.. . a? 1) 2. . . 432:. w J 2‘4“ 31).? . s x 1 ; .1 . {3.14151 {(931511I1U11. . . .. : .rflwipsz .)11\$~L.\11 . \4‘Iifi‘v .1554] 22?... .. 1.811b.~§}1§<;nw)<.1.115... .35». 12.1.3.3: 4.1... .1 7.... 1 )Jx‘fi. 5)...‘ .22: 1}V1r.1(.\’1\15\3 51.1 n . 1 . .. u... 1.. 11 4 .2115? . x «$39 a .751? L11}! 1..1.2sx!1n1?..117.3&u?1 .1]! \."vv. ’5.‘ Avg?" -‘ 1‘ m‘.¥l)l‘li~ovl\ in .I§\.&:.1. 4 . ‘1’} 55 13.1.3». 51.11)}! 1.431331... . 11L p ‘3 ’51 ah— i 1‘ fit. . V! in). If! 1.5.5.2 .\ .l . ‘i éiv‘lrl‘l i q{\' (a i‘. ‘sl. ltv“ "I, ,sl funds] . L§1\)1\))2.ll. .1. :1}? 31.25.? 11.5. asfl‘ .111 .2... . . , 71v). 1%.}..mm831135éflzzflhfl. his... .. .111}... .(1 ‘1},‘3: 11051 Rm...v;.1.m:24kwaszu..n§?« 2...}... .1193. #1133... 11. .4. (‘1 ‘Vhalnl. L o v... 2... iii-1.2.112; I... ’13 \rasguu .1 L . .3. .. . .2 . tilin111tfc~s I 1.) ittftfr- Ell _ L11... ...{1v§kl§t.r.41.i}oc 3...“. 1831 : £11 k $318.21... 4 . 10.1%. M7; a; 1. . . lv 1“ 1.11117... .... . ..l g 1)“. $11 .qub:ill1hzt:ti£l)l‘ 5.4%”.‘1. M? ii? 11 {.11.i..?l....1.1»1:42§i it ELCdQMNNH .FLEEA3615111k11 14F. §§.l.lfih}hr Ire). K 11.311.1th[41 Off .1141. . r . |1§£Egtut :11; 12:5... . l‘btl VIDIICII‘.‘ L It‘ -I. 1.- r}. 1111.. x233; A!!! . .) .1112}; nlxslAlEti(kur.xxL1.uNl 1 V . . 1 1.5.1..) 1.12.311. 1.1.. . $113.11 . s... .11.... 1b) 21.14111...“ ”ELF 1.11:1 . liaiifillhvl Lstitl .Illlmlrll‘ . 11.1... 1 .1!!!§ol1\l§lxi|nx£1f$l.£it§~ .1 3 1 Z .. $3.1 x)..1.......xh.9.\ ; . . . . 1 .4 ...;-.§.!!,1!.1t$ll..111 1 1.A|. 1....1...1.. ‘Vw.41 . .1 v ..1..... ..1\1J1.34111.S 3.111121... .1 . a . ,5§.§l.1a.s\.(n>1‘.1llr1$.. \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ “ University This is to certify that the thesis entitled The Association Between Corporate Social Performance and Financial Performance in the Market: An Empirical Study presented by Ali Mohamed El-Gawhary has been accepted towards fulfillment of the requirements for Efi-r-B-r— degree in Wmimistration MALM 3 Major professor U D... M,/‘«’>j/7,9 0-7639 ‘t' :5?! ‘3' “ "we". .J i W ‘if‘hnfifih. 5 OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book return to remove charge from circulation records © Copyright by Ali Mohamed El-Gawhary 1979 THE ASSOCIATION BETWEEN CORPORATE SOCIAL PERFORMANCE AND FINANCIAL PERFORMANCE IN THE MARKET: AN EMPIRICAL STUDY By Ali Mohamed El-Gawhary A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Accounting and Financial Administration 1979 ABSTRACT THE ASSOCIATION BETWEEN CORPORATE SOCIAL PERFORMANCE AND FINANCIAL PERFORMANCE IN THE MARKET: AN EMPIRICAL STUDY By Ali Mohamed El-Gawhary Ever since corporate social responsibility became increasingly popularized a decade ago, questions have been raised about what, if any, impact corporate social responsibility has on corporate economic performance. Some argue that since being socially responsible costs money, one should expect responsible companies to show a decrease in financial performance. Others argue that only profitable companies could afford social responsibility and therefore, social responsibility is a by—product of profitability. Others said no; companies are profit- able because they are socially responsible. Thus, it has been suggested that an appropriate concern for corporate social responsibility on the part of a company is a sign of good management and, therefore, is con- sistent with and necessary to sound investment policies. The objective of this study was to investigate the association between corporate social performance and market based financial per- formance. The general question which this research has attempted to answer is: Are corporate social performance and financial performance in the market compatible? Three specific operational questions were Ali Mohamed El-Gawhary formulated in such a way that answers to them would provide inferential evidence related to the answer for the general research question. The corporate social rating approach was used to identify those companies which are socially responsible and those which are not. This approach has been chosen to avoid the problems encountered when mea- suring and weighting the diverse aspects of corporate social perform— ance. Also, it is an attempt to overcome the deficiency of using one social area, such as pollution control, and neglecting all other rele— vant aspects. The study samples were selected from fifteen lists of companies which have been obtained from different sources. A control sample was selected to match the socially responsible companies in terms of industry code and asset size. For each of these three samples, a measure of financial perform— ance (defined as the risk adjusted return) was computed for each month during the study period (1971-1975). Parametric and nonparametric statistical comparisons were made between the three sample portfolios. Additionally, the socially responsible sample's market performance was compared to the performance of the general market. The results of the analyses indicated that the financial perform- ance of the socially responsible portfolio was not statistically dif- ferent from the financial performance of the socially irresponsible portfolio. However, the mean risk adjusted return of the socially responsible portfolio was somewhat higher than that of the socially irresponsible portfolio. Also, it is possible that a size effect might have contributed to an upward bias in the returns of the socially irre- sponsible portfolio. The financial performance of the socially Ali Mohamed El-Gawhary responsible portfolio was not significantly different from the general market. Finally, the financial performance of the control portfolio appeared to be higher than the socially responsible portfolio however, the statistical significance of the differential market performance was only moderate (p = .16). The following tentative conclusions were drawn from the study's results. First, the equity securities of companies which exhibit a high degree of social responsibility may tend to perform somewhat better in the market than the equity securities of socially irresponsible com- panies. Second, the equity securities of companies that are not identified as either socially irresponsible or highly responsible may tend to perform somewhat better in the market than equity securities of companies which exhibit a high degree of social responsibility. However, this second conclusion should be viewed with a degree of caution since the cumulative risk adjusted (excess) return of the socially responsible and control samples exhibited a tendency to move closer together over the last six months of the study period. ACKNOWLEDGMENTS I would like to express my sincere gratitude to professor Stephen L. Buzby for his continuous encouragement and for his invaluable guidance. Dr. Buzby, as chairman of the dissertation committee, pro— vided a sense of purpose and direction to the entire research project. His knowledge and perception had a significant impact on the structure and relevance of the completed research. His thorough analysis of numerous preliminary drafts of this paper has had a considerable influ- ence on its content and organization. Appreciation also is extended to the other members of the disser- tation committee. In particular, I would like to express my apprecia— tion to professor Kelly Price for his interest in the work presented here and his many valuable comments. In addition, my sincere thanks go to professors Fredric Jacobs and Simeon Okpechi for their coopera- tion and assistance. A special thank you goes to Dr. Gardner M. Jones, Associate Dean of the Graduate School of Business, for reading the manuscript and providing valuable comments and editorial assistance. Also, the tech— nical assistance of the computer laboratory personnel, in the computer programming and data manipulation phase of the research, is deeply appreciated. I am grateful to the Egyptian Government for providing me with the financial support that made it possible for me to devote full iii time to my doctoral studies and dissertation. Finally, but most important, a special note of thanks is reserved for my wife, Laila, and our children, Rehram, Tarek and Shamel. With- out my wife's continuous encouragement, patience, understanding and love, completing the doctoral program would have been literally im— possible. Without our children's inspiration, completing the task would not have been as meaningful. iv TABLE OF CONTENTS Page LIST OF TABLES. . . . . . . . . . . . . . . . . . . . . . . . . . vii LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . viii Chapter I. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . 1 Specific Statement of the Problem, Objectives and Hypotheses. . . . . . . . . . . . . . 3 Statement of the Problem. . . . . . . . . . 3 Statement of Objectives . . . . . . . . . . . . . 4 Statement of Hypotheses . 5 Overview of Research Methodology. . . . . . . . . . . 6 Relevance and Implications. . . . . . . . . . . . . . 7 Organization of Remaining Chapters. . . . . . . . . . 11 II. CORPORATE SOCIAL RESPONSIBILITY ISSUE . . . . . . . . . 13 Background. . . . . . . . . . . . . . . . . . . . 13 Social Accounting Literature. . . . . . . . . . . . 22 Demand for Social Accounting Information. . . . . . . 29 Social Reporting: State of the Art . . . . . . . . . 34 III. THEORETICAL CONSIDERATIONS AND REVIEW OF RELATED EMPIRICAL RESEARCH. . . . . . . . . . . . . 38 Investor Perceptions. . . . . . . . . . . 38 The Relationship Between Corporate Social Performance and Economic Performance . . . . . . . . 42 Review of Related Empirical Research. . . . . . . . . 48 IV. RESEARCH DESIGN AND METHODOLOGY . . . . . . . . . . . . 60 Restatement of Objectives and Research Questions . . . . . . . . . . . . . . . . . 60 Research Hypotheses . . . . . . . . . . . . . . . . . 63 Sample Selection. . . . . . . . . . . . . . . . . . 65 Corporate Rating Approach . . . . . . . . . . . . . . 67 Study Time Period . . . . . . . . . . . . . . . . . . 76 Chapter Financial Data Sources. . Specification and Measurement of Corporate Financial Performance. Statistical Testing Procedures. V. RESULTS OF THE ANALYSIS Analysis of Excess Return . Use of Nonparametric and Parametric Tests . Test of the First Hypothesis. Test of the Second Hypothesis . Test of the Third Hypothesis. Analysis of Beta and the Issue of Beta Stationarity . . . . . Additional Discussion and Interpre— tation of Results. VI. SUMMARY AND CONCLUSIONS Statement of the Problem. Purpose of the Research . Research Methodology. Findings and Conclusions. Limitations of the Study. . . Suggestions for Future Research . APPENDIX A Socially Responsible Sample . . . Socially Responsible Control Sample . Socially Irresponsible Sample LIST OF REFERENCES. vi Page 77 77 82 87 87 87 90 94 95 98 104 110 110 112 112 113 115 117 118 119 120 121 LIST OF TABLES Table Page 1 Fortune 500 Companies Making Social Respon~ sibility Disclosure . . . . . . . . . . . . . . . . . 37 2 Description of the Lists from which the SRG and the SIG were Chosen . . . . . . . . . . . . . . . 71 3 Industry Membership of Sample Companies . . . . . . . . 75 4 1971-1975 Average Asset Size of Sample Companies. . . . 76 5 Summary Statistics of the Monthly Excess Return (January 1971 - December 1975). . . . . . . . . . . . 88 6 Summary Statistics of Monthly Estimated Beta (January 1971 — 1975) . . . . . . . . . . . . . . . . 99 vii Figure LIST OF FIGURES Portfolio Cumulative Excess Returns of the Socially Responsible Sample and the Socially Irresponsible Sample . Portfolio Cumulative Excess Returns of the Socially Responsible Sample and the Control Sample. . . . . Moving Portfolio Betas of the Socially Responsible, Control and the Socially Irresponsible Samples. Percentage Changes in the Betas of the Socially Responsible Sample and the Socially Irresponsible Sample. Percentage Changes in the Betas of the Socially Responsible Sample and the Control Sample . viii Page 91 96 100 102 103 CHAPTER ONE INTRODUCTION The issue of corporate social responsibility involves many acti- vities that economists consider as externalities. The existence of an external economy (net social benefit) or external diseconomy (net social cost) results from an activity which affects parties external to business enterprises. These externalities are functionally internal but financially externalized costs and benefits that are not ”captured" by the standard financial accounts of the business enterprises. In the case of a diseconomy, no explicit costs would be recognized by the firm and external parties would not be adequately compensated for losses or damages. The problem of whether and how to correct for externalities in the form of environmental pollution, employee health and safety, product quality, and similar issues is complex and has received much attention in recent years. Pressures from consumer and environmental protection groups, law- suits, legislative action, proxy fights, and ethical investors have influenced corporate decision-making. Corporations have responded to these pressures with increased expenditures and actions in several areas of social concern. Accountants have responded by developing a number of proposals aimed at the construction and application of better methods of measuring, assessing and controlling the social performance of private enterprises. A number of individuals and 2 institutions have responded by considering social issues in their investment decision making processes. They support the notion of the social contract. According to the notion of the social contract, any individual firm agrees to perform certain socially desirable functions in return for certain rewards. As a deliverer of socially desirable goods, the firm is cast in the role of an agent of production. In this role, the ultimate test of a firm's success is whether its aggre- gate contribution to society is more than its aggregate consumption of the society's resources. The issue of whether investors (especially institutional investors) consider corporate social performance when making investment decisions raises the question of whether there is a relationship between corpo— rate social performance and financial performance. Questions have been raised about what, if any, impact corporate social responsibility has on corporate financial performance. The debate on the relationship between corporate social performance and financial performance can be classified into four views. First, socially responsible activities cost money, therefore responsible companies must show a decrease in financial performance. Second, only profitable companies could afford social responsibility, and therefore, social responsibility is a by- product of profitability. Third, companies are profitable because they are socially responsible; responsibility is a symptom of fore- sight, and foresight means good and profitable management. Fourth, company's financial condition does not tend to govern the extent of social involvement and that social responsibility is an independent decision of corporate management. The objective of this study is to investigate the association 3 between corporate social performance and financial performance in the market as reflected in security returns. The primary purpose of the empirical investigation is to provide inferential evidence related to the answer for the following question: Are corporate social perform- ance and financial performance compatible? Several studies have attempted to address this question using different methodologies. Unfortunately, these studies suffer from several potential limitations which, hopefully, will be mitigated in this study. Specific Statement of the Problem, Objectives and Hypotheses Statement of the Problem Spicer (1976 and 1978) found that companies with adequate pollu- tion control records were larger in size, less risky and more profit- able than companies with poorer pollution control records. Bowman and Haire (1975) found that companies which evidenced a middle—level amount of social activities were on the average more profitable than those companies with little reported activities or on the other hand, those companies with an extensive amount of such activities. Heinze (1976) concluded that companies' financial condition tended to govern the extent of social involvement. Longstreth and Rosenbloom (1973) found that 57 percent of the institutional investors surveyed by them consider social factors in their investment decisions, in addition to economic factors. They pointed out that a widely shared view holds that some investors while seeking to enhance their own economic interests, also believe that corporate performance with regard to social issues does impact on the expected return and risk of a security. Many institutional investors 4 in Europe and in the U.S. who were interviewed for a study of corporate social responsibility and the investor (Bowman, 1973) indicated that an appropriate concern for corporate social responsibility on the part of the company is a sign of good management and therefore consistent with and necessary to a good investment. If investors and consumers are conscious of firms' social activities and differentiate between firms on this basis in their investing and consuming activities, the effects of their behavior may be detectable by observing firms' security prices and earnings. Assuming that investors continue to act in accordance with their economic interests, it is possible that they consider corporate social responsibility as a surrogate for good management. If good management is considered to be a surrogate for a good investment, corporate social performance and financial performance in the market would not be incom- patible. Still, from an ethical point of view, some investors may seek out investments of companies that are socially responsible even though this may result in a lower rate of return and/or a higher level of risk than is offered by other investments. Statement of Objectives The objective of this study is to investigate the association between corporate social performance and financial performance in the market. The purpose of this investigation is to provide inferential evidence related to the answer for the following question: Are corpo- rate social performance and financial performance compatible? It was felt that the past research in this area suffer from several potential limitations, therefore, an alternative experimental design 5 and methodology was developed in order to answer the previous question. Three operational questions were formulated, to help in answering the general question, as follows: 1. Does a portfolio constructed from socially respon— sible stocks outperform a portfolio constructed from socially irresponsible stocks? Does a portfolio constructed from socially respon- sible stocks outperform a portfolio of randomly selected sample from the market in general? Does a portfolio of socially responsible companies outperform a comparable portfolio of unknown social performance companies selected to match the socially responsible portfolio in terms of industry code and asset size? Statement of Hypotheses Based on the research questions in the previous section, the following are the null hypotheses to be tested: 10 HO. There is no significant difference in corporate economic performance, in terms of excess return, between a port- folio of socially responsible companies and a portfolio of socially irresponsible companies. This hypothesis is designed to test the financial performance, in terms of excess return, of the socially responsible companies against the financial performance of the socially irresponsible companies 20 H0. The excess return of a portfolio of socially respon- sible companies is not significantly different from zero. This hypothesis is designed to test the financial performance of the socially responsible companies against the financial performance of the market in general. 3. HO. There is no significant difference in corporate economic performance, in terms of excess return, between a port- folio of socially responsible companies and a portfolio of unknown social performance companies selected to match the socially responsible portfolio in terms of industry code and asset size. 6 Since the socially irresponsible group is not matched to the socially responsible group in terms of asset size and industry code, the result of testing the first hypothesis will be less than conclu— sive. Therefore, the third hypothesis is designed to test the finan- cial performance of the socially responsible portfolio against the financial performance of a control portfolio selected to match the socially responsible portfolio in terms of asset size and industry membership. Overview of Research Methodology To test the hypotheses of the study, corporate social performance and financial performance must be defined and measured. The corporate rating approach has been used to identify those companies which are socially responsible and those which are not. This approach has been chosen to avoid the problems encountered when measuring and weighting the diverse aspects of corporate social performance. Also, it is an attempt to overcome the deficiency of using one social area, such as pollution control, and neglecting all other relevant aspects. Fifteen lists of companies that have been evaluated according to their social performance were obtained from different sources. Criteria were established for the selection of the study samples from these lists. Three samples were chosen for the purpose of this study. The first sample was chosen from those companies which were identified as socially responsible during the study period (1971—1975). The second sample was selected from those companies which were identified as socially irresponsible during the study period. The third sample was selected, as a control group, from those companies whose social performance were not known. This third sample was matched to the first sample in terms of asset size and industry category. A procedure for measuring the financial performance of the sam- ples was established using modern portfolio theory and the market model. Monthly moving beta coefficients were estimated by applying ordinary least squares regression procedures and the market model to return data for the fifty-two month period ending one month before the month of interest. A measure of corporate economic performance (defined as the risk adjusted return) was computed for each month and for each sample during the study period. Parametric and nonparametric statistical comparisons were made between the measures of economic performance computed for each of the three samples in addressing the major question posed in this research. A Kolmogorov—Smirnov two—sample test and a two-sample z—test were used to compare the financial performance of the socially responsible sample and the socially irresponsible sample. A one-sample z—test was used to compare the financial performance of the socially responsible sample to the market in general. A Wilcoxon matched-pairs signed- ranks test and a two-related samples z-test were used to compare the financial performance of the socially responsible sample and the con— trol sample. For each portfolio, 3 Kolmogrov-Smirnov two—sample test was used to test beta stationarity during the study period. Relevance and Implications An answer to the general research question of the compatibility between corporate social performance and financial performance may help in clarifying several issues which have implications for both 8 accounting and finance. Traditionally it has been assumed that invest- ors will act in accordance with their economic interests when choosing among alternative investments. This means that investors will seek the highest possible return given their preference for risk. Assuming that investors continue to act in accordance with their economic interests, it is possible that they consider corporate social respon— sibility as a surrogate for good management. There is evidence that some investors believe that a firm's involvement in social activities is a surrogate measure for the quality of management. For example, recall that Longstreth and Rosenbloom (1973) surveyed a number of institutional investors and found that some of the respondents apparently believe that a socially responsible firm is a well managed firm and that good management is associated with superior financial performance. That is, it makes good investment sense to invest in the securities of socially responsible companies. Furthermore, several writers, such as Bowman (1973), have suggested that socially responsible corporations, when compared to socially irresponsible corporations, should exhibit lower degrees of risk and/or higher degrees of profitability. However, there are a few writers such as Chastain (1975) who believe that the financial performance of the socially responsible firms will be penalized vis:a—vis the financial performance of the socially irresponsible firms. Given the conflicting opinions that exist with respect to the relationship between corporate social performance and financial per— formance, an empirical investigation of the relationship appears to be warranted. There are of course, a limited number of studies which have attempted to examine the assocaition between corporate social 9 performance and financial performance.1 In general, these studies have reported the existence of a positive association. Unfortunately, each of the existing studies possess several potentially serious limitations which render their results as questionable. Therefore, this study's results may contribute to a clarification of the associ— ation between corporate social performance and financial performance. With respect to the accounting implications of this study, several accountants have suggested that the accounting model should be expanded to encompass corporate social responsibility data. For example, Brummet (1973) called for an expansion of the accounting model to measure and communicate the impact of business organizations on society. The Study Group on the Objectives of Financial Statements (1973) has proposed that social reporting be considered as one of the objectives of financial statements. The implicit question in this proposal is whether corporations should be required to disclose more information than they now do on their social performance with respect to issues of social concern. The American Accounting Association Committee on Environmental Effects of Organization Behavior (1973, p. 94) stated that: Accounting has evolved to its present status as a profes- sion because it has changed to meet the needs of measuring and communicating economic and financial information to society. Society is now demanding environment information from organizations. In order that accounting may continue to receive professional endorsement from society, it must communicate needed ecological information to interested parties. lSee Bragdon and Marlin (1972), Spicer (1976 and 1978), and Sturdivant and Ginter (1977). 10 If the objectives of financial statements are to provide infor- mation relevant to a user's assessment of current cash flows to the firm and the prediction of future cash flows to the firm, then a ques- tion arises as to whether there is an association between corporate social responsibility and financial performance. As a case in point, it is clear that public concern over the social consequences of corporate activities can, and has, lead to various types of sanctions against socially irresponsible companies. Assuming that investors will con- tinue to act in accordance with their economic interests, investors would consider the socially irresponsible stocks as risky investments since such companies may be subject to government regulation and con- sumer retaliation which would affect their financial performance. Therefore, social responsibility might be considered as a symptom of management foresight, and foresight means good and profitable management. It follows that investors will seek corporate social responsibility information because they consider it as a surrogate for good management and good management is consistent with and neces- sary to a sound investment. The ultimate test of the relevance of corporate social responsi- bility data to the economic evaluation of the investment worth of a firm's securities will depend on the contribution of this data to improved predictions of financial performance. However, it seems that the existence of an association between corporate social and financial performance should serve as an indication that social data might have some predictive ability. Thus, if this study's results tend to indi- cate that such an association exists, it might be argued that there may be some economic merit to the inclusion of some social data in 11 the current accounting model. Of course, even if there is no associ— ation between corporate social performance and financial performance, it is entirely possible that social responsibility information would still be relevant for some investors. This is because there may be some investors who seek such information in order to avoid investments in the securities of socially irresponsible companies, based on purely ethical grounds. Organization of Remaining Chapters The rest of the dissertation will be structured as follows: Chapter II will provide an overview of the social responsibility issue and a review of the social accounting literature, the proposals for measuring and reporting social accounting information, and the state of the art of social reporting. Chapter III will look at the question of investor perceptions of the relationship between corporate social performance and financial performance. The theoretical relationship between social performance and financial performance will be established and a review of the relevant empirical research will be presented. Chapter IV will explain the research methodology to be used in answering the research questions. Also, it will discuss the statis- tical tests that will be used to test the research hypotheses. Chapter V will present the results of the analyses of the risk adjusted return and betas of the study samples (portfolios). A dis- cussion and interpretation of the results will be included in this chapter. Chapter VI will provide a summary of the research and a 12 reiteration of the major conclusions drawn from the study. Also, a discussion of the limitations of the study and suggestions for future research will be presented. CHAPTER TWO CORPORATE SOCIAL RESPONSIBILITY ISSUE The purpose of this chapter is to explore the corporate social responsibility issue. This purpose will be achieved by introducing the major approaches which characterize the relationship of business to society. The merits of each approach will be discussed. A review of social accounting literature will be presented. Proposals for measuring and reporting externalties will be introduced and discussed. Also, the potential users of social accounting information and the state of the art of social reporting will be introduced. Background During the past decade, there have been considerable social, political and economic pressures on corporate management to pay greater attention to its social responsibilities (Backman, 1975). Tradition- ally, the corporation was considered a good corporate citizen if it satisfied certain basic responsibilities: (1) if it provided a product or service valued by the society, (2) if, in its business operations, it stayed within the confines of the law, and (3) if it provided employment for members of the local community. However, since the early 1960's, the concept of a "good citizen" has undergone several fundamental changes. Satisfaction of the three criteria noted above was no longer deemed adequate to meet social responsibilities. Instead, 13 14 there has been an increasing concern for equal employment opportunity, the minimization of pollution, and the effective utilization of soci— ety's resources by corporations (Epstein, Flamholtz, and McDonough, 1976). Dwight Ladd (1963) has described the present relation between accounting and society. He believes that the major objective of organ- izations is no longer profit maximization and that stockholders' interests are no longer the primary responsibility of managers. Accounting principles, however, are still predicated on the premise that ownership interests are paramount. Thus, accounting continues to rely primarily on the notion of profit maximization as the basis for communicating the necessary and relevant information to society by which it can properly evaluate corporate performance. In Ladd's opinion, the objectives of organizations are to mainta- in their competitive position, to grow and to be "good citizens," that is, to respond to society's demand that they act as chief coordinators in using personnel, materials and monetary resources to produce goods and services, and in distributing returns to contributors. Since organ- izations thus hold power over a large part of our national resources they have a major responsibility to society in general. Sheldon and Parke (1975, pp. 87-88), pointed out that there are, historically, three major approaches characterizing the relationship of business to society. The first is that of traditional business and stems from the early industrial era. Business operates in a competitive market and functions to make a profit; hence, the businessman is re— quired only to deal fairly and honestly with his clients. Those who are committed to this attitude and who make contributions to the general I (. 15 welfare do so primarily by serving in community or charitable organi— zations. The second view has been gaining ground since the turn of the century. This approach holds that business has a responsibility to society with respect to its employees and products, and a responsi- bility to mirror the ideals and values of the society within its own microcosm. Businessmen then act affirmatively to promote safety, honesty, and efficiency; to contain and eliminate where possible those disruptions to the environment caused by their products or processes; and to create within their own institutions the conditions of nondis— crimination demanded by society. The third viewpoint is an activist one. It is held more by critics of business than by businessmen. In this third view, it is a primary obligation of business to use its power to promote social ends per- ceived as moral. An obvious example from the recent past, was the recommendation that firms refuse to sell to the U.S. Government unless it withdrew from the Vietnam war or refuse to deal with other corpora— tions whose services may further the presecution of the war. Those holding the traditional view rely on "market mechanisms" or government programs to resolve the social issues. Direct social action is often considered to be inappropriate since it could produce a diver- sion of management skills or a violation of responsibility to the owners. Friedman (1962) for example, has emphasized the primacy of the business firm's drive for profits. In his opinion, "there is one and only one social responsibility of business - to use its resources (for production) and engage in activities designed to increase its profits so long as it stays within the rules of the game" (p. 133). 16 Friedman views the market system both as the most efficient method of distributing resources and controlling costs and as bulwark of econ- omic and political freedom. He does not accept the concept of social responsibility because it, in his opinion, encourages interference in the functioning of the market system, causes irresponsible diversion of corporate resources from profits to social purposes, and results in dangerous meddling. Miller (1972) opposes the above idea and describes it as a narrow pursuit of profit. He said that our life would be so much simpler if our task and our only responsibility was the narrow pursuit of profits. As it happens, tremendous new demands are being made on this society, and they in turn are causing tremendous new demands on business. We should recognize that many of our most serious and urgent current de— mands are in the area of public, rather than private, goods. People want cleaner air and water, safer streets, less traffic congestion, better education systems, and so forth. These demands cannot be satis- fied in the traditional market place. The merit of the conservative position against corporate social responsibility is that the market system places limits on what a cor- poration can do to solve social problems. Business corporations are not able to solve all of society's ills. Society cannot expect business firms to produce goods and services efficiently and concurrently expect them to shoulder numerous unrelated social costs. A firm is limited to social projects that are in some sense related to its operating environ— ment and its own interests. However, one weakness in the conservative position against corporate social responsibility is the fact that it is not a useful guide for management decision making. Management is l7 admonished to focus only on the financial well—being of shareholders, yet the firm that takes a "public-be-damned" attitude toward nonshare- holders, is a likely target for harsh and perhaps even punitive legisla— tion. The firm that ignores the interests of society in a pell-mell pursuit of profit neglects the long-run consequences of such a strategy, consequences that could prove to be costly or even suicidal. Perhaps the most important critique of the conservative position against corporate social responsibility to date has been that of Kenneth Arrow (1973), a Nobel prize-winning economist. Arrow directly challenges the view that profit maximization is a necessary condition for achieving economic efficiency in a free market economy. Situations arise, Arrow points out, where profit maximization is not socially desirable. He stated: (T)here are two types of situations in which the simple rule of maximizing profits is socially inefficient: the case in which costs are not paid for, as in pollution, and the case in which the seller has considerably more knowledge about his product than the buyer, particularly with regard to safety. In these situations it is clearly desirable to have some idea of social responsibility that is to experience an obligation, whether ethical, moral, or legal (p. 309). Wallich (1972 a, p. 123), an equally vigorous advocate of the free market as Friedman, considers corporate social responsibility to be ”how business can rescue capitalism." If business does not behave respon- sibly, society will then act collectively through government to impose legal constraints on corporate behavior. The choice facing business is either to behave in such a way that it does not tend to cause or exacer— bate social ills or to ignore the effects of its operations on the pub- lic and be faced with new regulations and possibly higher taxes. The firm that pollutes may either stop polluting on its own or eventually 18 face enactment of antipollution laws. The firm that persists in mis— leading customers or mistreating employees invites legal and political challenges. The major reason for preferring corporate social responsibility to government regulation is that business can solve many societal problems more economically than if government action is relied upon: If corporations do not do what needs to be done, using stockholders' money, it will be done by government, using taxpayers' money. There are many reasons to think that a political body, working through a centralized system, will often get less value per dollar than the decentralized and results—oriented business system. In terms of how and at what prices to get a job done, there is much to be said for handling social responsibilities through corporate channels (Wallich 1972 b). Corporate responsibility is seen here as a less expensive alterna- tive to legal and political regulations which constrain managerial autonomy and efficiency. Virtually all those who argue for greater social responsibility by corporations make the point that governmental regulations carry onerous enforcement and implementation costs. Social responsibility is motivated by an attempt to avoid or forestall costly regulation and bureaucratic intervention. Shocker and Sethi (1974, p. 67) emphasized the notion of social contract, they argue: Any social institution--and business is no exception-- operates in society via a social contract, expressed or implied whereby its survival and growth are based on: (1) the delivery of some socially desirable ends to society in general and (2) the distribution of economic, social or political benefits to groups from which it derives its power. In a dynamic society, neither the sources of institu— tional power nor the needs for its services are per- manent. Therefore, an institution must constantly meet the twin tests of legitimacy and relevance by demon- strating that society requires its services and that 19 the groups benefitting from its rewards have society's approval. An individual firm, according to the notion of social contract, agrees to perform certain socially desirable functions in return for certain rewards. However, certain types of decisions of an individual firm may affect the actions of other firms or social groups or individ— uals in ways which are not reflected in the market process. Social accounting calls for the recognition of such nonmarket transactions in addition to the traditional market—based financial transactions. As a deliverer of socially desirable goods, the firm is cast in the role of an agent of production. In this role, the ultimate test of a firm's success is whether its aggregate contribution to the society is more than its aggregate consumption of the society's resources (Ramanathan, 1976, p. 519). Many economists agree that the source of the environmental problem is the fact that the price system is not applied to many of society's resources. William Baumol (1975) argues that a tax on pollution will produce environmental protection at a minimal cost to society. He pointed out that his proposal seeks to minimize the use of direct con- trols and to provide effective encouragement to industry to take appro— priate remedial action. The economist is impressed by the efficiency with which the economy utilizes resources that are supplied under the rules of the price system. Industry uses its raw materials with a degree of care and efficiency that is perhaps unparalleled in economic history. Yet, at the same time, the air supply deteriorates progressively, the rivers are transformed into sewers, and the neighbor- hoods into slums. What has gone wrong? Experience tells us what happens when costly resources are supplied free of charge (Baumol, 1975, p. 244). The prevailing view appears to be that the firm must periodically follow lines of actions that in some measure help society attain one or 20 more of its goals, as determined by the political process. Such poli— cies, even if not immediately profitable are presumed ultimately to re— dound to the benefit of the enterprise and its shareholders. Pressures from consumer and environmental protection groups, law- suits, legislative action, proxy fights, and ethical investors have influenced corporate decision-making. Corporations have responded to these pressures with increased expenditures and actions in several areas of social concern (Jensen 1976). These pressures which came from many sides, in turn, have resulted in a number of proposals aimed at the development and the application of better methods of measuring, assessing and controlling the social performance of private enterprise with respect to issues of social concern such as environmental protection, equal employment opportunity, quality of life and consumer policies. The Study Group on the Objectives of Financial Statements (1973) has proposed that social reporting be considered as one of the objec- tives of financial statements. This is implied by the following state- ment: An objective of financial statements is to report on those activities of the enterprise affecting society which can be determined and described or measured and which are important to the role of the enterprise in its social environment (p. 55). The implicit question in this proposal is whether corporations should be required to disclose more information than they now do on their social performance with respect to issues of social concern. 1Council on Economic Priorities (1973, 1974), Abt (1977), Linowes (1972, 1974), Baumol (1975), Kapp (1971), Beams and Fertig (1971), Estes (1976), Seidler (1975), Churchill (1973, 1978) and The American Institute of Certified Public Accountants (1977). 21 The resolution of this question poses a significant challenge for the future development of accounting theory and practice. It suggests a fundamental change in disclosure philosophy and significant expansion in the subject matter of accounting. The AAA Committee on Measurement of Social Costs (1974, p. 102) states: As accountants, we have little choice but to involve our— selves in the field of social accounting. A profession must be responsive to the needs of society if it is to remain viable. We must accept rather than avert the risks inherent in the movement into new fields, even at the expense of being considered social accounting activists. The need for us to develop new analytical tools to deal with broader measurement constructs is urgent. We may not even be able to wait the orderly development of tax— onomy. Along the same line, David Johnston (1978, p. 12) asserts the ne- cessity of refining corporate social measurement and reporting techniques and raising their quality to that of financial measurement, he said: Business executives might argue, 'why should we voluntarily refine our social measurement and reporting techniques if the government is increasingly going to require such report— ing?‘ It would be nice if one could project that if enough companies develop credible voluntary, 'anticipatory' social measurement and reporting techniques, more legislation for disclosure could be averted. That seems highly unlikely. Doing nothing in the short run will cost more in the long run. Thus the sooner corporations raise the quality of their social measurement to that of their financial measure— ment, the better equipped they will be to respond effici- ently to the demands for broadened accountability, whether from government, activist organizations, or the community. Flamholtz (1974) expects that in the near future, business corpo- rations will be accountable for social and financial goals, he argues: As a result of changing social ethos, business organiza- tions are increasingly expected to meet culturally and governmentally defined standards of corporate responsi— bility with respect to the environment, consumerism, minority employment, women's rights, and employee satis— faction. If current trends persist, it is likely that business corporations increasingly will be held account— able for social as well as financial goals. They will 22 be asked to demonstrate the extent of their social con- tribution. They may also come under the scrutiny of a social audit to assess the costs as well as the benefits of corporate activities (p. 305). Social Accounting Literature The proposed expansion of the subject matter of accounting carries with it implications for the specification of the decision—makers of interest, the contents of financial statements, reports and other cor— porate disclosure, the properties of events and phenomena to be mea— sured, legitimate measurement techniques to be employed, and the cost of information to decision-makers. A number of these issues have been the subject matter of committee deliberations and reports by all major accounting bodies (American Accounting Association, 1971-1976; National Association of Accountants, 1974). In addition, a number of accounting academics and others have attempted to expand the accounting model to include social performance evaluation. These attempts include normative proposals for a comprehensive theory of social accounting (Ramanathan, 1976; Estes, 1976), a step-by—step social audit process for inventorying the firm's social involvement (Bauer and Fenn, 1972; Abt, 1977) and workable methodologies for social performance evaluation (lionwes, 1972— 1973, 1974; Seidler and Seidler, 1975). These attempts among others will be reviewed below. Linowes (1974) has proposed an audit that would produce a socio— economic operating statement (SEOS). The SEOS would include both the marginal costs (those costs incurred above regular business expenses) of actions taken for other than business purposes and the costs of actions that should have been taken but were not. The net of these figures would produce the number of "social dollars” with which a firm 23 should be credited. That means that if a socially beneficial action is required by law and regulations, it is not included on a SEOS. On the other hand, if a socially beneficial action is required by law, but is ignored, the cost of such item is a "detriment” for the year. The ex- penditures made voluntarily by a business which are aimed at improving the welfare of employees, and/or conditions of the environment, and/or product safety would be entered on a SEOS as positive actions. The SEOS's overriding purpose, as Linowes (1974, p. 106) pointed out, is to give visibility to corporate social improvements and detri— ments. Its chief focus is on corporate effort, on degree of conscience exercised. It does not seek to assess program effectiveness or project the results of the social investment into the future. The question it seeks to answer is, how socially responsible is company A as compared with company B? Brummet (1973) has criticized the current model of accounting. He called for an expansion of the model to measure and communicate the im- pact of business organizations on society. He proposed that: If accounting is to remain the "language" of business organ- ization, it must increase its vocabulary. It must deal with measurement and communication in all of the spheres of organ- izational activity. It must become performance, accomplish— ment, and success oriented. It must supplement its net income and financial condition orientation with one of assessor and communicator of total company performance... The current practice of accounting is far from neutral in its effects on organizations and society. It may well cause great attention to be given to relatively trivial accomplishments while, in effect, hiding by default the most significant performance aSpects of a company (p. 11). He proposed a taxonomy for ”total performance” assessment. Total performance of a company, he said, is a function of net income, human resource contribution, public contribution, environmental contribution, 24 and product or service contribution. Abt (1977) has written more widely on the topic of the social audit than any other person. He has proposed and developed an ambitious social audit. It involves a social income statement and a social balance sheet. Abt Associates, Inc. have also developed a series of social accounting cost/benefit statements to tie in with five consulting services they offer in social accounting (an inventory of social ser- vices, a social performance evaluation, a social audit, a social evalu— ation of routine operations, and an analysis of social policy impact on financial performance). Abt Associates started doing a social audit in its 1971 annual report. Abt's ”social audit" appeared to be the great leap forward in social measurement. He used elaborate research to develop a totally quantified social income statement and social balance sheet, which included calculations of every conceivable cost and benefit of a com- pany's social impact, often through the use of "shadow prices." A serious problem with the overall structure of Abt's "social audit" is that all environmental costs are seen as liabilities, not as deductions from earnings, whereas benefits (even unexplained benefits like "envi— ronmental improvements") are added to earnings. This has the effect of reducing equity and increasing net income, thereby raising return on equity regardless of social performance. Although the Abt approach was exciting and challenging, it was generally rejected by the busi— ness community for being impractical. Linowes' approach has been rejected as well on the same basis. Seidler and Seidler (1974, p. 205) had commented on Abt Associ— ates' annual report as follows: 25 It is of interest to note that there may be some degree of enlightened self interest involved in this project ... The unique annual report received considerable pub— licity in major business publications; the result was certainly not adverse to the company's business. Equally interesting, Abt is a privately held company, under no obligation to publish and disseminate a formal annual report. Ralph Estes (1976) proposed a comprehensive social report of the direct effects of a single entity on society. He suggests that, when- ever possible, reports should be prepared for each community in which a component of the entity operates reflecting the social impact on that community alone. The report would include all significant social cost and benefits in gross without netting. As Estes noted; this model is somewhat idealistic and may not presently be practical for all firms. Seidler (1975) suggested two social income statement formats, one for a profit seeking organization and one for a not—for-profit organi— zation. These formats reflect the contribution of an entity to society. This social income statement adds socially desirable outputs for which no money is received (external economies) and deducts costs that the entity imposes on society but does not pay for (external diseconomies). The result is a net social profit or loss which will reflect the net contribution of the entity to society. A criticism of this statement is its dependence on the use of many subjective terms in measuring ex- ternal economies (or diseconomies). Marlin (1973) presents a detailed proposal for reporting on pollu- tion and pollution control. He derived a system which would provide for the reporting to the public of the source and type of pollution, the type of control employed, the magnitude of each type of effluent, and a comparison against a norm——the best available technology. 26 Preston (1978) suggested a comprehensive and standardized format for social performance analysis. The format contains a set of several social dimensions that would be relevant to nearly any business of significant size. Some of the information provided are quantitative while others are qualitative and descriptive. Such a format should distinguish between commitments or inputs, on the one hand, and out- puts or results on the other. In addition, appropriate comparative information should be provided whenever possible so that levels of social performance and cost benefit trade-offs can be better evaluated. Blake, Frederick and Myers (1976) introduced the concept of the social process audit and applied it in special case studies. This audit, as the authors noted, is designed to evaluate specific social programs established by management. It evaluates the program in the light of managerial or corporate goals and objectives. This technique does not assess the overall social impact of the corporation nor does it judge management according to some preconceived notion of good or correct behavior. Instead, the social process audit seeks to judge specific programs in terms of goals established for the program, the resources committed to the program, the mechanism by which the program is implemented and managed, and finally the results of the program. It seeks to provide managers with a technique that will enable them to determine whether the program is worthy of continuation according to their particular standards. Although the program management type of social audit has several advantages from the viewpoint of the corporation, it also has serious limitations which must be kept in mind. The program is of little value in the selection of priorities and it is not very useful for external 27 reporting purposes. As Butcher (1973, p. 15) pointed out: Relatively little information is generated about the im— pact of these programs on any particular social problem. By focusing on programs rather than problems, the possi~ bility of concentrating a broad range of capabilities on a single problem area might be overlooked. In a recent study, the American Institute of Certified Public Accountants (1977) offered some guidelines which should be considered in an initial practical social measurement system. The AICPA study did not prescribe a practical social measurement system. Rather, its major contribution to the literature was the pragmatic distinction it made between an ideal system and an initial practical system. The AICPA's ideal system will remain just that, an ideal, theoretical construct which will be "useful" only to the extent that it provides direction. The AICPA's study itself concedes that this system will not be achieved in the near future, if ever. On a more realistic level the AICPA's (1977, pp. 16-17) charac— teristics of an initial practical system would include: 1. Not all social phenomena will be measured; instead, empha— sis will be given to significant actions and impacts affecting areas of primary social concern. 2. Within each area of emphasis, measurement will be made of selected attributes, chosen because they indicate the essence of actions taken and impacts made by the company. 3. A variety of units of measure will be employed, and narrative descriptions will be used where quantitative measurements are not practical. 4. Although occasionally there will be attempts to assess impacts on the quality of human life directly, the measurements will usually relate to impacts on social conditions thought to affect the quality of human life to a significant extent. 5. Where the measurement of impacts on social conditions is not practical, an attempt will be made to measure actions 28 and their immediate results. (These may often be mea- sured, in any event, because of the intrinsic value of that information.) 6. The distinction between social and economic information will often be obscure. 7. The system will not possess complete neutrality. Sater (1971) made one of the earlier proposals for a social audit. His proposed audit was designed to guide investors in their investment decisions. He listed a number of dimensions along which firms might be evaluated: consumerism, type of international business, environmental protection, opportunities for minority group members and women, military and defense activity, contributions, and executive participation in com— munity affairs. He acknowledged that the relevance of these dimensions vary from industry to industry, and suggested that the selection of dimensions be made on the basis of knowledge of the given industry. Sater found difficulty in deciding how to weight a company that does well on one dimension and poorly on another, and concluded that the job of weighting should be left to the user of the audit. Shulman and Gale (1972) propose that a firm be evaluated in terms of how it serves its several constituencies. These constituencies con- sist of: employees, consumers of the organization's products and ser- vices, suppliers, communities in which the organization operates, society in general, and competitors. The authors made some sensible comments on the different requirements for using an audit for internal decision- making and for reporting to the public; they do not, however, go far toward making this plan operational. Dilley and Weygandt (1973) proposed a Social Responsibility Annual Report (SRAR). This SRAR does not attempt to judge the business's 29 performance on social issues. The report merely contains quantitative data that enable the reader to compare one company with the reader's own standards of required performance on a given social issue. They conducted a social audit of a publicly owned Midwest gas and electric utility company. They pointed out that no attempt has been made to reach a judgment on corporate social responsibility. However, in their opinion, this weakness is actually its strength since measurement of corporate social responsibility has not reached the point where judg- ments can be made by the social auditor. Thus, they suggest that the social auditor should limit himself to doing what is currently possible. Nonetheless, the social auditor can provide useful information to out- siders--information they can get nowhere else. Demand for Social Accounting Information The greatest demand for social accounting information appears to come from corporations themselves. Corporate directors need to know in some detail what sort of social programs the company is engaged in and what results it is getting. They also need complete information about the effects of the corporation on society. It is probably more important that they be fully informed as to negative effects, since this is where the criticism will be directed and it is where the directors may have to defend themselves in courts. Estes (1976, p. 3) stated that top manage- ment needs social performance information to respond to critical press, to testify knowledgeably when called on by congressional committees, to answer proxy challenges and stockholders' questions, and to insure that company policies are followed. The AAA Committee on Measurement of Social Costs (1974) emphasizes 30 the importance of social reporting for corporate management. The Com- mittee argues: A major advantage of social reporting for management will be the improved information pertinent to making compliance decisions (minimum vs. maximum) in response to both social pressures and legal requirements. The latter may be expected to change in the future, enhancing the need for flexibility and variety in reporting (pp. 103—104), External demand for social accounting information is even more di— verse. The external user groups are primarily concerned with the results of the social accounting process and how they might use it. Stockholders would receive additional information about items which will aid them in judging the social responsibility of their company. Employees and labor unions might receive information regarding job security, safety, dis— crimination, development, etc. Customers desire information regarding product performance, safety, disposability, advertising, guarantees, and the like. The public in general is concerned with a variety of charities, community programs, pollution control, energy conservation, and other com- pany impacts on the quality of life. There is some evidence of an increasing awareness and concern in the capital market with organizational behavior (AAA Committee Report, 1973). Some banks and other lending institutions make loans at a preferential rate to environment conscious firms, and refuse to assist in financing projects that will have an undesirable effect on the environment (Business Week, 1970b). The AAA Committee (1973, p. 88) suggested that: "These institutions that make direct loans are in a strong position to demand information relating to environmental impact as a requisite to the loan consideration." Johnston (1978, p. 5) summarized the rationale for social 31 measurement and reporting as follows: Internal: 1. To provide management with information on social perform- ance, both in terms of the social impact of the business, and of the effectiveness of specific social programs; 2. To help integrate social performance with financial per— formance through the assignment of managerial accountability; 3. To lay the foundation for the integration of social planning with more traditional corporate planning; and 4. To provide feedback on the effectiveness of what we call social policy planning (which is the identification of priority social and political issues and the development of responses to those issues). External: 1. To be able to respond to the expectations of various con- stituencies for information about the company's social performance, i.e., to manage disclosure; 2. To be in a better position to influence the public policy process; 3. To prepare for "social audits" by public interest organi- zations, government agencies, and legislators; 4. To improve the company's overall public image. Clearly, the internal and external reasons for social measure- ment are related. Thus, managerial accountability to the com- pany for social performance can improve the management of social performance by providing top management with information on that performance. At the same time, internal accountability is a direct result of the increasing level of external or public accountability. Why should investors and shareholders impose social performance criteria on the corporations in which they hold shares? Some of course are motivated by purely moral concerns. But there seems to be a growing tendency to consider social performance even when investment objectives are strictly economic. University trustees and investment committees now are concerned with the social impact of the corporations in which 32 the university owns securities. They aim to maximize economic returns without investing in companies that do "social injury". Thus, a number of universities have sought to develop a framework to compare invest- ments on the basis of social as well as financial performance. Various churches have also been interested in investing their money in socially responsible investments. The Trustees of Harvard University adopted the following policy statement in May 1972: The trustees believe the university, as an investor, must focus not only on a maximum economic return, but also on issues of social responsibility, as produced by the com- pany in which the university holds securities (Backman, 1975, p. 15). Harvard's definition of ”social responsibility" encompasses such areas as a company's position or involvement in labor relations, racial discrimination, minority employment practices, consumer protection, pollution control, military contracts, and international diplomacy. On December 8, 1978, Michigan State University became the largest university in the United States to announce that it would voluntarily sever financial stock holding ties with corporations doing business in South Africa (M.S.U. State News, December, 1978). The trustees put into effect a directive to MSU's money management firm to sell eight million dollars of stock in fifteen corporations with South Africa holding. This resolution was made since the trustees did not receive the necessary assurances from South African—related firms in the uni- versity's investment portfolio, of their intention to withdraw from South Africa. The board also resolved that no future investments would be made in companies listed with the Investor Responsibility Research Center as doing business in the apartheid country of South 33 Arrica. The least strenuous mode of applying social consideration to the investment process is what the Yale University Committee calls "the avoidance of social injury". This position requires that investor should not drink from a polluted income stream. It requires him not to make his capital available to an enterprise which injures his social values. Several mutual funds have been formed to invest in companies that, in one way or another, act in a socially desirable manner. For example, Dreyfus Third Century Fund, and Pax World Fund were established in the early 1970's to provide a portfolio of corporate securities that were selected on the basis of both social performance and financial perform- ance. The fund's management felt that social responsibility should be a criterion that should be used in making investment decisions, and thus, though the mechanism of a mutual fund, investors could use their financial influence to influence corporate behavior. In the opinion of the fund's management (Dreyfus Third Century, 1978, p. 3), the invest- ment should meet the traditional investment standards and the firm itself should show evidence in the conduct of its business, relative to other companies in the same industry or industries, of contributing to the enhancement of the quality of life. In making this latter assessment, the fund intends to consider performance by companies in the areas of: (l) the protection and improvement of the environment and the proper use of natural resources; (2) occupational health and safety; (3) consumer protection and product purity and (4) equal employ- ment opportunity. In addition, special consideration will be given to those companies which have, or are developing, technology, products or 34 services which will contribute to the enhancement of the quality of life in the United States. Within the group of securities which meets the fund's special stan- dards, the fund purchases and sells securities based on traditional investment considerations, including technical market factors as well as management's opinion of the fundamental value of securities. Securities are sold if at any time the management determines that they cease to meet such standards. The National Council of Churches is accumulating data on corporate responsibility in the following areas: the environment, consumer health, welfare and safety, foreign investment, military procurement and produc- tion, and minority employment. The Council gathers and compiles data from other sources to form a corporate information center. Its purpose is to exert moral suasion on churches and interested groups to affect corporate behavior directly and to influence the investment policies of foundations, mutual funds, etc. (Churchill, 1973, p. 18). The nation's two large philanthropic institutions——the Ford Founda- tion and the Rockefeller Foundation—-undertook studies of their invest- ment portfolios, using a social lens to determine whether the social improvement guidelines used in the grant-making area were being violated on the investment side. Social Reporting: State of the Art In the Unites States, most general social reporting appears in the annual report to stockholders, although some companies produce a separate report varying from a few to over 100 pages in length. Social reporting differs widely in scope, content, and comprehensiveness and in the mix 35 of narrative versus quantification. Preston (1978, p. 135) commented on social reporting as follows: The principal reason that voluntary social reporting is not growing even more rapidly appear to be uncertainty about the method of reporting and format to be adopted, and ignorance as to the results achieved, particularly the awareness, appraisal and use of such information by stockholders, employees, customers, public agencies, or members of the public. Most companies publish a mixture of required (in response to spe— cific government regulations) and non required social performance infor— mation in their annual reports. The format of this reporting varies widely from company to company, and sometimes from year to year at the same company. This inconsistency in social responsibility disclosures makes interfirm and interperiod comparability of corporate social per— formance extremely difficult. Although those particular disclosures are neither uniform nor comprehensive, the trend toward increased——and increasingly informative—-corporate social reporting is evidenced. A much smaller number of companies have published more convincing social reports, which generally contain a great deal of hard data and high-quality narrative. Although social performance is hard to compare even with these reports, these companies are at least measuring and reporting a wide variety of social performance. Only a few corporations have included enough information of a sufficiently negative kind to war— rant the phrase, balanced reporting. Ernst & Ernst has been tracking the socially relevant reporting of the "Fortune 500" companies since 1972 and has annually published its findings. The topics of interest are the environment, energy, fair business practice, human resources, community involvement and products. (Ernst & Ernst, 1978). Table 1 summarizes the number of firms of 36 Fortune 500 firms making social responsibility disclosures, and those that do not (based on the 1973-1977 Fortune 500 listing). The table reveals that there is a steady increase in the number of companies making social responsibility disclosure, and a decline in the number of companies making no disclosure except in 1977. Fry and Hock (1976) argue that corporations are using the social report to paint a picture of the firm as having the public interest at heart. In their opinion, the social reports contain more show than substance. The Ernst & Ernst Survey, however, warned against drawing conclusions about the social conduct of companies solely on the basis of the social responsibility disclosures in their annual reports. There are undoubtedly, the survey pointed out, a number of companies who en- gage in social responsibility activities and do not disseminate the information widely. While its wide distribution makes the annual report one of the most effective forums for disclosure, it is not the only communication medium of social responsibility information. Some companies disseminate this information through shareholder-meeting sum- maries, quarterly reports, separate publications, press releases, and filings with regulatory agencies. 37 TABLE 1 FORTUNE 500 COMPANIES MAKING SOCIAL RESPONSIBILITY DISCLOSURE* (Based on the 1973—1977 Fortune 500 Listing) 1973 1974 1975 1976 1977 Companies making social 59.6% 69.2% 85.0% 91.2% 89.2% responsibility disclosure in annual reports (298) (346) (425) (456) (446) Companies not making social 39.6% 30.2% 14.2% 7.6% 8.4% responsibility disclosure in annual reports (198) (151) (71) (38) (42) Reports not available .8% 6% .8% 1.2% 2.4% (4) (3) (4) (6) (12) (*) From Social Responsibility Disclosure, Ernst & Ernst, Cleveland, Ohio, 1978, p. 3. CHAPTER THREE THEORETICAL CONSIDERATIONS AND REVIEW OF RELATED EMPIRICAL RESEARCH This chapter will look at the question of investor perceptions of the relationship between corporate social performance and financial performance. The theoretical relationship between social performance and financial performance will be established and a review of the rele- vant empirical research related to the topic of the association between corporate social performance and financial performance will be presented. The limitations of these studies will be discussed and the justifica- tion for the present study will be provided. Investor Perceptions This section will address the issue of how investors perceive the relationship between corporate social performance and economic per- formance (in terms of dividends, capital gains and risk). In other words, the question is: to what extent is corporate social performance considered when making investment decisions? The answer will be achieved by summarizing the literature which addresses this question. The American Accounting Association Committee on Environmental Effects of Organization Behavior (1973, p. 89), suggested that: Investors are not only demanding that environmental acti— vities be disclosed; their portfolio policy may favor firms that are active in improving the environment, and many are selectively channeling their resources to these organizations and avoiding investment in others. 38 39 On the question of the usefulness and importance of social respon- sibility disclosures to institutional investors, Longstreth and Rosen— bloom (1973) surveyed institutional investors to determine the extent to which social factors were considered in investment decisions. One hundred ninety—six representative institutional investors were included in this study. They found that 57 percent of the institutions surveyed do consider social factors in their investment decisions, in addition to economic factors. However Buzby and Falk (1979) examined the effect of social responsibility disclosures on university investment decisions and reported results which tend to contradict those reported by Long— streth and Rosenbloom. They surveyed 500 university chief financial officers. Despite the common belief which suggests a growing demand by universities for external corporate social responsibility disclo— sures, the authors' findings suggest that a strong demand for social responsibility disclosures did not exist. They also concluded that adequate social responsibility information about corporations is not currently available to the investor. Belkaoui (1972) conducted a behavioral field experiment to measure the impact of the disclosure of pollution control expenditures in the annual reports, on the investment decision of investors from different professional groups. He found that the disclosure of pollution con— trol costs influenced investment decisions, especially for the bank officers group. Investor perceptions with regard to considering a corporation's social performance in investment decision making can be classified into four distinct viewpoints (Spicer, 1976; Longstreth and Rosenbloom, 1973), these are as follows: 1. 40 A traditional investment viewpoint: investors holding this viewpoint see little, if any, connection between a corporation's attention to social issues and the invest- ment worth of its securities. Therefore, investors will act in accordance with their economic interests when choosing among alternative investments in common stocks, i.e., they will seek the maximum possible return in terms of capital gains and dividends given their preference for risk. A modified investment viewpoint: investors holding this viewpoint also seek to enhance their own economic interests. However, they also believe that corporate performance with regard to social issues does impact on the expected return and risk of a security. Therefore, instead of a naive ad— herence to the traditional viewpoint, they consider corporate social performance when assessing the future levels of return and risk for alternative investments. A "passive” ethical investment viewpoint: investors holding this viewpoint, on the basis of moral or ethical tenants, avoid certain investments of companies that are causing unnecessary social injury. An "activist" ethical investment viewpoint: investors holding this viewpoint actively seek out investments of com- panies that are socially responsible even though this may result in a lower rate of return and/or a higher level of risk than is offered by other investments. The AAA Committee on External Reporting (1969, p. 81) stated that investors 41 "may wish to avoid some investments entirely for ethical reasons or because of personal biases." It is important to note that these viewpoints are not mutually exclusive, rather they represent four points on a continuum of opinion. Views of different investors may encompass two or more of these view— points. Longstreth and Rosenbloom's survey (1973) provides some evidence on the degree to which these viewpoints are spread throughout the in— vestment community. Based on 115 responses, the authors' findings are as follows: Sixty-six (57.4 percent) institutions indicated that in addition to strictly economic factors, they also consider social factors in the selection and retention of investments. Of these sixty-six institutions, thirty—two indicated that they avoid certain types of investments as being socially undesirable, immoral or contrary to insti- tutional purpose; seventeen institutions (mostly religious organizations and insurance companies) indicated that a portion of their investment portfolio is directed toward socially oriented investments where economic factors are not controlling; and thirty—nine institutions (mostly banks and mutual funds) suggested that there exists an associ- ation between socially responsible business enterprises and those that will produce a satisfactory monetary return. Forty-nine institutions (42.6 percent) indicated that they have recently reviewed, or currently are reviewing, the social aspects of their investment policy; and twenty—seven institutions (23.5 percent) indicated that they have estab— lished special procedural mechanisms to take social considerations into account in the investment process or have produced policy statements on the matter. 42 The results of Longstreth and Rosenbloom's survey (1973) indi— cated that increased attention is being given to corporate social per- formance on issues of social concern when making investment decisions. Two factors are believed to cause this perception: (1) a belief that there is an association between a corporation's social performance and its economic performance in the market (in terms of risk, dividends and capital gains); (2) an increase in the number of investors who, for moral or ethical reasons, believe that they should avoid investing in certain classes of firms thought to be causing unnecessary social injury of one type or another. To conclude, it appears that the modified viewpoint may be a widely shared view. It states that investors do not follow the traditional viewpoint blindly. While investors will seek to enhance their economic interests, at least some of them believe that a relationship between corporate social performance and economic performance exists. The next section will discuss the issue of the relationship between corporate social performance and economic performance and how investors and academicians perceive this relationship. The Relationship Between Corporate Social Performance and Economic Performance Ever since corporate social responsibility became increasingly popularized a decade ago, questions have been raised about what, if any, impact corporate social responsibility has on corporate economic per- formance. Some commentarors argue that since being socially responsible costs money, one should expect responsible companies to show a decrease in financial performance. This was the basis of Linowes' proposal (1974) 43 that corporations be given credit for the costs of social responsi— bility in the SEOS. Others argue that only profitable companies could afford social responsibility, and therefore, social responsibility is a by-product of profitability. Bragdon and Marlin (1972) said no; com- panies are profitable because they are responsible. Responsibility is a symptom of foresight, and foresight means good and profitable manage- ment. They feel that social responsibility is good business for a firm planning for the long run. Bauer (Ackerman and Bauer, 1976, p. 18) suggested that profitable companies might be responsible because they were competently run, had the business side of the enterprise in hand, and had the managerial time and skill to address themselves to being responsible. Moskowitz (1972, p. 71) asserts that: Socially sound investments need not, of course, be unsound financially. To the contrary, the argument has been made that the socially aware corporation possesses the special sensitivity that will enable it to surpass competitors. After Milton Moskowitz completed his article "Choosing Socially Responsible Stocks”, (Business and Society Review, Spring, 1972), he asked a number of people in investment firms and mutual funds to comment on the concept of the socially responsible investment portfolio. Royce Flippin, president of First Spectrum Fund, which secure and analyze in— formation on corporate responsibility with regard to prospective invest- ments, holds to the assumption that a company's economic interest cannot be separated from the public interest. Socially responsible companies, he says, are very likely to be good investment vehicles in that their behavior reflects those qualities of management vision and statesman- ship that also produce profits. He is not as interested in corporate benevolence as he is in generating a responsible manner--always bearing 44 in mind such factors as consumer protection, pollution control, and the protection of civil rights. Harold W. Janeway, Research Director of White, Weld and Company, made the following observations: (1) he believes that social criteria do have a growing relevance to investment decisions. However, one must be careful that the social aspect does not dominate...; (2) the more socially responsible companies tend to be more profitable, but it is not at all certain which comes first. One can also agree that the more profitable companies can afford to be more responsible; (3) from a per— formance standpoint it might make more sense to concentrate on avoiding the socially irresponsible companies than to try to profit from a port- folio of the most responsible companies. There are several ways in which a corporation's socially respon- sible activities may contribute to its profitability. It is said that social responsibility is a precondition for survival, if not for the individual business firm, of the business system as a whole. It is also argued that social responsibility makes it possible for a firm to attract a superior type of employee. Social responsibility makes a firm attractive to at least some investors, e.g., mutual funds, universities, churches and some individuals. Social responsibility may attract some customers to a firm's products or services. Finally, social respon- sibility should contribute to making the local community and the society at large a better place to do business (Bauer and Fenn, 1972, p. 63). Marlin (1973, p. 41) asserts that; even those who would restrict busi- ness goals to short-run profit maximization are aware that a company's public image has an impact on its profits or its ability to raise funds. Some consumers may differentiate between firms and between products 45 on the basis of the environmental issue. Moreover, they may be willing to pay higher prices for products which have less of an adverse environ- mental impact than competing products or which are produced by firms that have a reputation for a concern for the environment (The AAA Committee report, 1973; Business Week, 1970a). Further, Narver (1973) argues that certain corporate actions including those dealing with environ- mental improvement may change customer perceptions of the firms and increase both the demand and the elasticity of demand for the products of the firm. Bowman (1973) asserts that social responsibility and profitability are a consequence of good management. He said: Many people interviewed, both institutional investors in Europe and in America, indicate that an appropriate con— cern for corporate social responsibility on the part of a company is a sign of good management and, therfore, is consistent with and necessary to a good investment (pp. 32-33). Public concern over the social consequences of corporate activities can, and has, lead to various types of sanctions (e.g., legislation, litigation and consumer reactions) being directed against corporations in those areas of social concern where their performance is judged to be deficient. Under these conditions, it is entirely possible that a direct relationship exists between a corporation's social performance and its investment worth to investors who are attempting to act in their economic interests. An American banker stated that: ...beyond pure economic issues, any corporation must operate within the social and political structure of the times. As attitudes and problems change, the only way to maximize profits is to be responsive. Those firms that are not responsive will face large and sudden expenditures for meeting legislative standards, possible economic boycotts, or undercutting of their price/earnings ratio by well Street investors who 46 would consider firms with unresponsive managements to be more risky investments (Bowman, 1973, p. 33). Narver (1971) argues that given a society in which there is an ever-rising environmental awareness, there are many potential legal and economic risks to which the capital market is becoming increasingly sensitive. In his opinion, failure of corporate management to respond to the society's expectations in a "socially responsible” fashion may induce the capital market to perceive lower expected earnings and/or impute higher risk factors resulting in a lower present value of the firm. Along the same line, Bowman (1973) argues that the risk associ— ated with a given investment return is lessened for companies with ade- quate social concern. He concluded: ...many institutional investors now argue that the corpo- ration which is not responsive to corporate social respon- sibility will be a more risky investment. While it is true, perhaps, that the research mentioned does not focus on this kind of risk because it deals with statistical risk rather than latent risk (or ”systematic risk" rather than ”residual risk"), and though it is probably overly simple to summarize this argument so briefly, this is equivalent to stating that the corporation which is not socially responsible may be a security investment for its stock- holder in the class of portfolios suggested by point B on the chart. In other words, the investor could have obtained lower risk (in the sense used here) with the same return or, if he had chosen, a higher return with the same risk (p. 34). The relationship between corporate social performance and financial performance can be explained in different ways. Some take the position that corporate social activities cost money and only the richer com- panies can afford such luxury. Bragdon and Marlin (1972, p. 13) argued that this explanation is weak and they gave examples from the paper industry; e.g., Weyerhaeuser and Owens-Illinois, both adopted their environmental stance in the 1940's, long before they also became profit 47 leaders. By contrast, International paper was a profit leader for many years without cleaning up its mills. They concluded that pollution control does not seem to have been limited by the level of profits; it seems to have been an independent management decision. A second argument to explain the relationship between corporate involvement in social concerns and its financial performance is that both social responsibility and financial performance are consequences of good management (Bragdon and Marlin, 1972; Bowman, 1973). Creativity in dealing with social issues is likely to follow from general manage- ment competence, understanding of the ability to influence their envir- onment, and the ability to respond creatively to both public and gov— ernmental pressures for change. A third argument to explain this relationship is that; many cor- porate social activities enhance the company's "good citizen” reputa— tion (Nikolai et a1., 1976, p. 63). Such a reputation, which most companies are concerned with, benefits the company in numerous immea- surable ways including a favorable disposition toward the company with respect to the levying of taxes, the availability of a pool of potential employees and the provision of municipal services. Socially respon- sible companies are less likely to have long drawn—out battles with community groups and governmental agencies, and incur fines and penalties from governmental and private suits. Raising of new equity and debt capital might be facilitated if the company is socially respon- sible. Investors, especially the ethical investors, are becoming increasingly worried about the downside risk associated with companies that are socially irresponsible. This may reduce the price that the investors will pay for new equity and debt capital, thereby increasing 48 the cost of capital for the socially irresponsible companies. All these factors, among others, will contribute to the reduction of costs and increasing of revenues for the socially responsible companies. The next section will review the relevant empirical research related to the topic of the association between corporate social performance and finan— cial performance. The shortcomings in the past research will be dis— cussed and the justification for the present study will be provided. Review of Related Empirical Research The developments in social accounting described above provided the motivation for this study. A number of individuals and institu- tions are considering social iSSues in making investment decisions (Milton Moskowitz, 1972, p. 71). Many institutional investors in Europe and in the U.S. who were interviewed for a study of corporate social responsibility and the investor (Bowman, 1973) indicated that an appro- priate concern for corporate social responsibility on the part of the company is a sign of good management and therefore consistent with and necessary to a good investment. This widely shared view, Bowman said, could be explained in a number of ways: (1) good investments require good company management, and good company management is responsible, wordly and modern, and these traits are evidenced by concern about the involvement in the general socioeconomic problems of our times; or (2) profitable and successful companies have the resources to allocate a portion to social concerns, thus evidencing the power and flexibility of their resources; or (3) corporate activities and expenditures for social concern at an adequate level are really in the self—interest of the firm, permitting, 49 for example, better employee and management recruiting and solving pol- lution problems in a manner and at a pace and time advantages to itself before tightened legal requirement. A somewhat different way to capture these arguments is that the risk associated with a given investment re- turn is lessened with companies with adequate social concerns. In order to answer a question about the relationship between the amount of corporate social responsibility activities and the commercial success or profitability of firms, Bowman and Haire (1975) solicited the annual reports of 82 firms in the food industry listed in Moody's Manual of 1973. They found that companies which evidenced a middle- level amount of corporate social responsibility activities were on the average more profitable (over a 5—year period) than those companies with little reported activities or on the other hand those companies with an extensive amount of such activities. For the low, middle, and high levels of activity, these groups had 5-year average returns on investment of 10.2, 16.1, and 12.3 percent, respectively. Bowman and Haire argue that corporate social responsibility is not, in itself, a simple causal factor of profitability; rather, it is a signal of the presence of a style of management that extends broadly across the entire business function and leads to more profitable oper— ations. It is a diagnostic sign of an appropriate posture in dealing with a multivectored, changing environment. They also argue that re- sponsible activity is not related to a reduction in the return on equity, and that more is not monotonically better than less; rather, there is a u-shaped relationship in which some is better (in terms of association with profits) than little or none and that still more tends to be associated with less return than the middle ground. 50 Bowman and Haire's results and their argument about the association of corporate profitability and the amount of social activities support Janeway's observations. As mentioned earlier, Janeway (in Moskowitz, 1972) noted that from a performance standpoint, it might make more sense to concentrate on avoiding the socially irresponsible companies than to try to profit from a portfolio of the most responsible companies. Along the line of Bowman and Haire's study, Spicer (1976 and 1978) conducted a study to provide some empirical evidence relevant to the social performance disclosure question. He investigated the validity of the widely held view of some investors that a moderate to strong association exists between the investment value of a company's common shares and its social performance. This was achieved by testing for associations between a number of economic and financial indicators of investment value (profitability, size, total and systematic risk, price/earning ratio) and corporate performance on one social issue (pollution control) in a sample of companies drawn from a pollution prone industry (pulp and paper). Some statistically significant asso— ciations were found to exist although there was a reduction in the level of these associations over time. He argued that the reduction in the levels of observed associations over time is due primarily to the marked improvement in the pollution control records as a response to the man- dates of antipollution statutes passed and regulations promulgated or threatened in 1970—1972 period. Bragdon and Marlin (1972) were interested in the question of the compatability between a good pollution control record and a good profit record. The pollution control records of 17 companies in the paper industry were compared to their profit records. Pollution control 51 adequacy was measured by three indices derived from the Council on Economic Priority study. Profits were measured in five different ways. They found strong correlations between profitability measures and pollu- tion control indices. The gross evidence, at least in the paper in- dustry, seems to refute the view of the incompatability between environ- mental virtue and financial rewards; that corporate managers can either control pollution or maximize profits but not both; that the investor either invest in a profitable company or socially responsible company but that no company is likely to be both. The results of Spicer's study may be compared to Bragdon and Marlin's study. It is important to note that both studies utilized the data reported by the Council on Economic Priorities on the pollution control records of 24 companies in the pulp and paper industry (Spicer deleted 6 companies in his study). Spicer's findings, while not strictly comparable with Bragdon and Marlin's findings due to a slightly different sample and a different time period being selected, are not inconsistent with those of Bragdon and Marlin. Heinze (1976) examined the relationship between corporate social involvement and financial position. The study aimed to answer these questions: Does a poor financial position influence the extent of corporate social involvement? Are social involvement and firm's financial position correlated? If so, what financial variables are most highly related to social involvement? Twenty—eight companies were selected. These companies were among fifty companies which were rated by the National Affiliation of Concerned Business Students (NACBS) in regard to their social involvement. The relationship between social involvement (based on the average rating by the NACBS) and financial 52 position (seven financial variables were measured) was examined by means of multiple linear regression and correlation analysis. The social involvement rating was the dependent variable. A high correlation (.65) was revealed and the financial variable most highly related to social involvement was profitability. He suggested that financial condition tends to govern the extent of social involvement. As the correlation coefficient was not unity, he pointed out, it should be also noted that some other factor(s) were also influential in the determination of the extent of corporate social involvement. Fry and Hock (1976) analyzed the annual reports of 135 firms in 15 industries. They were interested in answering the question of what variables (assets, earnings, sales, equity, return on investment, public image of the industry) were related to corporations' social responsive- ness. The degree of corporate "claimed" responsiveness was obtained by searching for evidence of a social responsibility orientation in either the photographs or the text in the annual reports. The public image of the industry was obtained by a questionnaire. The study showed that sales, earnings, equity, assets, and public image of the industry were all related to the emphasis given to social responsiveness in the annual reports. However, return on investment was not significantly related to the claimed responsiveness. Since sales, assets and equity can be considered "size-related" variables, they concluded that the most important factor in determining the annual report space allocated to socially responsible activities was the size of the company. The second was the type of industry, and earnings was a poor third. The big firms and bad—image industries, both of which tended to claim responsiveness, performed no better than others. 53 They noted: These results suggest that the degree of unresponsiveness is a function not of profits but of the type of industry the firm operates in. It indicates that unresponsiveness does not necessarily lead to large profits. In addition, the research suggests that, in general, removing that unresponsiveness—-becoming more responsive——will not adversely affect profits (p. 65). In a research study, Sturdivant and Ginter (1977) investigated two related questions: (1) Do corporate policies and practices reflect the values and attitudes of the top management? Thus, if a given management group believes the role of business in society to be narrowly defined, one would expect its policies and practices to reflect that belief and (2) Is there a relationship between ”perceived corporate social respon- siveness and long—run economic performance?” To answer these questions, they needed representative samples of ”good" and "bad” firms in terms of social performance. They used lists of the "best," ”honorable mention,’ and ”worst” firms originally devel- oped by Milton Moskowitz in the old Business and Society Review news- letter. To answer the first question, the authors surveyed management attitudes on thirteen factors, five of which proved to be statistically different between the best and worst firms. As expected, the honorable mention firms tended to fall in between the two extremes, although they were usually closer to the best firms than the worst. On the subject of economic performance, the authors' findings confirm the assertion that there is a relationship between social performance and economic performance. The survey revealed that both the best and honorable men- tion firms outperformed the worst firms by a significant degree in terms of earnings per share growth over the ten-year period of the study (1964- 1974). There was, however, no significant difference between the best 54 and honorable mention firms. As the authors note (p. 38): While the findings certainly will not support the argument that socially responsive companies will always outperform less responsive firms in the long-run, there is evidence that, in general, the responsively managed firms will en- joy better economic performance. It would be simple minded, at best, to argue a one-on—one cause-effect relationship. However, 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. Preston (1978) provided Sturdivant with a list of Fortune 500 and . . . l . . . of high-reportlng f1rms and asked h1m to prov1de the average executlve attitude score, for high—reporting and other corporations. Preston found that high~reporting companies scored lower than other companies on the Sturdivant-Ginter scale (the scale assigns lower values for gen- erally more liberal social attitudes). As Preston pointed out (p. 148): It can be concluded that within this small sample of firms, those with a higher level of corporate social disclosure also had a more socially conscious group of senior execu— tives. This association does not indicate anything about the direction of effects involved. It may be that the firms became involved in social reporting because their executives already had liberal social attitudes. It could also be that reporting activity served to stimulate or in- crease social concern and involvement among the executives. Probably both effects are at work. It is reassuring to note that the attitude scales and extent of reporting are asso- ciated in the expected direction. Belkaoui (1972 and 1976) investigated the market reaction to the disclosure of pollution control expenditures in the annual report. Two hypotheses were tested: (1) During the months following the disclosure date of pollution control expenditures (annual report date), the common stock prices of disclosing corporations will be favorably affected; 1These are the firms that report an extensive amount on social activities. 55 (2) Corporations identical in asset size and industry classification that do not disclose such expenditures will not be as favorably affected. The author found that for a period of four months after disclosure the performance of the stocks of the disclosing firms was better than the market, indicating a sharp reaction to disclosure, then declined during the rest of the period. This implies a substantial but temporary effect on the stock market performance. This result, Belkaoui said, follows the efficient market hypothesis in its semi-strong form. Under the naive investor hypothesis it verifies the existence of an "ethical investor" who entered the stock market to buy shares on the strength of the better social image conveyed by the disclosure. Belkaoui's study, in general, tends to refute the suggestion that the most deficient reporters of social costs will be rewarded more in the capital market. On the basis of these results, the author said, man- agers may be advised to allocate a proportion of their resources to pollution control and to report these expenditures to the stockholders. Ingram (1977 and 1978) investigated three related questions involving the issues of whether investors find social responsibility disclosure to be relevant for their investment decisions as inferred from the impact of the disclosures on security returns. A sample of 287 firms was selected from the data published by Ernst & Ernst for the years 1971—1975. Three hypotheses were tested using methodologies adopted from the efficient market literature and the application of both parametric and nonparametric statistics. The first hypothesis concerned whether those firms making social responsibility disclosures and those making no disclosure were all from the same or identical population with respect to return behavior. The portfolios' return series were 56 then examined and evidence was collected to warrant rejection of this hypothesis. The second hypothesis concerned whether firms whose ex- penses for social programs were relatively large in proportion to their earnings were from the same population with respect to return behavior as those firms whose expenses were smaller in proportion to their earnings. Evidence was not found to warrant rejection of this hypothe- sis. The third hypothesis tested whether interaction effects existed among disclosure categories (environmental, fair business, personnel, community and product) and between disclosure categories and the other variables (industry, year of disclosure, excess earnings and risk class) which would help to explain the variation in the excess returns of the securities. Interactions were detected between the disclosure cate- gories (especially environmental) and time period, excess earnings, and risk. The author concluded that these interaction effects indicate that investors do respond differently to disclosures made by firms with dif- ferent earnings and risks but not to disclosures made by firms in dif- ferent industries. Ingram's study presents evidence to support the contention that investors in equity securities use social responsibility information in making their investment decisions. Furthermore, the results of the study indicate that investors do not respond to all types of disclosures in the same manner and do not necessarily respond to the same type of disclosure consistently from firm to firm. Instead, the response is conditional upon the firm's financial performance and risk character- istics. All the past studies, in the topic of the association between corporate social performance and financial performance, suffer from 57 potential limitations which, hopefully, may be mitigated in this study. Various studies have attempted to measure social performance using the social responsibility disclosures in the annual reports as a surro- gate for social performance. It is hazardous to draw conclusions about the conduct of companies that make little or no social responsibility disclosures in their annual reports because it is not the only accept- able location for such information. If the annual report does not include social information, it does not mean that the company is not active in this area. It is also reasonable to say that disclosure in this area is not neutral. Furthermore, with specific reference to Ingram's study (1977 and 1978), all the companies which made social responsibility disclosures were treated on an equal basis. No attempt was made to investigate whether the information provided was an indica- tion of good or bad social performance in each specific area; rather the classification made by Ernst & Ernst was used, i.e., firms making monetary disclosures, firms making nonmonetary disclosures, firms (from the same industry as those in the above groups) making no disclosures, and firms (from industries other than those represented above) making no disclosures. Then the security returns were compared for the four groups. In another study the amount of social responsibility disclosures in terms of the number of pages of social disclosure in the annual reports was used as a surrogate for social performance (Bowman and Haire, 1975 and Fry and Hock 1976). The assumption that the larger the report, the better the social performance, is difficult to support. There is no evidence that the number of words in the social report is correlated with the actual performance in this area. 58 Other studies used subjective rating of companies with respect to social performance made by interested groups such as NACBS (Heinze, 1976) or interested individuals such as Milton Moskowitz (Sturdivant and Ginter, 1977). Only one list of companies from only one source was used in each of these two studies. If more than one list were used, this would give more confidence in terms of a consensus of the different sources with respect to the evaluation of corporate social performance. In the studies conducted by Bragdon and Marlin (1972) and Spicer (1976 and 1978), an objective rating of social performance was obtained by deriving pollution control indices based on the Council on Economic Priorities evaluation of the paper companies' pollution control records. However, these studies were limited to one area of social performance, i.e., pollution control and limited to the paper industry. Further- more, the sample was small. Pollution control, while important, is just one area of corporate social involvement and there are many other areas of social responsi— bility that could be considered when measuring corporate social perform- ance. Thus, a company may be strong in one area, such as pollution control, while it may be weak in another, such as minority hiring. The present study, as will be discussed in the next chapter, used the corporate rating approach to identify those companies which are socially responsible and those which are not. This approach has been chosen to avoid the problems encountered when measuring and weighting the diverse aspects of corporate social performance. Also, it is an attempt to overcome the deficiency of using one area of corporate social involve— ment, such as pollution control, and neglecting all other relevant 59 aspects. None of the preceding studies used a control sample (except Belkaoui, 1976). For example, Spicer (1976 and 1978) found that com- panies with adequate pollution control records tended to be larger in size, less risky, and more profitable. Spicer made no attempt to con— trol for company size in his comparison of pollution control and profit- ability. One would argue that Spicer's sample of companies with ade- quate pollution control records were more profitable because they were larger in size and not because they were socially responsible. In this study, as will be discussed in the methodology section, a control group has been chosen from companies with unknown social performance, i.e., the control group will match the socially responsible group in terms of asset size and industry category. This procedure, hopefully, will control as much as possible for confounding variables that might dis- tort the results. Only the studies by Ingram (1977 and 1978) and Belkaoui (1972 and 1976) used risk adjusted returns for comparing companies' financial per- formance. Comparing companies' financial performance on the basis of book rate of return alone ignores the existence and operation of the capital market. Investors are interested in both, total return and risk. Companies with higher total return might have higher risk and when risk adjusted returns are computed for those companies, they might be considered as inefficient portfolios. In this study, financial per- formance is compared in terms of the excess return (risk adjusted return). The next chapter will discuss the research questions, the hypotheses to be tested and the methodology that will be used to answer the research questions. CHAPTER FOUR RESEARCH DESIGN AND METHODOLOGY The purpose of this chapter is to explain the research methodology to be used in answering the research questions. A discussion of the research hypotheses and sampling criteria will be presented. Also, the statistical testing procedures to be used in this study will be deline- ated. Restatement of Objectives and Research Questions The objective of this research is to investigate the association between corporate social performance and financial performance. This objective was derived from an examination of the existing literature and related empirical evidence on corporate involvement in social actions and the relationship of that involvement to the firm's financial performance. Some commentators argue that since being socially responsible costs money one should expect responsible companies to show a decrease in financial performance (Beams and Fertig, 1971 and Chastain, 1975). However, Bowman and Haire (1975) found that companies which evidenced a middle-level amount of activities in corporate social responsibility were on the average more profitable than those companies with little reported activities or those companies with an extensive amount of such activities. Others argue that from a performance standpoint it 60 61 might make more sense to concentrate on avoiding the socially irrespon- sible companies than to try to profit from a portfolio of the most responsible companies. Bowman (1973) argues that the risk associated with a given in- vestment return is lessened for companies with adequate social concern. He argues that corporate involvement in social activities may benefit a corporation if investors associate lower risk with the common stocks of corporations that show adequate social concern than with those that do not. Bowman demonstrated that a socially irresponsible corporation may be a security investment for its stockholders in the class of in- efficient portfolios. In other words, the investor could have obtained lower risk with the same return or, if he had chosen, a higher return with the same risk. Narver (1971) argues that there are many potential legal and economic risks to which the capital market is becoming increasingly sen- sitive. In his opinion, the failure of corporate management to respond to society's expectations in a "socially responsible" fashion may induce the capital market to perceive lower expected earnings and/or impute a higher risk factor resulting in a lower present value of the firm. Spicer (1978) concluded that companies with adequate pollution control records tended to have on the average, lower total and system— atic risk than those companies with poorer pollution control records. Some argue that the risk reducing effects will more than offset the expenditures made for pollution control. "Between firms competing in lHarold Janeway, research director of White, Weld and Company, in Milton Moskowitz (1972). 62 the capital markets those perceived to have the highest expected future earnings in combination with the lowest expected risk from environmental and other factors will be most successful at attracting long-term funds" (U.S. News and World Report, 1970, p. 41). Bragdon and Marlin (1972) assert that social responsibility and profitability are compatible. They argue that social responsibility is a symptom of foresight, and foresight means good and profitable manage- ment. In another argument, socially responsible companies are very likely to be good investment vehicles in that their behavior reflects those qualities of management vision and statemanship that also produce profits.2 Moskowitz (1972) noted that socially responsible investments need not be financially unSOund. Assuming that investors continue to act in accordance with their economic interests, they might consider the socially irresponsible stocks as risky investment. Irresponsible companies could be subject to government regulation and consumer retaliation which would affect their financial performance. It follows that investors might then consider corporate social responsibility as a surrogate for good management. If good management is considered to be a surrogate for a good investment, corporate social performance and financial performance in the market would not be incompatible. Conse- quently the general question which this research is attempting to answer is: Are social performance and economic performance compatible? Unfor- tunately, the preceding general question is difficult to answer directly, therefore, three specific operational questions are introduced below. 2Royce Flippin, president of First Spectrum Fund, in Milton Moskowitz (1972). 63 These specific questions are formulated in such a way that answers to them will provide inferential evidence related to the answer for the general question mentioned above. 1. Does a portfolio constructed from socially responsible stocks outperform a portfolio constructed from socially irresponsible stocks? 2. Does a portfolio constructed from socially responsible stocks outperform a portfolio of randomly selected sample from the market in general? 3. Does a portfolio of socially responsible companies out- perform a comparable portfolio of unknown social perform- ance companies selected to match the socially responsible portfolio in terms of industry code and asset size? Research Hypotheses Based on the research questions in the previous section, the following are the null hypotheses to be tested in this study: 1 . . . . . . . HO: There is no s1gn1f1cant d1fference 1n corporate econom1c performance, in terms of excess return, between a port— folio of socially responsible companies and a portfolio of socially irresponsible companies. 2 . . . H0: The excess return of a portfollo of soc1ally respon31ble companies is not significantly different from zero. Hg: There is no significant difference in corporate economic performance, in terms of excess return, between a port- folio of socially responsible companies and a portfolio of unknown social performance companies selected to match the socially responsible portfolio in terms of industry code and asset size. The first hypothesis is designed to test the economic performance of socially responsible companies against the economic performance of socially irresponsible companies. If the socially responsible group outperforms the socially irresponsible group, this result could be attributed to one or more (since they are not mutually exclusive) of the following: 64 l. Socially responsible companies are rewarded in the market place, 2. Corporate social performance and economic performance (in terms of excess return) are compatible, 3. Social responsibility and profitability are a consequence of good management. 4. There could be a size effect and/or industry membership effect since both groups are not matched. If the socially irresponsible group outperforms the socially responsible group, this result would be consistent with those who main- tain that social activities cost money and companies involved in such activities will experience a decrease in financial performance. Also, there might be a size effect and/or industry effect. Since the socially irresponsible group is not matched to the socially responsible group in terms of asset size and industry code, the result of testing the first hypothesis will be less than conclusive. There is a possibility that company size and/or industry membership, among other factors, may distort the results. This is one reason for including a control group in the study to control for confounding vari- ables such as those mentioned above. The second hypothesis is designed to test the economic performance of socially responsible caompanies against the market, i.e., a randomly selected sample from the market in general. Rejection of the null hypothesis will lead to one of two conclusions. First, socially responsible companies might outperform the market in general. Alterna- tively, there is a possibility that an opposite conclusion could be obtained. In this case, the conclusion would be that socially 65 responsible companies have, in general, lower performance than the market. The third hypothesis is designed to test the economic performance of the socially responsible companies against the economic performance of selected companies with unknown social performance, which have been matched to the socially responsible sample in terms of industry code and asset size. Rejection of the null hypothesis will result in one of two possibilities. First, the socially responsible portfolio might outper- form the control portfolio, in terms of excess return. This result will provide additional evidence that social responsibility and profitability are indeed compatible, that investors profit from holding socially responsible stocks. Second, if the control portfolio outperforms the socially responsible portfolio, this would imply one or more of the following possibilities: 1. that the control group might have good social performance too, but for some reason it was not included in the socially responsible portfolio, 2. at least by avoiding the stocks of companies that have been identified as socially irresponsible, investors will still be making better investment decisions. 3. a high degree of social involvement visia-vis a moderate degree may impair financial performance. Sample Selection Three samples were chosen for the purpose of this study, they are: l. a sample of companies which were defined as socially responsible during the study period. This sample is 66 called the "socially responsible group" (SRG). 2. a sample of companies which were defined as socially irre— sponsible during the study period. This sample is called the ”socially irresponsible group" (SIG). 3. a sample of companies as a control group. This control group has been chosen to match the SRG in terms of industry code and asset size. This control sample was chosen from companies whose social performance was not known. This sample is called the "control group" (CG). The reason behind including a control group in the study is that: the socially irresponsible group was not matched to the socially respon— sible group in terms of asset size and industry category, therefore, any significant difference in financial performance could be attributed, among other factors, to a size effect, industry effect or both. For example, it may be that large companies, in general, are more profit- able and successful and even less risky than small companies. Spicer (1976 and 1978) concluded that companies with adequate pollution control records tended to be larger in size, less risky and more profitable than companies with poorer control records. It would also have been desirable to include a control sample for the SIG. However, the author was unable to match the SIG to a control group in terms of asset size and industry category. The control group has been chosen to match the socially respon- sible group as closely as possible on the basis of their average total asset size during the study period (l971-l975). Matching the firms on total asset size would control for a potential confounding variable, i.e., company size, that might affect both groups in terms of riskiness 67 and/or profitability. In addtiion, the control group has been matched to the SRG with respect to industry membership. Since market risk is determined in part by industry membership, a disproportionate industry membership in either the socially responsible or control groups could cause risk adjustment differences between the two groups if inherent riskiness of these industries change during the test period. Matching the firms by the 3-digit Standard Industrial Classification (SIC) code might control for a factor that might affect both groups in a systematic manner . Corporate Rating Approach It is extremely difficult to construct standards by which a com- pany's social performance can be accurately measured. There are many areas of social responsibility that could be considered when measuring corporate social performance. However, it is impossible, by the avail- able state of the art, to measure corporate social performance quanti- tatively in some areas, such as fair business activities, community involvement, product performance, consumerism, etc. Even in the areas of pollution control and equal employment opportunity which could be quantified and measured (e.g., pollution abatement levels of compliance, amount of emissions per ton of production, and percentage of minority and women in employment categories), a company may be strong in one area, such as pollution control, while it may be weak in another, such as minority hiring. Furthermore, the problem of weighting the diverse aspects of social performance are still largely unresolved. The corporate rating approach is an attempt, largely by groups and individuals outside the corporate structure, to determine which 68 companies are responding well to social and environmental demands and which companies are acting in an irresponsible manner. The move toward corporate social rating has been motivated by universities, foundations, church groups, and certain mutual funds that wish to maintain port- folios of "socially responsible" stocks (Butcher, 1973). Each of these interested groups is working to establish its cri- teria for evaluation. Some of the current rating are highly subjective, relying on some general, composite image of the corporation and its industry. Though some ratings are subjective, they do reflect the corpo— ration's image in the mind of a public and therefore are important. Heinze (1976, p. 49) noted that "because it is presently impossible to measure the true extent of a company's social involvement, it is also difficult to shed any light on the validity of a subjective evaluation." Two major sources of corporate social performance evaluations are dis- cussed below. The periodical, Business and Society Review is almost totally devoted to social responsibility issues and reports on specific corpo— rations' social performance. The periodical publishes both the positive and negative aspects of corporate social performance, but the number of corporations evaluated and the areas of performance discussed, are limited. Business and Society Review may be one of the more objective sources for corporate social performance information. The most scientific work in the field of corporate social rating is being done by the Council on Economic Priorities (CEP).3 The studies 3The Council on Economic Priorities was founded in 1969 as a non- profit organization dedicated to evaluating corporate performance in social impact areas. The key goal underlying the Council's research 69 conducted by the CEP have involved in-depth analysis and reporting with respect to particular firms and industries. The companies are ranked within the industry according to their actual performance with respect to one social area. These studies are substantially more informative than the studies based on an opinion—poll but suffer from the limitation of their narrower scope, a single industry, and even a single location, in some instances, and a single area of social concern. To test the hypotheses of this study, two representative samples of companies were chosen, one sample represents socially responsible firms and the other represents socially irresponsible firms. To avoid the difficulties encountered with measuring corporate social perform- ance, the corporate social rating approach has been used to identify the two representative samples for this study. Interested groups and individuals have published lists of those companies that are responding well to social and environmental demands and those companies which are acting in an irresponsible manner. The search for the lists (of is ”the provision of comparable reports on corporate environmental and equal employment opportunity performance" with the belief that corpora- tions should compete for social as well as earning performance, and to achieve this, full social performance reporting is necessary. The CEP believes that investors, motivated by a return on their invest- ment, as well as by social concern, are able to evaluate the farsighted- ness of the management of the firm in which they invest. Some see a direct relationship between good performance and long-term profit- ability, believing that the best corporate managers do a good job in both financial and social arenas because they can anticipate, learn and respond to the changing environment of business (CEP, 1977, p. 46). The CEP has consistently been an advocate of corporate disclosure of concrete and comparable information about the social consequences of industrial activities, especially in the areas of environmental concern. The Council noted in testimony on the subject in 1975 before the Secu- rities and Exchange Commission that this disclosure was necessary and desirable, not prohibitively expensive, and if done by all competitive companies, not damaging to proprietary interests (CEP, 1977, p. 45). 70 (companies that have been evaluated as to their social responsiveness) to be used in this study began with the year 1970 and continued through 1978. The following sources have been consulted during the process of searching: Business and Society Review, Business Week, Human Resources Network, Interfaith Center on Corporate Responsibility, Council on Economic Priorities, Dreyfus Third Century Mutual Fund, Pax World Fund, National Affiliation of Concerned Business Students, and Mr. Milton Moskowitz, senior editor of Business and Society Review. Fifteen lists have been obtained for companies evaluated according to their social performance. The fifteen lists obtained, cover the period from 1970 through 1975. They consist of nine lists for socially responsible companies, two lists for socially irresponsible companies, and four lists that rank companies according to their social performance. A brief description of the lists, sources, evaluation period and the number of companies on each list, is presented in Table 2. To facilitate the process of selecting the study samples from these lists, each of the four lists, which ranked companies according to their social performance, was further classified into three equal groups of companies (as good, intermediate, and bad social performance). To over- come some of the ambiguity regarding the classification of companies with "intermediate" social performance and to guard against an overlap in ranking, the intermediate group in each list was dropped. Only the top third of the list was considered as socially responsible and the bottom third of the list as socially irresponsible. This procedure resulted in only two types of lists, i.e., lists of socially responsible companies and lists of socially irresponsible companies. Thus, the 71 TABLE 2 DESCRIPTION OF THE LISTS FROM WHICH THE SRG AND THE SIG WERE CHOSEN Classi- Ranked fied as as to to their their social List Compiler No. of social perform- Evalu- com- perform- ance ation panies ance Good Bad period 1 Moskowitz (1974 a & b) 18 X 73-74 2 Moskowitz (1974 a & b) 29 X 73-74 3 Moskowitz (1974 a & b) 20 X 73-74 4 Bus. & Society Review1 43 X 72-75 5 Moskowitz (1972) 14 X 72 6 Moskowitz (1975) 10 X 75 7 Moskowitz (1975) 2 10 X 75 8 Bus. & Soc. Rev. (74/75) 11 X 74 9 Bus. & Soc. Rev. (75/76)2 13 X 75 10 Business Week3 9 x 70-72 11 Bus. & Society Review4 36 X 72 12 Bus. & Society Review5 45 X 72 13 NACB86 50 x 72 14 CEP (1977)7 50 x 70-74 15 Dreyfus Third Century Mutual Fund (l972-78)8 x 72-78 1The editors of Business and Society Review periodically identify those companies which stand out for unusually important action in the public interest. In the section entitled "Corporate Performance Round- up", the editors single out companies which they deem to be ”superior" or especially outstanding in meeting social responsibilities. 2These companies have been given awards from Business and Society Review for their Superior social performance in 1974 and 1975. 3These companies have been given awards for their superior social performance by Business Week during the period 1970-1972. In the Summer of 1972, Business and Society Review send a ques- tionnaire to 215 individuals and groups engaged in monitoring consumer practices. The receipients were all working in the area of consumer affairs and thus were assumed to be able to give knowledgeable answers. They were asked to "compare corporations according to their social responsiveness and accountability to the public interests." They were then invited to ”rank corporations listed on the enclosed sheet (which grouped companies according to retailing groups) according to the recipient's view of the companies' efforts to provide adequate protection 72 and product information to the consuming public." They were further asked to refrain from ranking those retailing groups about which they did not feel qualified. Forty-nine recipients, 23 percent of the total invited, responded in a manner that could be recorded. The survey results were published in the Autumn 1972 issue of the Business and Society Review. Business and Society Review sent a questionnaire to 800 persons, most of them were on corporate staffs working in the urban and public affairs areas. The questionnaire listed 45 corporations, and the recipients were asked to use their knowledge of each company's social programs to rate them. They were asked to skip companies about which they had inadequate knowledge. The Survey results were published in the Spring 1972 issue of the Business and Society Review. 6In the Spring of 1972, Kirk Hanson, president of the National Affiliation of Concerned Business Students (NACBS), a non-profit edu- cational organization of graduate business students formed to promote research on the social role of the corporation conducted a survey to rate 50 companies according to their social performance. Three hundred graduate business students with a special concern for corpo- rate social policy were sent questionnaires. One hundred and fifty students returned the questionnaires rating the overall social perform— ance of the fifty corporations. The editor of Business and Society Review commented on the survey results, he said: "It is important not to overestimate the significance of this survey...it is not a factual analysis of policies or programs. Such a survey runs the risk of rating highly these corporations which have publicized their social involvement more than others" (p. 20). The results of the survey were published in the Summer of 1972 issue, Business and Society Review. 7In a research report, the CEP (1977) developed a ranking of fifty companies based on actual pollution control performance. Companies from four of the worst polluting industries (petroleum refining, iron and steel, pulp and paper and electric utilities) were ranked within their respective industries. The Council noted that "investors moti— vated by a return on their investment, as well as by social concern are able to evaluate the far-sightedness of the management of the firm in which they invest" (p. 46). 8The portfolio of the Dreyfus Third Century Mutual Fund during the period 1972-1978 will be used as a representative of socially responsible mutual funds. The socially responsible mutual funds have been founded to invest in companies that, in one way or another, act in a socially responsible manner. 73 socially responsible lists have been increased by 4 to become 13 lists, and the socially irresponsible lists have been increased by 4 to become 6 lists. The criteria that were established for the selection of the study samples from these lists are discussed below: 1. Any company that appeared in at least three of the socially responsible lists was included in the SRG sample. 2. Any company that appeared in at least two of the socially irresponsible lists was included in the SIG sample. It is important to note that the more lists used to include a com— pany in the sample, the better, in terms of the consensus of different sources upon the evaluation of corporate social performance. On the other hand, if the number of lists required to include a company in the sample was increased (more than three lists for the SRG, or more than two lists for the SIG), the sample size would decrease and might not be adequate for portfolio formulation and the statistical analyses. 3. The lists used in criteria one must cover a period of three years (they need not be consecutive) or more during the study period (1971—1975). Similarly, the lists used in criteria two must cover a period of two years (they need not be consecutive) or more during the study period. Also, no more than one evaluation year was allowed to be missing on both ends of the study time period. For example, if a com- pany was listed on three of the socially responsible lists during the period 1971-1973, it was not qualified for inclu- sion in the sample. On the other hand, if a company was listed on three of the socially responsible lists during the period 1972-1974, it was qualified for inclusion in the sample. 74 Similarly, if a company was listed on two of the socially irresponsible lists during the period 1971-1972, it was not qualified for inclusion in the sample. On the other hand, if a company was listed on two of the socially irrespon- sible lists during 1972 and 1974, it was qualified for inclu- sion in the sample. This criterion was established to assure consistent evaluation of the social performance during a period of time which coincides with study time period. It was expected that the image of the company, in the mind of the public and in particular, the investor, would last for at least one year after the evaluation date. Also, if a company was evaluated as socially responsible in one year, it was assumed that the company was involved in social activities in the preceding year too. Similarly, if a company was evaluated as socially irresponsible in one year, it was assumed that the company was socially irresponsible in the preceding year too. Therefore, it was suggested that no more than one evaluation year should be missing on both ends of the study time period. The three year period of evaluation for the SRG and the two year period of evalu— ation for the SIG were considered to be sufficient for the five year period of the study. 4. Any firm in the SRG which could not be matched to a control firm according to the 3-digit SIC code and/or asset size has been excluded. 5. Complete data on stock prices, dividends, and capital changes for each company must be available from August, 1966 through December, 1975. Following the preceding criteria, 21 firms have been chosen as 75 socially responsible (SRG) and 14 firms as socially irresponsible (SIG). Twenty—one firms were chosen as a control group (CG) to match the SRG in terms of asset size (the average asset size during the study period within plus or minus 15%) and industry category. The major industry groupings for each of the samples is presented in Table 3. As can be seen from Table 3, for each firm in the SRG there is a control firm in the same industry category. By contrast, seven out of the fourteen firms in the SIG are from different industries. TABLE 3 INDUSTRY MEMBERSHIP OF SAMPLE COMPANIES 3-Digit SIC Number of Companies Code CG SIG U) W O 200 204 205 230 240 260 273 280 283 284 291 300 322 331 357 363 371 372 491 492 541 602 631 679 'P‘HJNDF‘F‘H‘I NJHJF‘I I PJKJI I HIAIAIAIA IIAIA II I Ira Irora I l IIwIAIacpIAIA I I I I IF’H‘NJPJH‘HJI NDF‘F‘I I FJRDI I H‘F‘H‘H‘FJI east N l..: N H |._.I .£.\ Total 76 Table 4 presents the asset size distribution of the three groups. The Table indicates that the average asset size of a company in the SRG is almost identical to the average asset size of a company in the CG. By contrast, the average asset size of a company in the SIG is more than 200% larger than the average asset size of a company in either the SRG or the CG. TABLE 4 1971-1975 AVERAGE ASSET SIZE OF SAMPLE COMPANIES Asset Size (000,000 deleted) SRG CG SIG $200 or less 3 3 0 $ 201 — 500 2 3 1 $ 501 - 1000 4 4 1 $1001 - 2000 5 3 4 $2001 — 5000 4 5 3 $5001 — 10,000 2 2 3 Over $10,000 _l_ _l_ ._2 Number of firms in each group 21_ 21_ 14 Average asset size in millions of dollars 2242.8 2160.6 4637.7 Study Time Period The process of searching for lists began with year 1970, and con- tinues through 1978, and only a few companies were evaluated during the years 1970, 1976, 1977, and 1978, thus, the time period 1971-1975 has been chosen as the study period since most of the evaluation lists have been issued during this period. Since the impact of social respon— sibility on corporate financial performance may be long-term in nature, 77 a relatively long time period may be desirable. However, the length of the study time period was limited by the availability of corporate social rating lists and the logical criteria that were established in the preceding section for the selection of the study samples. Also, this time period was considered to be sufficient for the statistical analyses. Financial Data Sources Monthly stock returns for the New York Stock Exchange (NYSE) firms were generated from the Center for Research on Security Prices (CRSP) tape. Monthly closing stock prices for the American Stock Exchange (ASE) firms and bid and asked prices for the Over-the-Counter firms were collected manually from the ISL Daily Stock Price Record and the Wall Street Journal. Data on cash dividends, stock dividends and stock splits were gathered from the ISL Daily Stock Price Record, Moody's and Standard and Poor's Annual Dividend Record. Specification and Measurement of Corporate Financial Performance Since the study is designed to test the association between corpo- rate social performance and financial performance, the latter must be identified and measured for each company in the sample. There are two major approaches to the measurement of financial performance, the accounting-based approach and the market-based approach. The financial statements produced by the accounting system provide the most commonly used estimate of returns. However, the rate of return to investors in the firm varies from the perspective of different sets of economic agents such as the common equity owners, all equity owners, 78 creditors, and lessors of plant and equipment. There are several difficulties in using the book rate of return to compare profitability of firms or even industries (Sunder, 1977). Using the book rate of return to measure corporate financial performance does not permit the control of the difference in risk between firms. In the legal and economic standards of fair or normal returns, it is explicitly recognized that the return on investment in any business should be com- mensurate with the risk involved. It might be the case that the firm or industry with a low rate of return is actually earning more than a normal rate of return if its risk is very low. Accounting based—mea- sures of profitability ignore the existence and operation of the capital market. The book values usually are the historical amounts of capital invested. In an opportunity sense, the measure of investment in a firm is the market price of its securities and not the book value. Although a contemporaneous correlation between accounting and market-based mea— sures is not ruled out and is indeed observed in several empirical tests,4 the correlation, however, is still far from perfrect, and there is little reason to prefer the accounting measures over the market measures. Even if the above mentioned objections are ignored, some diffi— culties remain in the interpretation of accounting measures. One of these is caused by the diversity of accounting numbers used to arrive at a single point estimate of accounting variables found in the finan— cial statements. Generally Accepted Accounting Principles do not 4See Beaver, Kettler and Scholes (1970), Bildersee (1975), and Griffin (1976). 79 define a unique set of accounting procedures. Therefore, interfirm and inter—period comparability of financial statement—based measures of profitability are limited. Another difficulty is caused by infla- tion, i.e., the numerator of the rate of return ratio is based on part historical and part current data, while the denominator is based only on historical data. A procedure for measuring the financial performance of companies using modern portfolio theory and the capital-asset pricing model cir- cumvents several of the difficulties encountered in using and inter— preting the accounting—based estimates mentioned above. The amount of investment in a firm at any time is the opportunity cost of such an investment measured by the market value of the firm's securities, not the sunk costs presented by the book values. In the market oriented approach, an internal measure of risk is estimated for each capital asset or portfolio of assets. The existence of a linear relationship between this measure of risk and average return when the market is in equilibrium allows us to make risk-adjusted estimates of normal and abnormal (excess) returns on individual assets and portfolios. Sharpe (1963) extended the market model which was first suggested by Markowitz (1952 and 1959) to a simplified portfolio model and to a capital asset pricing model. Sharpe's diagonal model states that the return on security i in period t (Rit) is linearly related to the return on the market portfolio (Rmt) by: ~=.+. +. Rit 0'1 B1 Rmt Bit (1) 80 Where, Rit is the monthly return of security i in month t, ~mt is the return on the market portfolio, in month t, di is the regression constant for security i, Bi is the contribution of security i to the portfolio risk, i.e., the security's systematic risk, eit is the individualistic factor unique to security i or unsystematic with respect to the market place. In other words, the portion of security i's return that varies independently of Rmt' The model that has been used in this study's regression analyses is similar to (1) above, it is of the form: = + pt O'pt 8pt Rmt + 8pt (2) Where, N P R = . pt 1&1 th / Np and, N = sample size for each portfolio, p = (SRG, SIG, CG) The monthly return series of each portfolio (sample group) was regressed against the monthly return series of the market portfolio, and an assessment of apt and Bpt were obtained. The monthly return of each portfolio is the average of the monthly returns of all the securities in that portfolio. The monthly return of each security (Rit) is defined as: ~ it = [(Pit + Dit) / Pi,t—l] 1 (3) Where, Pit is the price of security i at the end of month t, D. is dividends per share paid on security i in month t, It 81 The return on the market is computed using the CRSP value weighted index. This index of common stock performance is market value weighted, i.e., the price of each stock is weighted by the number of shares out- standing. The monthly returns for the market (Rmt) are defined as: ..—.. + - mt [(Pmt Dmt) / Pmt-l] 1 Where, mt is the common stock total return during month t, P is the value of the Standard and Poor Composite Int Index at the end of month t, mt is the estimated dividends received during month t and reinvested at the end of month t. The data used for estimation procedures is from the 52 month period ending one month before the month of interest (the month for which an excess return is to be estimated). For example, in order to estimate the excess return for January 1971, return data from August, 1966 through November, 1970 was used.5 Thus, the risk estimates for later months of the study period will use the return data from earlier months of the study and therefore are not independent of the latter. This monthly moving estimate procedure will permit the updating of the beta estimates every month, hence it would catch most of the changes in the risk estimate. This moving procedure, although it does not totally correct for any unknown shift in beta, should provide 5See Fisher (1971) for an argument to the effect that it helps to discard the data for the period immediately preceding the point of estimation -- in this case, December, 1970. 82 corresponding excess returns that will exhibit less bias than those procedures which assume parameter stationarity (Collins and Dent, 1979). The forecasted or expected returns were compared to actual returns during the study period for each month in the following manner for each portfolio (sample group). A ept = Rpt — O'pt — Bpt Rmt (4) The procedure described above resulted in a series of 60 monthly excess returns (ept) for each portfolio (sample group) for the years 1971 through 1975. Statistical Testing Procedures Since the samples for this study were not randomly selected and in order to avoid making assumptions about the distribution underlying the variable of interest, that is, the excess return, nonparametric tests were suggested for investigating hypotheses one and three. A Kolmogorov-Smirnov two—sample test (Siegel, 1956) was used to investigate hypothesis one. If the two samples have in fact been drawn from the same population distribution, then the cumulative dis- tributions of both samples may be expected to be fairly close to each other, inasmuch as they both should show only random deviations from the population distribution. If the two sample cumulative distribu- tions are ”too far apart" at any point, this suggests that the samples come from different populations. Thus, a large enough deviation between the two samples' cumulative distributions is evidence for rejecting H0. The K—S two-tailed test is sensitive to any kind of difference in the distributions from which the two samples were drawn - 83 differences in location (central tendency), in dispersion, in skewness, etc. The K—S test, when compared with the t—test, has high power effi- ciency (about 96 percent) for small samples (Siegel, 1956, p. 136). To investigate hypothesis two which states "the excess return of a portfolio of socially responsible companies is not significantly different from zero," a one sample z-test was used. This parametric test was chosen because the author was unable to discover a nonpara— metric test that can be used to test this hypothesis. If the mean risk adjusted return (ept) is not significantly dif- ferent from zero, this result means that on the average the financial performance of the socially responsible portfolio is not different from the financial performance of a randomly selected portfolio from the market in general. It also means that the portfolio's return is consistent with its expected return given its estimated beta coeffi- cients and that the beta coefficients explain the return series. If the mean risk adjusted return of the portfolio is significantly greater than zero, this result means that the financial performance of that portfolio is higher than the financial performance of a randomly selected portfolio from the market in general, with identical beta. In other words, the performance of that portfolio exceeds its expected performance given its beta. It also means that the portfolio's return is not totally explained by estimated beta coefficients, rather there might be some other factors which contribute to the difference. The SRG and the SIG were not matched in terms of asset size or industry category and these factors might impose extraneous differences between these two groups. One way to overcome the difficulties imposed by extraneous differences between groups is to match or otherwise 84 relate the two samples studied. Hypothesis three was designed to test the differences in financial performance (in terms of excess returns) between the SRG and a matched sample as a control group. A Wilcoxon matched-pairs signed—ranks test was used to test hypothesis three. This nonparametric test is most useful when comparing two related samples. The test considers the relative magnitude as well as the direction of the differences between the matched pairs. The difference score for each matched pair, di’ represents the difference between the pair's scores for the two groups. All di's are ranked without regard to sign. Then the sign of the difference is affixed to each rank. Now, if the two groups are equivalent, that is, if H is true, then we would expect that the sum of the negative ranks 0 and the sum of the positive ranks will be almost equal under H If 0' the sum of the positive ranks is very much different from the sum of the negative ranks, we would infer that the SRG differs from the CG, and thus we would reject HO. It is important to note that when the Wilcoxon test is used for data which in fact meet the conditions of the parametric test, its power efficiency is about 95.5 percent for large samples (N>25) and not much less than that for smaller samples (Siegel, 1956, p. 83). The time period used in testing the research hypotheses was 60 months. If any of the research hypotheses could not be rejected using the 60-month period, a sensitivity test will be performed using dif- ferent time periods. This procedure is necessary to see how sensitive the results are to a change in the test period. The test periods that were chosen for the sensitivity analysis must be between 30 and 60 months. It was not desirable to test a period less than 30 months 85 because of the limited number of observations. Also, no test was made for a period beyond the 60-month study period because the classifica- tion of companies with respect to their social performance might have changed after the end of the study period. For hypotheses one and three, two periods were chosen for the purpose of the sensitivity analysis. These two periods were chosen in such a way that the difference between the two cumulative excess returns of the two portfolios were at the maximum. For hypothesis two, the period at which the absolute difference between the cumulative ex- cess return of the SRG and the zero point was at the maximum was chosen for the sensitivity test. Only one point was chosen because if we could not reject the hypothesis at this point it was felt that the probability of rejection at any other point would be low. A nonparametric statistical test is a test whose model does not specify conditions about the parameters of the populations from which the sample was drawn. Since it was desirable to use parametric tests in addition to the nonparametric tests because they are more powerful when all the assumptions are in fact met in the data, the data were tested for the assumptions required for a parametric statistical test. The monthly risk adjusted return series for each sample were tested for normality. Also, the assumption of equality of variance was tested between the SRG and the CG and between the SRG and the SIG. The results of the tests suggest that the data (excess return series) conform to normality and equality of variance which are required for a parametric test. Hence, a two-independent samples z-test was conducted to test hypothesis one in addition to the K-8 test. Also, a two-related samples z-test was performed to test hypothesis three in addition to 86 the Wilcoxon test. A z-test is a test of equality between two popu- lation means. Utilizing two statistical models to investigate and test the same hypothesis should produce a high degree of confidence in the combined results of these tests than would be the case if one test were used alone. An important assumption undelrying the procedure of analyzing the risk adjusted return series is that the relative market risk (beta) remains relatively stable during the study period. In this study, sequential beta estimates were derived for each group of firms using a 52-month moving estimation period ending one month before the month of interest. This procedure was suggested to reduce as much as possible any systematic bias in the calculated risk adjusted return during the test period. However, the stationarity of the beta coefficients during the study period was tested by splitting the over-all time series of monthly beta observations into two non-overlapping, equal sub-periods (30 months each). A Kolmogorov-Smirnov two—sample test was performed on each beta series in order to compare sub-period one to sub-period two. The results of the statistical tests are presented in Chapter five. CHAPTER FIVE RESULTS OF THE ANALYSIS This chapter presents data on the analysis of the risk adjusted return and beta of the study sample groups (portfolios) during the period of January 1971 to December 1975. Also, interpretations of this study's results are provided in this chapter. Analysis of Excess Return As explained in the methodology section, monthly moving beta coefficients were estimated by applying ordinary least squares regres- sion procedures and the market model to return data for the fifty-two month period ending one month before the month of interest. Monthly excess return series for the 60-month study period were computed for each portfolio (sample group). Table 5 persents summary statistics of the excess return series for each sample group. The contents of Table 5 will be discussed in the context of those sections of this chapter to which the data apply. Use of Nonparametric and Parametric Tests A parametric statistical test requires certain conditions about the parameters of the population from which the research sample was drawn. The meaningfulness of the results of a parametric test depends on the validity of these assumptions. Because of the rigid requirements for the parametric tests, nonparametric statistical tests 87 88 TABLE 5 SUMMARY STATISTICS OF THE MONTHLY EXCESS RETURN (January 1971 — December 1975) Statistics SRG CG SIG Sum: .027331 .148191 .081169 Mean .000455 .002469 .001353 Variance .000225 .000324 .000289 Standard Deviation .015 .018 .017 Range: Minimum -.026 -.O36 -.038 Maximum .039 .046 .040 Normalty: Kurtosis —.554 -.525 .221 Skewness .324 .072 .041 Studentized Range 4.389 4.563 4.604 were suggested to test the hypotheses of this study. Nonparametric statistical tests do not specify conditions about the parameters of the population from which the sample was drawn. It is obvious that the fewer or weaker the assumptions that define a particular statistical model, the less qualifying we need to do about our decision arrived at by the statistical test associated with that model. The data used in the statistical tests were essentially the risk adjusted return series for each portfolio. Since it was deemed 89 desirable to use parametric tests in addition to the nonparametric tests, the data were tested for normality and equality of variance. As can be seen from Table 5, the excess return series during the study period has kurtosis between -.554 and +.221, and the acceptable range for normality is from -1 to +1. The excess return series has skewness between .041 and .324, and the acceptable range for normality is from —.5 to + .5. The Studentized range was calculated for each series and it was between 4.389 and 4.604, and the acceptable range for normality is less than or equal 5 (Roberts, 1974). The assumption of equality of variance was tested between the SRG and the CG (F=.694), and the SRG and the SIG (F=.778). Both tests were insigificant (p = .50) so, this implies that the assumption of equality of variances between groups holds. The results of these tests suggest that the data (excess return series) conform to normality and equality of variance assumptions which are required for a parametric test. Therefore, except for the fact that sampling was not random, the use of parametric tests appear to be acceptable for this study. The parametric tests have been used in addition to the nonpara- metric tests (Kolmogorov—Smirnov and Wilcoxon tests). It should be noted that utilizing two statistical models to investigate and test essentially the same hypotheses introduces an element of redundancy into the results of the analyses. However, the two tests utilize different parts of the information contained in the data and, in addi- tion manipulate these data differently. Therefore, it follows that if the two different statistical procedures yield essentially equivalent results, it is possible to place greater confidence in the results of 90 these tests than would be the case if one test were used alone. It should be noted that all of the statistical tests used in the analyses of the data were two—tailed (i.e., the alternative hypotheses were non- directional). Test of the First Hypothesis The first null hypothesis was as follows: 1 O O O I I I 0 HO: There 18 no s1gn1f1cant d1fference 1n corporate econom1c performance, in terms of excess return, between a port— folio of socially responsible companies and a portfolio of socially irresponsible companies. As a starting point in evaluating the first hypothesis, the monthly cumulative excess returns were computed for the 60—month period for each of the SRG and the SIG portfolios. Figure 1 displays the behavior of the cumulative excess returns of the SRG and the SIG. Using visual ob- servation, the data plotted in the Figure show that the cumulative excess return for the SIG was higher during most of the period, then declined during the last ten months until it was below the cumulative excess return of the SRG during the last five months. In order to statistically test for the existence of a significant difference between the excess return series (i.e., to test hypothesis one) both nonparametric and parametric tests were applied to the excess return series data. A Kolmogrov—Smirnov (K-S) two-sample test was used to compare the excess returns of the SRG and the SIG. This nonparametric test is sensitive to any type of difference in the two distributions--median, dispersion, skewness, etc. The test result was not significant (p=.92). That means that there is no significant difference between the financial performance of the SRG and the SIG with respect to the excess return () SRG 91 FIGURE 1 Portfolio Cumulative Excess Return of the Socially Responsible Sample and the Socially Irresponsible Sample SHE —-+—n—-4 C36? JI— / . I 8):. l I 4 E I , A I '0 . ' .4» l 3 .I ‘) 4— ‘ ’3 ' . Q , a —— " t 4?- "% ,1 ad I a - fl :1 0 , 'J F'- vi 1, ’91 5 : cg + 5‘ I g) I C) (.3 C) C) C) C3 (D D Q C) V CD (‘1 'd O F D c— P" SNHDIBH SSHJX? BAIIUWOWOJ F50 r \ ID 40 (“x UNTH 92 series during the study period (1971-1975). A two-independent samples z-test was conducted on the same data. The null hypothesis of no difference could not be rejected (p = .40). This result supports the K-S test, i.e., there is no significant dif— ference between the socially responsible group and the socially irre— sponsible group in terms of the excess return series. While the pre— ceding two statistical tests performed on the data showed no signifi— cant difference between the SRG and the SIG it may be of interest to note that, as shown in Table 5, the average monthly excess return for the SRG during the study period was —.00045 and it was —.00135 for the SIG. That is, the mean excess return for the SRG was higher than the mean excess return for the SIG. The results of the two preceding tests indicate that the first hypothesis which states "there is no significant difference in corporate economic performance, in terms of excess return, between a portfolio of socially responsible companies and a portfolio of socially irresponsible companies,” could not be rejected. In order to test the sensitivity of the results of the preceding statistical tests to the time period covered, it was considered neces— sary to test hypothesis one for shorter time periods. Two sub-periods were chosen for the application of the sensitivity analysis. The highest two points between 30 and 60 months (at which the difference between the cumulative excess returns of the two groups were at rela- tive maximums) were chosen. A K-S two-sample test and a two—independent samples z-test were performed on the same data but for periods of 33 and 50 months. The difference between the cumulative excess returns of the SRG and the 93 SIG were at relative maximums during these two periods. The results of the K-S test (p = .287) and z—test (p = .210) were not significant for the 33-month period. Also, the results of the K—8 test (p = .393) and z-test (p = .357) were not significant for the 50-month period. The sensitivity test results indicate that the differences between the excess return series of the SRG and SIG were not significant even when the differences were examined at the two relative maximum points of the cumulative excess return series. Two reasons might cause the lack of significant difference between the financial performance of the SRG and SIC. First, the time period of the study was relatively short and the impact of social responsibility on corporate financial performance might be a long-term effect. For example, the cumulative excess return of the SIG declined substantially relative to those of the SRG, in the last seven months of the study period. Second, the SRG companies were smaller in size visJa-vis the SIG companies. If large firms tend to be more profitable than small firms, it is possible that a matched sample of SIC would have exhibited lower financial performance than the SRG. To conclude, the SRG might in fact have performed somewhat better in the market than the SIG. This conclusion has been arrived at by taking into consideration the poten- tial differential size effect and the trend in the cumulative excess returns mentioned above. Also, recall that the mean excess return of the SRG was higher than that of the SIG. To conclude, it seems that social responsibility and financial performance in the market are not incompatible. 94 Test of the Second Hypothesis The second null hypothesis was as follows: H2: The excess return of a portfolio of socially responsible companies is not significantly different from zero. This hypothesis was investigated by performing a one-sample z-test. The null hypothesis of no difference could not be rejected (p = .82). This result means that the mean risk adjusted return (apt) is not sig- nificantly different from zero. Thus, the financial performance of the socially responsible portfolio is not different from the performance of a randomly selected portfolio from the market in general. It also means that the portfolio's return is consistent with its expected return given its estimated beta coefficients and that, beta coefficients explain the return series. In order to test the sensitivity of the preceding result to the time period covered, hypothesis two was tested using a shorter time period. The sensitivity test was performed for a period of 33 months. This period was chosen because the difference (in absolute value) be- tween the cumulative excess return of the SRG and zero point was the maximum during this period. The null hypothesis of no difference could not be rejected (p = .298). This result implies that the financial per- formance, in terms of excess return, of the SRG was not different from the market in general even at the point where the difference between the financial performance of the SRG and the market in general was examined at the relative maximum point. The preceding results tend to suggest that socially responsible companies are not rewarded or penalized in the market for their social activities. 95 Test of the Third Hypothesis The third null hypothesis was as follows: HO: There is no significant difference in corporate economic performance, in terms of excess return, between a port- folio of socially responsible companies and a portfolio of unknown social performance companies selected to match the socially responsible portfolio in terms of industry code and asset size. As was the case for hypothesis one, the starting point to evaluate hypothesis three was the cumulative excess returns of the SRG and the CG. Figure 2 displays the cumulative excess returns of the SRG and the CG. The Figure reveals that the cumulative excess return for the CG was higher than the cumulative excess return for the SRG. Following month ten, the differences in cumulative excess returns became progres- sively larger. In order to statistically test for the existence of a significant difference between the excess return series (i.e., to test hypothesis three) both nonparametric and parametric tests were applied to the excess return series data. A Wilcoxon matched-pairs signed—ranks test was used to compare the excess return of the SRG and the CG. This nonparametric test is most useful when comparing two related samples. Hypothesis three was re— jected (p = .16) in favor of the alternative which states ”a portfolio of companies selected to match the SRG in terms of asset size and industry code outperforms the SRG in the market place." A two-related samples z-test also was performed to test the same hypothesis. The null hypothesis of no difference was rejected. The result of the z—test was significant (p = .20) and it supports the result of Wilcoxon test but at a slightly lower significance level. O SRG 96 FIGURE 2 Portfolio Cumulative Excess Return of the Socially Responsible Sample and the Control Sample C? 4- C.“ x *_T E: D (O C" I ‘1’ " --II- CD _ (xi 0 '7 fl :9 ‘3 .‘ 4L— . , «'4 4 , 4 C.) <> I '— 3 4 a“ I .9 .- ’4 ‘ - 2» \dl L TL _L {L j' 1 IL C) c5 (5 d5 :5 to C3 C3 Q C) C) C3 Q C) C'.) C) "<3 03 (\J ”:1 C) m '- C) c— M SNHOIHH SSHJXE EAIIUWOWOJ MONTH 97 One might argue that we should not have rejected hypothesis three given the moderate levels of alpha obtained. Although the results of the two statistical tests were not highly significant, one should bear in mind that an 80 to 84 percent chance of being right should not be ignored. Perhaps, given the moderate levels as alpha, it would be better to reserve judgment at this point rather than to accept or re- ject the hypothesis. The sensitivity of these results (to a change in the time period) were tested following the same procedure used in testing hypotheses one and two. A Wilcoxon matched-pairs signed-ranks test and a two—related sample z-test were used to test the same hypothesis but for 36 months and 47 months. The difference in the cumulative excess returns between the two groups were at relative maximums during these two periods. The results of the Wilcoxon test (p = .033) and z-test (p = .041) were sig- nificant for the 36-month period. The results of the Wilcoxon test (p = .031) and z-test (p = .040) were also significant for the 47-month period. These results show that there is significant difference in financial performance, in terms of excess return, between the SRG and the CC for the 36 and 47 month periods. In addition to the fact that there was a high, statistically sig- nificant difference between the excess return series of the SRG and the CG for the 36 and 47 month periods, Table 5 shows that the mean excess return of the CG was higher than the mean for the SRG. Given the preceding findings and the moderate degree of statistical signifi- cance found for the 60 month period, it appears that the excess return series of the CG was superior to that of the SRG for the study period. However, it should be pointed out that the excess returns (see Figure 2) 98 of the SRG and CG exhibited a tendency to move closer together over the last six months of the study period. Hence, it is possible that had the study been for a longer period of time, the differences between the two portfolios might have disappeared. I Analysis of Beta and the Issue of Beta Stationarity One important assumption underlying the procedure of analyzing the excess return series for this study's sample groups of firms is that the relative market risk (beta) of the three groups remained relatively stable over time. A significant change in beta from the estimation period to the test period could cause a systematic bias in the calcu- lated residuals or risk adjusted returns during the test period. In this study, sequential beta estimates were derived for each group of firms using a 52-month estimation period ending one month before the month of interest. Table 6 represents summary statistics of the beta series for the three groups. The Table reveals that on the average the SRG and the CG experienced higher betas than the SIG. On the average, the SIG beta was .916 compared to the average betas for the SRG and the CG of 1.072 and 1.095, respectively. This indicates that on the average, the SIG was less risky than the average firm on the New York Stock Exchange. This result is not surprising since the SIG companies are among the largest companies in the samples (the average company size in the SIG was more than twice the average company size in the SRG and the CG). 99 TABLE 6 SUMMARY STATISTICS OF MONTHLY ESTIMATED BETA (January 1971 — December 1975) Statistics SRG CG SIG Mean 1.072 1.095 .916 Minimum .981 1.033 .849 Maximum 1.197 1.190 .975 The moving beta estimates for the SRG, the SIG and the CG are plotted in Figure 3. Inspecting Figure 3 reveals that the SIG's rela- tive risk appears to exhibit some stationarity throughout the study period. By contrast, the SRG and the CG appear to exhibit somewhat of an upward shift in relative risk. The stationarity of the beta coefficients during the study period was tested by splitting the over-all time series of monthly beta obser- vations into two non-overlapping, equal sub-periods (30 months each). A K—S two-sample test was performed on each beta series comparing sub— period one to sub—period two. The null hypotheses of no differences between sub-period one and sub-period two was rejected for each of the three portfolios (p =.0004 or less). The result implies that there was a shift in beta between the first half of the study period and the second half. To assess whether the relative risks of the three groups behaved differently during the study period, monthly changes in the betas of O SRG 100 FIGURE 3 Moving Portfolio Betas of the Social Responsible, Control and the Socially Irresponsible Samples ..A— 1»- .1. \\ (’ 0 + 0 ‘ 0 “L ‘3 {D q— 9‘ {D {b I q— \ L % l t) J 4 l I i C) C5 Cl) C. C.) CD L3 C). CD L3 ( \1 F- CD ’7') CD c- :— s—» H 1 39 OllOleOd (5 IJ‘ ,ij (I (11'. (.3 C3 MONTH 101 each group were computed. The monthly percentage changes for the SRG and the SIG, the SRG and the CG are displayed in Figures 4 and 5 respectively. It appears from inspection of Figure 5 that the risk changes of the SRG and the CG were similar over the entire interval of January, 1971 to December,l972 and, this implies that some charac- teristics of the SRG and the CG are similar. However, the monthly percentage changes in beta displayed in Figure 4 show that the risk changes for the SRG and the SIG were different. This implies that there are some characteristics of the SRG which are different from those of the SIG. An important assumption underlying the procedure of analyzing the risk adjusted return series is that the relative market risk (beta) remains relatively stable during the study period. In this study, sequential beta estimates were derived for each group of firms using a 52-month moving estination period. The moving estimate procedure although it does not totally correct for any unknown shift in beta, should produce corresponding excess returns that will exhibit less of a bias than those procedures which assume parameter stationarity. Of course, if one knew precisely where (if at all) a shift in beta took place, separate estimate could be derived for each set. The problem, of course, is that beta is unobservable and it is difficult to specify exactly where the shifts take place. Moreover, there may be multiple shifts during the period of analysis, thus complicating the identifi- cation task. Notwithstanding the previous comments, the moving beta procedure is considered a useful and reasonably simple way of esti- mating beta and produces less bias in the excess return than those procedures assume parameter constancy. (W SRG SH} 102 FIGURE 4 Percentage Change in the Betas of the Socially Responsible Sample and the Socially Irresponsible Sample -,_ 4.— x. Ek::::FE:F—flqw— + - 0‘ v *4.» - T': :I I . , ,,‘_--llll!!-p. «I: T” 3 - “ :‘ I - .. ‘ l '35 I <3 I W" x l €\ ‘ ‘* - ., A: ' Q" '1" Q I ‘ 3 .. :1: _,_ - - “ 5' «— :‘j‘ a) v, \ . x). :‘ «r— f; fl? 53% ’ .- -'4- .. a ‘Ilro '1 .v :' JI— 3 -c /'":d' - \ n, I I g I g) I g I C) C) CD CD C O Q C O Q C) v— 1— 8138 N1 3‘3NbH3 HOUINHJHHCI .1 .‘J IO MONTH 10 () SRG 103 FIGURE 5 Percentage Changes in the Betas of the Socially Responsible Sample and the Control Sample :3 4.. _ c: .4.— Q I ‘ u ' c‘ C “l— J -C " .‘ 9P 0 g r: O 1' 4— Q" «I» a» $5 = ., C3 ‘1 ( d ‘3 9' ' 4L ‘ -o~ 3’: ‘ .9 .— 0. -~ 9 o.“ .... "4 3p, :1. L 1 1 I L l m I O c; ‘ d) ' d) ' C'D C3 C) C) C) C) C) C) :3 C3 C5 C5 L'A L3 LJA UT‘ CE |-- UJ GD 22 LL. (.5 2? Cf I: L) L3 CD (I }__ 27 Id LJ 1.1.: M O N T H 104 Additional Discussion and Interpretation of Results The results of testing the first hypothesis indicated that while the mean excess return of the socially responsible portfolio was higher than the mean of the socially irresponsible portfolio, the differences in the excess return series were not statistically significant. Mere- over, the relative risk of the portfolio of socially responsible com- panies was higher than the relative risk associated with the portfolio of socially irresponsible companies. The results of the sensitivity tests indicated that the differences between the SRG and the SIG, in terms of excess return, were not sig- nificant even when those differences were at the relative maximum levels. However, the behavior of the cumulative excess returns of the SRG and the SIG suggest the existence of a potential trend in the finan- cial performance of both groups. The excess returns of the SRG were drifting upwards while the excess returns of the SIG were drifting downwards, in the last six months of the study period. Finally, a factor which might have contributed to the lack of a significant difference between the SRG and the SIG is that the average size of the companies in the SRG was 2242.76 million, while the average asset size of the companies in the SIG was 4637.66 million. The soci- ally irresponsible companies were among the larger companies in the samples. It is possible that, in general larger companies tend to be more profitable. The potential size effect on security returns lead to the inclusion of a control group in the study since it was not possible to include a SIG sample of companies which matched the SRG in terms of asset size and industry category as was the case for the SRG. If it were possible 105 to have matched the SRG with the SIG, this procedure might have indi- cated a decreased return and/or increased risk for the SIG, which when taken together might have tended to lower the risk adjusted return for the SIG vis4a-vis the SRG. The results of testing the second hypothesis supported the con- clusion that the financial performance of the SRG was not different from that of the market in general. The sensitivity test indicated that there was no significant difference even when the absolute difference between the cumulative excess return of the SRG and the zero point was at the relative maximum point. The results of testing the third hypothesis suggested that the financial performance of the SRG was lower than the financial perform- ance of the CG but, the attained significance level for the 60-month period should be considered as in a gray area. However, the sensi— tivity tests indicated the existence of a significant difference be- tween the SRG and the CG. This implies that the financial performance of the SRG was increasing relative to that of the CG, over the last six months of the study period. Based on the findings reported above, the statistical test based conclusion is that there was no difference between the SRG and SIG portfolios, in terms of excess returns over the study period. However, if several other factors are taken into consideration, an alternative conclusion might be supportable. Thus, the difference in means, the trend in the return series, and the differences in the sizes of the companies in the two portfolios could be construed as an indication that the financial performance of the SRG was somewhat better than that of the SIG. It was also found that the excess return series of II 106 the SRG was neither superior or inferior to the market in general. With respect to the SRG versus the CG, the results seem to support a conclusion that the CG outperformed the SRG over the study period. However, it is possible that had the study period been longer, the difference in performance might have disappeared. This study's results do not appear to be consistent with the results reported in most of the previous research related to the topic. Recall that several studies concluded that socially responsible corpo- rations, when compared to socially irresponsible corporations, exhibited lower degrees of risk and/or higher degrees of profitability. For example, Sturdivant and Ginter (1977) found that socially responsible companies (the best and honorable mention) outperformed the socially irresponsible companies in terms of earnings per share growth over a period of ten years. Ingram (1977 and 1978) concluded that the firms making social responsibility disclosures experienced higher security returns than those companies making no disclosures. Spicer (1976 and 1978) found that companies with adequate pollution control records were larger in size, less risky, and more profitable than those com- panies with poorer pollution control records. There are many reasons that might cause the inconsistency between this study's results and the past research results. First of all, the samples and the time period in this study were different from the other studies. Also, this study used risk adjusted (excess) returns as the measure of financial performance. Basically this study found that in general, the estimated beta coefficients for the SRG and the CG were consistently higher than those for the SIG. Although beta coefficients for the SIG were lower, the average risk adjusted return 107 during the study period was lower than those of the SRG (although it was not statistically less) and the CG. This result indicates that the total return for the SRG and the CG were larger than the total return for the SIG. The preceding evidence provides a potential explanation as to why the other studies' results indicate that the SRG outperformed the SIG. For example, if Sturdivant and Ginter (1977) or Spicer (1978) had used risk adjusted returns rather than accounting based returns, their results might have come out differently. Alternatively, if this study had used the total return without adjusting for risk, the results would have indicated that the SRG outperformed the SIG. Another factor that might have contributed to a seeming difference in the results of this study versus much of the existing research, is size. Recall that the average size of the firms in the SIG was twice that of the firms in the SRG. Further, recall that Spicer (1978) found that companies with adequate pollution control records were larger in size, less risky, and more profitable. We might infer from Spicer's results that company size, risk and profitability were related variables. That means that the larger the company the higher the profitability, and the larger the company the less risky. One could argue that Spicer's sample of companies with adequate pollution control records were more profitable because they were larger in size. It is interesting to note that the findings of this study do not necessarily conflict with all of the studies which have examined the association between social and financial performance. In particular, there appears to be a similarity between this study's findings and those of Bowman and Haire (1975). They found that companies which 108 evidenced a middle-level amount of activities in corporate social responsibility were on the average more profitable than those companies with little reported activities or those with an extensive amount of such activities. From low, middle and high level activity, these groups had average return on investment of 10.2, 16.1, and 12.3 per- cent, respectively. If the companies in this study's CG can be thought of as similar to Bowman and Haire's firms with a moderate involvement in social activities, then it would be possible to use the findings reported in this chapter to arrive at a conclusion similar to the one arrived at by them. In summary, this study's findings tend to support the following set of conclusions. Statistically, it can be argued that all the three groups are equal, with respect to their financial performance in the market over the 60—month period covered in this study. But, if one group tends to do better, in terms of financial performance in the market, than the others, it is the control group rather than the socially responsible group. It can be concluded that, with respect to performance in the market, there might be some merit in avoiding highly socially responsible companies. However, it must be kept in mind that the directly preceding conclusion only applies to the 60 month period covered by this study since there was some weak evidence that the SRG and CG excess returns were converging in the last six months of the study period. With respect to the SRG and the market in general, it appears that there is no financial penalty or reward for investing in the equity securities of those companies that exhibit a high degree of social involvement. Turning to the implications of the findings for 109 the socially responsible firms versus the irresponsible firms, the conclusion is that there may be no difference in their financial per- formances. However, if there is a difference, the evidence seems to favor the socially responsible companies over the irresponsible com- panies. From a purely economic point of view, the findings of this study are not inconsistent with the suggestion that the financial perform— ance of this study's portfolios can be ordered from best to worst as follows: the control group, the socially responsible group, and then the socially irresponsible group. Perhaps Janeway (Moskowitz, 1972) was correct when he observed that from a performance standpoint it might make more sense to concentrate on avoiding the socially irrespon— sible companies than to try to profit from a portfolio of the most responsible companies. CHAPTER SIX SUMMARY AND CONCLUSIONS This chapter will provide a summary of the statement of the prob- lem, purpose of the research and research methodology used to answer the research questions. Also, a summary of findings and conclusions will be presented. The limitation of the study and suggestions for future research will be discussed. Statement of the Problem During the past decade, there have been considerable social, political and economic pressures on corporate management to pay greater attention to its social responsibilities. Some corporations have responded to these pressures with increased expenditures and actions in several areas of social concern. Some accountants have responded by developing a number of proposals aimed at the construction and application of better methods of measuring and controlling the social performance of private enterprises. A number of individuals and insti— tutions have responded by considering social issues in making their investment decisions. The issue of whether investors consider corporate social perform- ance when making investment decisions raised the question of whether there is a relationship between corporate social performance and economic performance. Questions have been raised about what, if any, 110 111 impact corporate social responsibility has on corporate financial per- formance. Some take the position that socially responsible actions cost money, therfore responsible companies must show a decrease in financial performance. Others argue that only profitable companies could afford social responsibility, and therefore, social responsi- bility is a by—product of profitability. Others said no; companies are profitable because they are responsible. In general the available empirical evidence tends to indicate that a positive association exists between corporate social responsibility and financial performance. For example, Spicer (1976 and 1978) found that companies with adequate pollution control records were larger in size, less risky and more profitable than companies with poorer pollu- tion control records. Heinze (1976) concluded that companies' finan- cial condition tended to govern the extent of social involvement. Sturdivant and Ginter (1977) found that the best socially responsible companies outperformed the worst socially irresponsible companies in terms of earnings per share growth. However, Bowman and Haire (1975) found that while companies which evidenced a high level of social activities were on the average more profitable than those companies with little reported activities, those companies with a moderate level of social activities tended to be more profitable than the firms with an extensive amount of such activities. Unfortunately, each of the studies which addressed the issue of the association between corporate social and financial performance was subject to certain limitations. 112 Purpose of the Research The primary purpose of this research was to investigate the asso- ciation between corporate social performance and financial performance in the market. The objective of this investigation was to provide inferential evidence related to the answer for the following question: Are corporate social performance and financial performance compatible? Since it was felt that the past research in this area suffered from several potential limitations, an alternative methodology was used to answer the previous question. Three operational questions were formulated, to help in answering the general question, as follows: 1. Does a portfolio constructed from socially responsible stocks outperform a portfolio constructed from socially irresponsible stocks? 2. Does a portfolio constructed from socially responsible stocks outperform a portfolio of randomly selected sample from the market in general? 3. Does a portfolio of socially responsible companies outperform a comparable portfolio of unknown social performance companies selected to match the socially responsible portfolio in terms of industry code and asset size? Research Methodology Fifteen lists of companies that have been evaluated according to their social performance were obtained from different sources. Criteria were established for the selection of the study samples from these lists. Three samples were chosen for the purpose of this study. The first sample was chosen from those companies which were identified as socially responsible during the study period (1971—1975). The second sample was selected from those companies which were identified as socially irresponsible during the study period. The third sample 113 was selected, as a control group, from those companies whose social performance were not known. This third sample was matched to the first sample in terms of asset size and industry category. A procedure for measuring the financial performance of the samples was established using modern portfolio theory and the market model. Monthly moving beta coefficients were estimated by applying ordinary least squares regression procedures and the market model to return data for the fifty—two month period ending one month before the month of interest. A measure of corporate economic performance (defined as the risk adjusted return ) was computed for each month and for each sample during the study period. Parametric and nonparametric statistical comparisons were made between the measures of economic performance computed for each of the three samples in addressing the major question posed in this research. A Kolmogorov—Smirnov two-sample test and a two-sample z—test were used to compare the financial performance of the socially responsible sample and the socially irresponsible sample. A one—sample z-test was used to compare the financial performance of the socially respon— sible sample to the market in general. A Wilcoxon matched—pairs signed- ranks test and a two-related samples z-test were used to compare the financial performance of the socially responsible sample and the con- trol sample. Findings and Conclusions The economic performance of the socially responsible portfolio was not statistically different from the economic performance of the socially irresponsible portfolio since the results of the K-S and 114 z-test were not significant. Two reasons might cause these results. First, the average company size in the socially responsible group (SRG) was less than fifty percent the average company size in the socially irresponsible group (SIG). Given that the socially responsible com- panies were smaller in size visJa—vis the socially irresponsible com- panies, if large firms tend to be more profitable than small firms, it is possible that a matched sample of SIG would have exhibited lower risk adjusted returns than the SRG. Another factor which might have contributed to the lack of significance is the time period used in the study. The study's time period was relatively short (only five years) and the impact of social responsibility on corporate financial perform- ance might be a long—term effect. For example, the cumulative excess return of the SIG declined substantially relative to those of the SRG, in the last seven months of the study period. While the excess return series of the SRG and SIG portfolios were not significantly different, an analysis of the basic data would not be inconsistent with a conclusion that the SRG might in fact have per- formed somewhat better in the market than the SIG. This conclusion has been arrived at by taking into consideration the potential dif- ferential size effect and the trend in the cumulative excess returns mentioned above. Also, recall that the mean excess return of the SRG was higher than that of the SIG. The results of the study also indicated that the financial perform- ance of the SRG was not statistically different from the market in general. However, the financial performance (excess return) of the SRG appeared to be lower than that of the CG. The preceding observa- tion is based on the fact that there was a moderate degree of statistical 115 significance (p = .16) between the excess returns of the two groups. Also, the mean excess return of the CG was higher than that of the SIG. The following tentative conclusions have been drawn from the study's results. First, the equity securities of companies which exhibit a high degree of social responsibility may tend to perform somewhat better in the market than the equity securities of socially irresponsible companies. Second, the equity securities of companies that are not identified as either socially irresponsible or highly responsible may tend to perform somewhat better in the market than the equity securities of companies which exhibit a high degree of social responsibility. However, this second conclusion should be viewed with a degree of caution since the cumulative risk adjusted (excess) returns of the socially responsible and control samples exhibited a tendency to move closer together over the last six months of the study period. Limitations of the Study Several limitations of the research warrant discussion. First, the firms in this study were not randomly sampled. Therefore, it may be inappropriate to generalize the results of this study beyond the firms in the samples. Second, the time period in this study was rela— tively short (only five years), while the impact of social responsi— bility on corporate financial performance might be a long—term effect. Third, the sample size was relatively small. The reliability of the empirical results is related to the appro- priateness of the aSSumed return generating process and the procedures used to obtain the results. For example, the use of the market model 116 as the assumed process by which period—by—period returns are generated for each firm implies that the relative risk measure, beta, for each firm is stationary and can be estimated with little or no error. While, beta was estimated at the portfolio level (portfolio level betas are relatively more stable than at the firm level) there was some evidence of a shift in beta for all three groups. However, a moving beta esti— mate procedure was used in this study. Although this moving procedure does not totally correct for any change in beta, it should produce less of a bias in the excess return than those procedures which assume para- meter constancy. Furthermore, the validity of the market model has been questioned by Roll (1977) among others. He pointed out that the assumed linearlity relation between expected return and beta follow from the market portfolio's efficiency and are not independently test— able. Also, he has demonstrated various practical difficulties in testing the validity of the capital asset pricing model, one of them being insufficient information abOut non—marketable assets. The conclusions and interpretations related to the findings are dependent on the approach utilized to classify companies as socially responsible or socially irresponsible. Because it was extremely diffi- cult to construct criteria by which a company's social performance can be accurately measured, the corporate rating approach was relied upon in classifying companies as socially responsible and socially irre- sponsible. We must keep in mind that while the ratings reflect the corporation's image in the mind of a public, the ratings are subjec- tive. We must also keep in mind that because it is presently impos— sible to measure the true extent of company's social involvement, it is also difficult to shed any light on the validity of a subjective 117 evaluation. Suggestions for Future Research Several limitations described in the preceding section may possibly be overcome by further research. This study did not include a control sample for the socially irresponsible portfolio. A suggestion for future investigation would be to address the same research question posed here by including an additional sample in the design as a control portfolio for the socially irresponsible portfolio. Perhaps, as time passes it will even become possible to form matched samples of socially responsible and irresponsible companies. An additional suggestion for future research is to increase the sample size and extend the time period of the study. The association between corporate financial performance and social performance might be better explored in a longer time period Such as ten years, than in the five year time period used in this study. 118 APPENDIX A Socially Responsible Sample Company Name Aetna Life & Casualty Co. Bank American Corp. Chase Manhattan Corp. Consolidated Edison of N.Y. Cummins Engine Inc. Dow Chemical Co. El—Paso Co. First Pennsylvania Corp. Jewel Companies Inc. Kimberly Clark Corp. Maremont Corp. McGraw-Hill Inc. Mobil Oil Corp. Owens-Illinois Inc. Phillips Van Heusen Corp. Quaker Oats Co. Ralston Purina Co. Shell Oil Co. Weyerhaeuser Co. Whirlpool Corp. Xerox Corp. *Asset size in millions of dollars. Asset Size* 148. 1,989. 1,156. 6,121. 542. 4,647. 2,153. 299. 657. .483 1,144 172. 424. 11,516. 1,722. 175. 623. 1,168. 5,667. 2,717. 716. 3,333. Based upon size during the study period (1971—1975). 399 158 027 968 731 256 712 082 367 099 799 794 742 197 189 589 733 991 095 309 3-Digit SIC Code 631 679 602 491 371 280 492 602 541 260 371 273 291 322 230 200 204 291 240 363 357 the average total asset 119 APPENDIX A--Continued Socially Responsible Control Sample Company Name Atlantic Richfield Co. British Petroleum Co. Cpc Intl Inc. Connecticut General Insurance Co. Continental Group Inc. First National City Corp. Georgia—Pacific Corp. Grolier Inc. Honeywell Inc. Houdaille Industries, Inc. Marine Midland Banks Inc. Morgan (J. P.) & Co. Pacific Gas & Electric Philsbury Co. St. Regis Paper Co. Southland Corp. Texas Eastern Corp. Timken Co. Union Carbide Corp. Vanity Fair Mills Inc. White Consolidated Industries, Ltd. Inc. *Asset size in millions of dollars. 3-Digit SIC Asset Size* Code 5,591.652 291 11,225.599 291 1,272.837 204 161.213 631 1,880.104 322 2,101.860 679 2,126.627 240 418.353 273 2,462.794 357 169.223 371 337.172 602 836.033 602 6,089.987 491 621.246 200 1,201.350 260 441.730 541 2,270.358 492 520.710 371 4,787.938 280 191.382 230 663.652 363 Based upon the average total asset size during the study period (1971—1975). 120 APPENDIX A——Continued Socially Irresponsible Sample Company Name American Can Co. American Electric Power American Home Products Corp. Bethlehem Steel Corp. Colgate—Palmolive Co. Florida Power & Light Co. Good Year Tire & Rubber Co. Great Atlantic & Pacific Tea Co. Gulf Oil Corp. Nabisco Inc. Northrop Corp. Standard Oil Co. Texaco Inc. U. S. Steel Corp. of Calif. *Asset size in millions of dollars. 3—Digit SIC Asset Size* Code 1,669.369 322 5,854.039 491 1,145.150 283 4,024.218 331 1,107.139 284 2,951.134 491 3,956.311 300 1,005.427 541 10,758.399 291 854.907 205 405.164 372 9,843.551 291 14,199.885 291 7,152.559 331 Based upon the average total asset size during the study period (1971—1975). LIST OF REFERENCES Abt, Clark C., The Social Audit for Management, Amacom, 1977. Ackerman, Robert W., and Raymond A. Bauer, Corporate Social Respon— siveness, The Modern Dilemma, Reston Publishing Co., Inc., 1976. American Accounting Association, ”An Evaluation of External Reporting: A Report of the 1966—1968 Committee on External Reporting,” The Accounting Review, Supplement to Vol. XLIV, 1969, pp. 79-123. , "Report of the Committee on Nonfinancial Measures of Effectiveness," The Accounting Review, Supplement to Vol. XLVI, 1971, pp. 165—211. , "Report of the Committee on Measures of Effectiveness for Social Programs,” The Accounting Review, Supplement to Vol. XLVII, 1972, pp. 337—396. , "Report of the Committee on Environmental Effects of Organization Behavior,” The Accounting Review, Supplement to Vol. XLVIII, 1973, pp. 72—120. , "Report of the Committee on Measurement of Social Costs," The Accounting Review, Supplement to Vol. XLIX, 1974, pp. 99—113. , "Report of the Committee on Social Costs," The Accounting Review, Supplement to Vol. XLX, 1975, pp. 51-89. , ”Report of the Committee on Accounting for Social Per- formance,” The Accounting Review, Supplement to Vol. XLXI, 1976, pp. 39—69. American Institute of Certified Public Accountants, The Study Group on the Objectives of Financial Statements, Objectives of Financial Statements, AICPA, 1973. , The MeaSurement of Corporate Social Performance, AICPA, 1977. Arrow, Kenneth, "Social Responsibility and Economic Efficiency," Public Policy (Fall 1973), pp. 303—317. Backman, Jules, ”Social Responsibility: The Problem and its Setting," in Social Responsibility and Accountability, edited by: Jules Backman, New York University Press, 1975, pp. l-23. 121 122 Bauer, Raymond A. and Dan H. Fenn, Jr., The Corporate Social Audit, Russell Sage Foundation, 1972. Baumol, William J., The Environmental Protection at Minimum Cost: The Pollution Tax,” in Social Accounting: Theory, Issues, and Cases," edited by Lee Seidler and Lynn Seidler, Melville Pub— lishing Co., 1975, pp. 240—246. Beams, Floyd A., and Paul A. Fertig, "Pollution Control Through Social Cost Conversion,” Journal of Accountancy (November 1971), pp. 37- 42. Beaver, W., P. Kettler and M. Scholes, "The Association Between Market Determined and Accounting Determined Risk Measures," The Accounting Review (October 1970), pp. 654—681. Belkaoui, Ahmed, ”The Impact of the Disclosure of Pollution Control Information on Investors: A Behavioral Field Experiment and a Market Reaction Investigation," Ph.D. Dissertation, Syracuse University, 1972. , "The Impact of the Disclosure of the Environmental Effects of Organizational Behavior on the Market," Financial Management (Winter 1976), pp. 26—31. Bildersee, John S., "The Association Between a Market Determined Risk and Alternative Measures of Risk," The Accounting Review (January 1975), pp. 81—98. Blake, David H., William C. Frederick, and Mildred S. Myers, Social Auditing: Evaluating the Impact of Corporate Programs, Praeger Publishers, 1976. Bowman, Edward H., "Corporate Social Responsibility and the Investor," Journal of Contemporary Business (Winter 1973), pp. 21—43. , and Mason Haire, "A Strategic Posture Toward Corporate Social Responsibility," California Management Review (Winter 1975), pp. 49-58. , "Social Impact Disclosure and Corporate Annual Reports,” Accounting, Orggnizations and Society, Vol. 1, No. l (1976), pp. ll—21. Bragdon, Joseph H. Jr., and John A. T. Marlin, ”Is Pollution Profit— able?" Risk Management (April 1972), pp. 9-18. Brummet, R. Lee, ”Total Performance Measurement, Management Accounting (November 1973), pp. ll—15. Business and Society Review, "Corporate Performance Roundup," 1972-78 issues. 123 , "Corporate Responsibility Awards, 1974: This Year's Good Guys," (Winter 1974-1975), pp. 12-20. "The 1975 Good Guys," (Winter 1975—1976), pp. 18-27. v "Industry Rates Itself," (Spring 1972), pp. 96—99. , "How Business School Students Rate Corporations,” (Summer 1972), pp. 20-21. , "How Consumer Organizations Rate Corporations,” (Autumn 1972), pp. 94—96. Business Week, ”The Trade—Offs: A Better Environment," (April 11, 1970a), p. 73. , "Pollution Foes Find a Friend at the Bank," (July 4, 1970b), pp. 20—21. Butcher, Bernard L., ”The Program Management Approach to the Corporate Social Audit,” California Management Review (Fall 1973), pp. ll—l6. Buzby, Stephen L., and Haim Falk, "Demand for Social Responsibility Information by University Investors," The Accounting Review (January 1979), pp. 23-37. Chastain, Clark E., ”Corporate Accounting for Environmental Information,” Financial Executive (May 1975), pp. 45~50. Churchill, Neil C., "The Accountant's Role in Social Responsibility," in The Accountant in a Changing Business Environment, edited by Williard E. Stone, University of Florida Press, 1973, pp. 14-27. , "Corporate Accountability," Working Paper, Harvard Business School, March 1978. Collins, Daniel W., and Warren T. Dent, "The Proposed Elimination of Full Cost Accounting in the Extractive Petroleum Industry: An Empirical Assessment of the Market Consequences," Journal of Accounting and Economics (March 1979), pp. 3—44. Council on Economic Priorities, ”Minding the Corporate Conscience," Economic Priorities Report, Vol. 4, No. 1, 1973. , "Minding the Corporate Conscience," Economic Priorities Report, Vol. 5, No. l, 1974. , The Pollution Audit: A Guide to 50 Industrials for Respon— sible Investors, CEP, 1977. Dilley, Steven C., and Jerry J. Weygandt, "Measuring Social Respon— sibility: An Empirical Test," The Journal of Accountancy (September 1973), pp. 62—70. 124 Dreyfus Third Century Fund, Inc., Prospectus, September, 1978. Epstein, Marc, Eric Flamholtz, and John J. McDonough, "Corporate Social Accounting in the U.S.A.: State of Art and Future Prospects,” Accounting, Orggpizations and Society, Vol. 1, No. l (1976), pp. 23-42. Ernst and Ernst, Social Responsibility Disclosure, 1977 Survey of Fortune 500 Annual Reports, 1978. Estes, Ralph W., Corporate Social Accounting, John Wiley & Sons, Inc., 1976. Fisher, Lawrence, ”On the Estimation of Systematic Risk," Paper pre— sented at the Wells Fargo Symposium, July 26—28, 1971. Flamholtz, Eric, Human Resource Accounting, Dickenson Publishing Co., 1974. Friedman, Milton, Capitalism and Freedom, The University of Chicago Press, Chicago, 1962. Fry, Fred L., and Robert J. Hock, "Who Claims Corporate Responsibility? The Biggest and the Worst," Business and Society Review (Summer 1976), pp. 62-65. Griffin, Paul A., ”The Association Between Relative Risk and Risk Estimates Derived from Quarterly Earnings and Dividends," The Accounting Review (July 1976), pp. 499—515. Heinze, David C., "Financial Correlates of a Social Involvement Mea— sure," Akron Business and Economic Review (Spring 1976), pp. 48—51. Ingram, Robert Wilson, ”The Impact of Social Responsibility Disclosures on Security Returns," DBA Dissertation, Texas Tech University, 1977. "An Investigation of the Information Content of (Certain) Social Responsibility Disclosures,” Journal of Accounting Research (Autumn 1978), pp. 270-285. Jensen, Robert E., Studies in Accounting Research #14, "Phantasmagoric Accounting: Research and Analysis of Economic, Social and Environ— mental Impact of Corporate Business," American Acc0unting Associ- ation, 1976. Johnston, David C—H, "Social Measurement and Reporting,” Corporate Responsibility Planning Service, Human Resources Network, #47% May 5, 1978. Kapp, K. William, The Social Cost of Private Enterprise, Schocken Books, 1971 125 Ladd, Dwight R., Contemporary Corporate Accounting and the Public, Richard D. Irwin, Inc., 1963. Linowes, David F., "Measuring Social Programs in Business," Paper delivered to the American Accounting Association, Baton Rouge, La., April, 1972. , "Let's Get on with the Social Audit: A Specific Proposal," Business and Society Review (Winter 1972—1973), pp. 39-42. , The Corporate Conscience, Hawthorne Books, Inc., 1974. Longstreth, Bevis, and H. David Rosenbloom, Corporate Social Respon— sibility and the Institutional Investor, Praeger Publisher, 1973. M.S.U. State News, Michigan State University, December 8, 1978. Markowitz, Harry, "Portfolio Selection," Journal of Finance (March 1952), pp. 77—91. , Portfolio Selection: Efficient Diversification of Invest- ments, John Wiley, 1959. Marlin, John Tepper, ”Accounting for Pollution,” Journal of Accountancy (February 1973), pp. 41-46. Miller, Arjay, "The Social Responsibility of Business," A Look at Busi— ness in 1990, Washington, D.C., White House, Conference on the Industrial World Ahead, 1972, pp. 85-89. Moskowitz, Milton R., ”Choosing Socially Responsible Stocks," Business and Society Review (Spring 1972), pp. 71—75. , ”Social Responsibility Portfolio 1973," Business and Society Review Newsletter (15 January 1974a), p. 1. ”46 Socially Responsible Corporations,” Business and Society Review Newsletter (2 July 1974b), p. 8. , "Profiles in Corporate Responsibility," Business and Society Review (Spring 1975), pp. 28—42. Narver, John C., "Rational Management Responses to External Effects," Academy of Management Journal (March 1971), pp. 99-115. National Association of Accountants, "Report of the Committee on Accounting for Corporate Social Performance," Management Accounting (February 1974), pp. 39—41. Nikolai, Loren A., John D. Bazley and R. Lee Brummet, The Measurement of Corporate Environmental Activity, National Association of Accountants, 1976. 126 Preston, Lee E., "Analyzing Corporate Social Performance: Methods and Results," Journal of Contemporary Business (Winter 1978), pp. 135-150. Ramanathan, Kavasseri V., "Toward A Theory of Corporate Social Accounting," The Accounting Review (July 1976), pp. 516-528. Roberts, Harry V., Conversational Statistics, The Scientific Press, 1974. Roll, Richard, "A Critique of the Asset Pricing Theory's Tests, Part 1: On Past and Potential Testability of the Theory," Journal of Financial Economics (March 1977), pp. 129-176. Sater, Clair, ”A Supplement to the Bottom Line: Rating Corporations on Social Responsibility,” Stanford University Graduate School of Business Bulletin (Summer 1971), pp. 18-21. Seidler, Lee J., "Dollar Values in the Social Income Statement," in: Social Accounting: Theory, Issues and Cases, edited by Lee J. Seidler and L. L. Seidler, Melville Publishing Co., 1975, pp. 1-13. Seidler, Lee J., and L. L. Seidler, Social Accountigg: Theory, Issues, and Cases, Melville Publishing Co., 1975. Sharpe, William, "A Simplified Model for Portfolio Analysis,” Manage- ment Science (January 1963), pp. 277—293. Sheldon, Eleanor B., and Robert Parke, "Social Indicators and Corporate Social Performance: A Case of Delayed Parrallelism," in: Social Responsibility and Accountabiligy, edited by Jules Backman, New York University Press, 1975, pp. 77-92. Shocker, A. D., and S. P. Sethi, ”An Approach to Incorporating Societal Performances in Developing Corporate Action Strategies," in: The Unstable Ground: Corporate Social Policy in a Dynamic Society, edited by S. P. Sethi, Melville Publishing Co., 1974, pp. 67-80. Shulman, James S., and Jeffrey Gale, "Laying the Groundwork for Social Auditing," Financial Executive (March 1972), pp. 38—42. Siegel, Sidney, Nonparametric Statistics, McGraw—Hill Book Company, 1956. Spicer, Barry H., "Investors, Environmental Protection and the Capital Market: An Empirical Investigation," Ph.D. Dissertation, Univer- sity of Washington, 1976. , "Investors, Corporate Social Performance and Information Disclosure: An Empirical Study," The Accounting Review (January 1978), pp. 94-111. Sturdivant, Frederick D., and James L. Ginter, "Corporate Social Respon— siveness: Management Attitude and Economic Performance," California Management Review (Spring 1977), pp. 30—39. 127 Sunder, Shyam, Oil Industry Profits, American Enterprise Institute for Public Policy Research, Washington, D.C., 1977. United States News and World Report, 1970, p. 41. Wallich, Henry, "How Business Can Rescue Capitalism," Fortune (March 1972a), pp. 123-124. , Editorial page, Wall Street Journal, May 31, 1972b. llIllllIIIIIIIIIIIIIIIIIIIlllIIIIIIITINIIIIIIIIIIlIIIIIIIIll