little empirical evidence exists which documents that firms rated high in terms of CSR perform poorly in terms of financial performance.

In the course of this study we identified 22empirical studies (including this one) which attempted to gauge the degree of association between CSR and financial performance. Of these 22 studies which we have examined, only one reported a negative correspondence between social responsibility and financial performance.

5.

5. – Some forms of CSR may enhance, and not detract, from financial performance.

This proposition is the most controversial of our observations. However, we believe that it is the most consistent reading of the available empirical data. The conclusion is also intuitively appealing. This is especially true that if by social responsibility we focus on the following limited set of socially responsible activities: environmental pollution, employee and consumer relations, and product quality. Each of these areas are inextricably linked with financial performance.

What this last proposition does not imply is that one should expect corporations to go beyond their areas of expertise, and to solve social problems of which they are not even indirectly responsible for. Peter Drucker (1989) perhaps put it best when he recently wrote that “We know that corporations first social responsibility is to do their job. We know secondly that they have responsibility for their impacts – on people, on the community, on society, in general. And finally we know that they act irresponsibly if they go beyond the impacts necessary for them to do their own job, whether it is taking care of the sick, producing goods, or advancing learning.” (p. 86)

6.

6. – Stakeholder theory, especially as developed by Ullmann, is a useful but not a complete paradigm to model CSR.

To the extent that proposition 5 holds, it suggests a limitation of the stakeholder theory. As Ullmann (1985) has written, “Economic performance determines the relative weight of social demand and the attention it receives from top decision makers. In periods of low profitability and in situations of high debt, economic demands will have priority over social demands … Economic performance influences the financial capability to undertake costly programs related to social demands.” A model, however, which, a priori, disallows the possibility that CSR causes better financial performance is incompletely specified.

Conclusion

What makes this area of inquiry so interesting is that with each answer, new and exciting questions emerge. The relationship between CSR and financial performance is complex and nuanced. This study has emphasized the recurrent and paradoxical finding that firms which have been perceived as having met social-responsibility criteria have generally been shown to have financial performance at least on a par, if not better, than other firms.

Two areas of further interest are the role of the CEO in establishing CSR goals, and how corporations, through the annual report, defend and justify CSR expenditures. Although our understanding of CSR is by no means complete, it is an area of research that has proven to yield interesting and important results.

Appendices

Appendix A: Summary of 21 Empirical Studies

Authors/Date	Social responsibility criteria	Financial performance criteria	Results	Comments
Bragdon and Marlin (1972)	Council on economic priorities air and water pollution measures	Various measures of financial accounting returns	Lower levels of pollution were correlated with better financial performance	Authors correlated the pollution control indexes with profitability indexes (1965–1971) for 17 companies in the pulp and paper industry
			(+)
Vance (1975)	Milton Moskowitz’s social responsibility ratings	Percentage change in stock price	All but one of the 14 firms in the sample had performance records considerably worse than the NYSE composite index	Author examined updated financial performance (1972–1975) of original Moskowitz sample
			(–)
Bowman and Haire (1975)	Proportion of annual report prose devoted to social responsibility issues	Return on equity	Mean return on equity for firms with “some discussion” was 14.3%, while the mean return on equity for firms with “no discussion” was 9.1%	Authors examined 82 firms in the food-processing industry between 1969 and 1973. Authors claim social disclosure is a surrogate for social responsibility. Some evidence provided which suggests relationship between social responsibility and financial performance may be U-shape
			(+)
Fogler and Nutt (1975)	Three pollution indexes	Financial accounting earnings and stock price data	No significant relationship was found between financial performance and pollution ratings	Authors examined performance of 9 firms between March 1971 and March 1972 after substantial publicity was released about their pollution tendencies
			(0)
Belkaoui (1976)	Disclosure of pollution control information in 1970 annual reports	Market-based returns adjusted for risk	The 50 experimental firms, in which pollution information was disclosed, out-performed the control sample in terms of stock returns	In the 4 month period following disclosure, the market made a temporary conversion of the positive effect of pollution control expenditures in higher share valuation
			(+)
Sturdivant and Ginter (1977)	Milton Moskowitz’s social responsibility ratings	Ten year earnings per share growth	There was a significant difference in EPS growth between the best and worst social performers. Socially responsible firms outperformed their non responsible counterparts	Authors examined 28 firms between 1964 and 1974 who passed data requirements. They conclude that there is evidence that, in general, the responsively managed firms will enjoy better economic performance
			(+)
Alexander and Buchholz (1978)	Milton Moskowitz’s social responsibility ratings	Market-based returns adjusted for risk	No significant relationship between social responsibility ratings and market-based returns	Authors examined stock market performance of 46 firms between 1970 and 1974. They concluded that their results are consistent with efficient markets. Further, the effects of the degree of social responsibility on stock prices were either non-existent or had occurred prior to 1970
			(0)
Chugh et al. (1978)	Firms belonging to high pollution industries	Market-based estimates of beta	Between 1970 and 1972 estimated betas of “polluter” firms shifted up	Authors compared 59 experimental firms, in high pollution industries, to 60 control firms. The authors attributed the shift in estimated betas to the increased water and air pollution control legislation during the 1970 to 1972 time period
			(+)
Anderson and Frankle (1980)	Annual report disclosures (1972) related to social responsibility issues	Market-based returns adjusted for risk	In 6 month period following annual report disclosure there is no difference between disclosing and non-disclosing firms. Examination of March returns, however, gives credence to the possibility of a positive impact	Authors compared stock market returns between 210 disclosing firms and 113 non-disclosing firms. The authors concluded that the results strongly support the contention that the market values social disclosure positively. The ethical inventors may exist and, in fact dominate the market
			(+)
Freedman and Jaggi (1982)	Council on economic priorities air and water pollution measures	Various measures of financial accounting returns	In general there is no association between pollution measures and financial performance. However, evidence is reported which suggests that for very large firms with poor economic performance, pollution disclosure are more detailed	The authors examined the relationship among pollution disclosures, pollution performance, and economic performance for 109 firms in highly polluting industries during 1973 and 1974
			(0)
Shane and Spicer (1983)	Council on economic priorities air and water pollution measures	Market-based returns adjusted for risk surrounding publication of CEP studies	The results indicated that the CEP firms experienced, on average, relatively large negative abnormal returns. Moreover, returns for those companies that revealed to have low pollution-control performance rankings were found, on average, to have significantly more negative returns than companies with high rankings	The authors examined stock market performance of 58 firms (pulp and paper, electric power, iron and steel, and petroleum industries only) between 1970 and 1975. The purpose of this paper was to investigate the question of potential information content of socially-oriented disclosures produced outside the firm
			(+)
Cochran and Wood (1984)	Milton Moskowitz’s social responsibility ratings	Various measures of financial accounting based returns and excess market valuations	Firms with older assets have lower social responsibility ratings. There is also a marginally significant positive association between social responsibility and financial performance	Financial performance was examined for nearly 40 firms between 1970 and 1979
			(+)
Chen and Metcalf (1984)	Two pollution indexes	Various measures of financial accounting based returns, estimated betas, and price earnings ratios	Controlling for firm size, there is no statistical association between pollution indices and financial indicators	In re-examining an earlier study by Spicer, the authors concluded that there is no relationship between a pollution index and financial indicators. The authors concluded that given the visibility of large firms and the severe effects of pollution from large operations on the environment, a large firm tends to do more, either voluntarily or involuntarily on pollution control
			(0)
Aupperle et al. (1985)	CEOs’ concern for society as reflected in mail questionnaire	Return on assets adjusted for risk	No significant relationships were found between a strong orientation toward social responsibility and financial performance	The authors examined the association between attitudes of CEOs (for 241 firms who were listed in Forbes 1981 Annual Directory and answered a mail questionnaire) and financial performance
			(0)
Freedman and Jaggi (1986)	A pollution index	Market-based returns adjusted for risk surrounding annual report date	The test results did not indicate any difference between investor reaction to extensive disclosures and investor reaction to minimal disclosures	This study examines investors’ differential reaction to extensive pollution disclosures in annual statements compared with those firms that make minimal disclosures. All 88 firms belonging to chemical, paper and pulp, oil refining, and steel industries that disclosed some information were examined for 1973 and 1974
			(0)
Baldwin et al. (1986)	Investment in South Africa	Market-based estimates of beta	Excluding firms which do business with South Africa from investment portfolio produces a “minute” increase in risk	Authors attempted to estimate the penalty (as non-market risk) that would have to be incurred as a result of not being able to invest in firms that do business with South Africa. The procedure was to delete excluded companies from S&P 500, and to then come up with the combination of stocks that eliminated the most non-market risk
			(0)
Rockness et al. (1986)	Amount of chemical waste disposal as reported by EPA and US House Subcommittee on Oversight and Investigations	Various measures of financial accounting based returns	Higher ROE was associated with smaller amounts of on-site chemical waste disposal	This study examined 21 firms in the chemical industry between 1980 and 1983. It also examines the disclosure of environmental performance in the annual report with respect to hazardous waste disposal
			(+)
McGuire et al. (1988)	Fortune magazine’s annual survey of corporate reputations	Various measures of financial accounting based returns, market-based returns adjusted for risk, and market-based estimates of beta	ROA and total assets showed positive relationships and operating income growth had a negative correlation. Accounting and stock-market based risk measures tended to be negatively associated with social responsibility	The authors examined the association between financial performance and social responsibility for 98 firms during the 1977–1984 time period. The authors concluded that it may be more fruitful to consider financial performance as a variable influencing social responsibility than the reverse
			(+)
Cotrill (1990)	Fotune magazine’s annual survey of corporate reputations	Market concentration, market share, industry	There was a positive association between market share and CSR. In addition there was an industry effect as well	The author examined 118 firms in over 18 industries during 1982 and 1983. The author wrote that the biggest surprise was that the industry effect was not more fully accounted for by competition levels
			(+)
Patten (1990)	Sullivan principles (A code of behavior mandating equal economic opportunities for non-white workers in South Africa)	Market-based returns adjusted for risk and trading volume around the signing of the principles	No price reaction. Authors did report a volume reaction	The author compared price and volume reaction between 37 firms who signed Sullivan principles in 1977 and 37 control firms. The results indicated that, at least in terms of volume, the information did impact upon stock market behavior
			(0)
Roberts (1992)	Council of economic priorities evaluations of social disclosure, dollars contributed by PACs, public affairs staff members, sponsorship of philanthropic foundation	Various measures of financial accounting based returns, market-based estimates of beta, size, etc.	There was a positive association between CSR and economic performance	The purpose of this study was to test “a stakeholder theory”. The author examined 80 firms between 1984 and 1986 which met data requirements. The author concluded that the empirical results support stakeholder theory

Appendix B: Socially-Screened (Group 1) Versus Control Firms (Group 2)

	Group 1 – Socially-screened	Group 2 – Control
1	Campbell Soup Co	Unilever PLC – Amer Shrs
2	Quaker Oats Co	Borden Inc
3	Sara Lee Corp	Smithfield Companies Inc
4	Ben & Jerry’s Homemde – CL A	Dreyer’s Grand Ice Cream Inc
5	Heinz (H. J.) Co	CPC International Inc
6	General Mills Inc	Ralston Purina Co
7	Kellogg Co	American Maize-Prods – CL A
8	Hershey Foods Corp	Savannah Foods & Inds
9	Tootsie Rool Inds	Mel Diversified Inc
10	Hartmarx Corp	Crystal Brands
11	Liz Claiborne Inc	Benetton Group SPA – ADR New
12	Miller (Herman) Inc	Kimball International – CL A
13	Weyerhaeuser Co	Georgia-Pacific Corp
14	Gannett Co	Times Mirror Co-Del – SER A
15	Knight-Ridder Inc	New York Times Co – CL A
16	Houghton Mifflin Co	Western Publishing Group Inc
17	Wellman Inc	Courtaulds Plc – ADR
18	Baxter International Inc	Smithkline Beecham PLC – ADS
19	Merck & Co	American Home Products Corp
20	Johnson & Johnson	Bristol Myers Squibb
21	Marion Merrell Dow Inc	Imcera Group Inc
22	Procter & Gamble Co	Colgate-Palmolive Co
23	Clorox Co-Del	NCH Corp
24	Avon Products	Intl Favors & Fragrances
25	Fuller (H. B.) Co	Loctite Corp
26	Rubbermaid Inc	Illinois Tool Works
27	Stride Rite Corp	Wolverine World Wide
28	Cummins Engine	Brunswick Corp
29	Digital Equipment	Hewlett-Packard Co
30	Apple Computer Inc	Tandy Corp
31	Pitney Bowes Inc	General Binding Corp
32	Tennant Co	Tokheim Corp
33	Maytag Corp	Whirlpool Corp
34	Lifeline Systems Inc	Pico Products Inc
35	Huffy Corp	Harley-Davidson Inc
36	Easstman Kodak Co	Canon Inc – ADR
37	Xerox Corp	Fuji Photo Filmm – ADR
38	Polaroid Corp	Ricoh Co Ltd – ADR
39	Delta Air Lines Inc	AMR Corp-Del
40	Federal Express Corp	Airborne Freight Corp
41	Nynex Corp	Bellsouth Corp
42	Hawaiian Electric Inds	Puget Sound Power & Light
43	Brooklyn Union Gas Co	Peoples Energy Corp
44	Chambers Development – CL A	Attwoods Plc – ADR
45	Penney (J. C.) Co	Ito Yokado Co Ltd – ADR
46	K Mart Corp	Coles Myer Ltd – ADR
47	Wal-Mart Stores	Woolworth Corp
48	Dayton Hudson Corp	Price Co
49	Nordstrom Inc	TJX Cos Inc – New
50	Rouse Co	Vornado Inc
51	Safety-Kleen Corp	Sotheby’s Holdings – CL A
52	Ryder System Inc	Rollins Truck Leasing
53	Disney (Walt) Company	Bally Mfg Corp