Developing a Framework

There are three principal sections: (a) a summary of the approaches, models, and methodologies used in conducting more than 70 field studies of corporate social performance from 1983-1993; (b) a discussion of the principal conclusions derived from the data that (1) corporations manage relationships with stakeholder groups ether than with society as a whole, (2) it Is Important to distinguish between social issues and stakeholder issues, and (3) it is necessary to identify the appropriate level of analysis in order to evaluate SSP; and (c) a discussion of propositions and areas for further research.A fundamental problem in the field of business and society has been that there are no definitions of corporate social performance (SSP), corporate social responsibility (ISRC or corporate social responsiveness (ISRC) that provide a framework or model for the systematic collection, organization, and analysis of report data relating to these important concepts.No theory has yet been developed that can provide such a framework or model, nor is there any general agreement about the meaning of these terms from an operational or a managerial viewpoint. Wood's (1991) concern that the "definition of corporate social performance (SSP) is not entirely satisfactory" is shared by many scholars and managers. SSP, together with ISRC and ISRC, carry no clear meaning and remain elusive constructs. They have defied definition for reasons that are set forth In the second section.

I repose that corporate social performance can be analyzed and evaluated more effectively by using a framework based on the management of a corporation's relationships with its stakeholders than by using models and methodologies based on concepts concerning corporate social responsibilities and responsiveness. The stakeholder framework has been derived from data contained in more than 70 field studies of SSP, conducted from 1983-1993.During this research program there have been three principal stages in the development of the methodologies for data collection, analysis, and evaluation: (a) 1983-1985: 30 field studies; (b) 1986-1988: 28 tidies; and (c) 1989-1993: 20 studies. A RESEARCH PROGRAM TO ANALYZE AND EVALUATE SSP stage 1: 1983-1985 When this research and teaching program on ISRC was Initiated In 1983, there had been only one significant empirical study of SSP In Canada, Corporate Social Performance in Canada (the Royal Commission on Corporate Concentration [RCA], 1977).The situation in the United States was very much the same: "actual empirical research designed to test the multitude of definitions, propositions, concepts, and theories that have been advanced has been scarce" (Puerperal, Carroll, & Hatfield, 1985). To develop a methodology in 1975, the researchers in Canada had used the corporate social response matrix, which had been developed by Preston (1977), who writing of Corporate Social Performance in Canada.

The focus of Proton's matrix or framework was the management of social issues by corporations.It was assumed that managers followed stages of a process identified as corporate social involvement. The stages of this process were defined by Preston as follows: (a) awareness or recognition of an issue, (b) analysis and planning, (c) response in terms of policy development, and (d) implementation. This analytic framework was implemented using survey instruments and guidelines developed by Kelly and McGrath (1979). These materials provided the basis for the methodology that was developed for use in 1983.The nine companies selected initially for study at that time had been among those studied seven years earlier, thus providing the opportunity for noting changes and trends in performance.

(A more detailed description can be found in Clanks, 1988. ) Proton's (1975) framework, however, provided no definition of what was, or was not, a social issue; nor was there guidance or a corporation's managers or researchers in determining whether a social issue was one about which the company should become concerned and involved.In 1983, at the beginning of the research program, several human resource issues were identified as important enough for most corporations to regard them as issues to be managed: communications with employees; training and development; career- planning; retirement and termination counseling; layoffs, redundancies and plant closings; stress and mental health; absenteeism and turnover; health and safety; employment equity and discrimination; women in management; performance appraisal; day care. (Clanks, 1988: 52) Because these are all issues, the assumption was made that they are also social issues.Consequently, the next assumption was made: Corporations and their managers should be concerned about, and responsive to, these social issues if they were to be evaluated as socially responsible. As researchers, we had introduced, without explicit acknowledgment or understanding, a set of normative assumptions about how corporations should behave and how their performance should be evaluated.

Stage 2: 1986-1988 The development of Carol's (1979) original model represented an advance over Proton's (1975) framework and introduced a new conceptualization of SSP.Carroll was attempting to (a) reconcile the achievement of both corporate social and economic objectives, (b) to reconcile ISRC with ISRC, and (c) to focus on the most important element, SSP. Carol's model was both comprehensive and integrative. The strength of its influence can best be Judged by its longevity and that of its progeny. The model defined ISRC in terms of principles or categories and ISRC in terms of processes or strategies toward both social responsibilities and social issues. Social issues were defined by Carroll as consumerism, the environment, coordination, and so on, and were used as surrogates for actual performance.

It was plausible that corporations were expected to "do something" about these issues. But why they were expected to do something and what they were expected to do were not easily explained. Carol's model, in the form of a three-dimensional cube, was complex and difficult to test. It did not lend itself to the development of a methodology that could be used in the field to collect, organize, and evaluate corporate data. Warwick and Cochran (1985), building on Carol's integrative view of he management of social issues were separate, alternative corporate concerns.Their model, based on Carol's, recognized and incorporated economic performance as the first among the dimensions or elements of social responsibility, without excluding the other responsibilities defined by Carroll: legal, ethical, and discretionary.

Their model, again like Carol's, was an attempt to show that there is an underlying and continuous interaction between and among the principles of social responsibility, the processes of social responsiveness, and the policies and programs developed to address social issues. Models and frameworks are helpful for clarifying theories and abstract concepts or constructs.But to be useful in practice, a model or framework must be applicable to the conditions that it is attempting to describe, analyze, or predict. Empirical testing of a model is important to establish its validity. Whereas Proton's corporate social response matrix was limited to policies and programs responding to social issues, the Warwick and Cochran model, based on the Carroll construct, included the dimensions of corporate social responsibility and the processes of corporate social responsiveness.

By the end of the third year of field research, 30 studies had been completed using the initial methodology based on Proton's matrix.Changing methodologies is not done lightly, because data obtained previously must be reorganized to be useful. But because the Warwick and Cochran model appeared to be suitable for testing in the field and, in terms of the management of social issues, was compatible with Proton's approach, the decision was made to revise the methodology to use the new model for studies beginning in 1986. Details of the methodology developed for using and testing the model in the lied have been described elsewhere (Clanks, 1988).

Only the most important conceptual difficulties and problems are discussed in the second section.The principles of social responsibility. Under the heading of principles of social responsibility in the Warwick and Cochran model, the elements or dimensions of social responsibility are defined as economic, legal, ethical, and discretionary, following Carol's original classification. Consequently, the methodology developed for the field studies required that data be gathered on each of these four dimensions.

Obtaining economic data presented few problems, with annual reports ND data on industry profitability usually available.Being profitable for the preceding five years was established as the measure that a company had been fulfilling its economic responsibilities. Databases of the financial press were checked to provide data about litigation and allegations of illegal corporate behavior. Government departments, unions, and municipalities in company towns were also checked to discover data about environmental or safety problems. If no evidence was found, the assumption was made that the company was fulfilling its legal responsibilities. This, of course, was an easy test to pass.

Ethical responsibilities were more difficult to define and test.There are no generally accepted ethical principles that can be cited or enforced, as with accounting principles. The existence of a corporate code of conduct, practice, or ethics is certainly evidence that a company is aware of some responsibilities but does not tell the researcher how the code is being implemented or whether it is simply window dressing. Many company codes were primarily defensive, aimed at protecting the company and its property from its employees except in terms of the extent of the corporation's philanthropic activities and the tauter of its involvement in the communities in which it did business.As Carroll (1979) noted, "discretionary responsibilities of business are volitional or philanthropic in nature, and, as such, are also difficult to ascertain and evaluate. " Given the four dimensions of corporate social responsibility defined by the model, the corporate studies provided little empirical data to show that a company was not socially responsible, unless there was a history of unprofitably, coupled with evidence of illegal or unethical corporate behavior.

It developed that the model did not provide a distraction means by which the concept of social responsibility could be tested with reasonably accessible corporate data. Votary's (1973) criticism of the term corporate social responsibility remained valid: The term is a brilliant one; it means something, but not always the same thing, to everybody. To some it conveys the idea of legal responsibility or liability; to others it means socially responsible behavior in an ethical sense; to still others, the meaning transmitted is that of "responsible for", in a causal mode; many simply equate it with a charitable contribution.Voted, 1973: 1 1) The processes of social responsiveness.

The processes of social responsiveness were defined by both models in terms of corporate strategy or posture toward social issues. Carroll (1979) identified these processes of response as being reactive, defensive, accommodative, or proactive. Warwick and Cochrane model (1985) used the same four categories of social responsiveness.As Wood (1991 : 703) correctly observed: "These approaches may indeed characterize various organizational responses to social pressure, but they are not themselves processes. " Consequently, he research question for the field studies became one of determining the types of behavior that could serve as reliable indicators of, or surrogates for, these differing characterizations of corporate postures or strategies toward social responsiveness and social issues.In an attempt, therefore, to describe a corporation's social responsiveness and analyze its elements, the methodology that was developed in 1986 included the following descriptions of the data to be gathered: A corporation's statement of mission or purpose, its code of conduct or ethics, and the structure of TTS processes for managing issues in such areas as environmental scanning and analysis, the integration of social issues into policy and planning, and the internal linkages in a corporation whereby strategic decisions about social issues are integrated into operations by means of objective setting, performance appraisal and rewards; and the extent of public policy involvement, either directly or through trade associations.

(Clanks, 1988) When questions arose from student researchers and managers in the field, it became apparent that there was no logical explanation for he inclusion of statements about corporate mission or purpose, together with evidence of public policy involvement, under the heading of social responsiveness rather than under the management of social issues.It was not clear whether policies, programs, and performance data concerning codes of ethics, conduct, or practice should be included under the headings of ethical responsibilities or management of social issues. The fundamental problem was, and remains, that no definition of social responsiveness provides a framework for the systematic collection, organization, and analysis of corporate data. The term social responsiveness carries no clear meaning time, energy, and paper have been consumed in attempts to explain the term. But it remains an elusive construct, lacking both logic and rigor, which limits seriously its usefulness for empirical research.

Although the categories were confusing, the terms used by the model to describe a corporation's strategy or posture toward the management of issues were helpful in the field.Strategies, posture, and behavior that are reactive, defensive, accommodative, or proactive can be demonstrated by the presence or absence of policies and programs concerning relevant issues and by the reparation's performance in implementation. The following extract, from the field study of Canada's second largest bank in 1986, illustrates this point: A characterization of the company's attitude towards social responsiveness can be summed up by a couple of statements from interviews with the Bank's representatives. The Manager of Media Relations said of the Bank: "We are not a government, we are a bank. We do not set social policy, we look to government for social policy.

In another interview with a Vice President of Human Resources, it was said: "The government is into every nook and cranny of our business. " These tenements, and many others, indicate that the social orientation of the company, using the READ scale, is, at best, accommodative. (Vincent, Lowlier, & Starts, 1986: 6) Performance and nonperformance are concrete, measurable criteria. If an issue is being managed, there will be data.

The terms reactive, defensive, accommodative, and proactive have been incorporated into the READ scale, which was developed to evaluate corporate performance and is discussed in the second section. Stage 3: A New Framework is Developed From 1986-1988, researchers gathered case study data about 28 companies, using the new methodology.Data had now been collected from more than 50 corporations about policies, programs, and issues concerning the social and physical environments, public affairs and government relations, community relations and charitable donations, employee relations, and human resource management, as well as customer and shareholder relations. In short, the data that were being collected fit into categories that could be classified, as later became apparent, in terms of the management of a corporation's relationships with its stakeholder groups. The methodology, however, required that the data be organized to fit the Warwick and Cochran model, which was based on distinctions among the principles of corporate social responsibility, the processes of corporate social responsiveness, and the management of social issues.

These distinctions, which had intuitive appeal on the printed page, failed the test of practicality.Attempts were made to fit the data to the methodology, but finally it became clear that the categories of the model were not applicable to the data that were being gathered and that the classifications of the model were not grounded in the realities of corporate practice. As the volume of data ND the number of studies grew, it became increasingly difficult to achieve consistency in the collection and classification of these data to conform with the methodology. The model and, consequently, the methodology were at variance with the way in which corporations actually manage their relationships with employees, customers, shareholders, suppliers, governments, and the communities in which they operate.Although the term stakeholder management was not necessarily in use, it groups or constituencies, which could be defined as stakeholder groups, and that Hess relationships were either being managed, or not being managed, for better or worse. Whether these groups of customers, employees, shareholders, etc.

, were classified as internal or external stakeholders was irrelevant, Just as it was irrelevant for the companies themselves whether these groups were described as stakeholders at all. What was relevant to the research program was that the data that had been collected and analyzed corresponded with the concepts and models of stakeholder management (Freeman, 1984), rather than with the concepts and models of corporate social responsibilities, responsiveness, and performance.The data showed that, in the normal course of conducting their business, corporate managers do not think or act in terms of the concepts of corporate social responsibilities and responsiveness, nor of social issues and performance. The following statement from Corporate Social Performance in Canada illustrates this point and also provides an early example of the use of the term stakeholder issues. It is also worth pointing out that in many cases public affairs departments were not established to handle social responsibility issues as such but to help the organization respond more competently to a whole anger of "stakeholder issues," including the company's relationships with employees, media, and with government. RCA, 1977: 81) ISRC and ISRC are concepts that have been generated outside business.

They have normative connotations lacking clarity and specificity and have the disadvantage of sounding like Jargon. "Socially responsible to whom? ", "Socially responsive about what? ", "Social performance judged by whom and by what standards? ": These are legitimate questions to which business people have not received satisfactory or meaningful responses. Understandably, they have resisted attempts to make them responsible for social issues that they do not perceive as corporate or business issues. Managers are trained in the management of the processes of production, marketing, finance, accounting, and human resources.

Managers understand the meaning of responsibility in the context of these functional disciplines, and they understand the meaning of accountability for the results of their decisions. Obligations and responsibilities to customers, shareholders, employees, and other important constituencies are defined by most companies, together with corresponding accountabilities. Consequently, there are data concerning the management of report relationships with these constituencies or stakeholder groups. Managers do not find it difficult to understand the concepts and models of stakeholder management. They recognize that important issues of concern to groups of stakeholders may be identified as stakeholder issues as well as social issues.

For example, occupational health and safety or employment equity and discrimination are issues of sufficient concern to society as a whole to result in legislation and regulation, but they are also issues of concern for all corporations in terms of their relationships with employee stakeholder groups and government. Similarly, the social issues of product safety or truth in advertising have also led to legislation and regulation, but from a corporate perspective, these are stakeholder issues involving obligations and responsibilities to both customers and governments. Social issues concerning environmental pollution are of concern to a variety of government operations, employees, and customers.Research Design and Data Collection From its beginning in 1983, the design of this research has been influenced by several factors. MBA students at the University of Torso's Faculty of Management revived most of the necessary research time and effort, studying individual companies in groups of two or three and writing the case studies as their term project for a second-year elective course on corporate social responsibilities (Clanks, 1988, 1991). To describe and evaluate a company's performance, the researchers had to gain the confidence of the relevant managers so as to be able to ask the right questions and obtain written material about policies and programs.

Both researchers and managers needed a framework and guide to facilitate the provision, analysis, and evaluation of data. It was essential for such a framework and guide to be expressed in terms that would be understood in a corporation as well as in a classroom. Proceeding from the conclusion that a "stakeholder management" model provided the most appropriate organizing principle, an inventory of representative stakeholder issues was developed from the data contained in the field studies. This inventory, or index, of approximately 50 issues is shown in Table 1 . This index is described as "representative" because it lists the issues identified most frequently in the studies. It is reasonably comprehensive, but not exhaustive.

It can revere as a stimulus to some managers to consider a wider range of stakeholder issues than has been their practice. This index provides a uniform entry and coding system and is central to the organization of the data in each study for the computerized database. Information pertinent to each of the stakeholder issues is organized into four areas: description, performance data, evaluation, and analysis. To facilitate data collection and comparisons, it was necessary to define clearly the issues identified in Table 1 .

It was also important to define the performance data that were being requested from the companies being studied. This guide for researchers and managers is illustrated in the appendix.Clanks described the data as follows: The corporations are asked to provide the descriptive data covering the company itself and relevant stakeholder and social issues. This material is then edited and returned to the company with requests for the performance data identified in the guide. Interviews with appropriate executives are then held in order to check and explore the implications of the performance data that have, and have not, been supplied Experience shows that corporations find this task worthwhile. Few have hitherto identified stakeholder and social issues so comprehensively.

(1991 : 344) Sixty-five of the more than 70 corporations that have been studied are among the largest 250 companies in Canada, in terms of sales or assets, or are subsidiaries of companies listed in the Fortune 500.Ten of the 14 largest financial institutions in Canada have been studied, as well as the two largest transportation companies, the two largest steel companies, the three largest publishing companies, the three largest breweries, the largest electric and gas utilities, and the largest nickel, auto arts, pulp and paper, and telecommunications companies, together with four of the five largest integrated oil companies and six large retail companies. The universe of companies studied is large and diverse, containing companies with various forms of ownership: Canadian, U. S. And foreign, public and private. Most companies in the OF CONCLUSIONS FROM THE RESEARCH The principal conclusions drawn from the research program are as follows: It is necessary to distinguish between stakeholder issues and social issues because corporations and their managers manage relationships with their stakeholders and not with society.

It is necessary to conduct analysis at the appropriate level: institutional, organizational, or individual.It is then possible to analyze and evaluate both the social performance of a corporation and the performance of its managers in managing the corporation's responsibilities to, and relationships with, its stakeholders. Distinguishing Between Social Issues and Stakeholder Issues A multitude of issues have been described as social issues in the SSP literature. Under the rubric of the Social Issues in Management division of the Academy of Management, an extraordinarily wide range of subjects pertaining to business and society is discussed at conferences and written about in Journals.

It has become difficult, if not impossible, to define what is, or what is not, a social issue. The difficulties that have been encountered in defining ISRC, ISRC, and SSP can be attributed in part to the broad and inclusive meaning of the word social.The connotation of social is society, a level of analysis that is both more inclusive, more ambiguous, and further up the ladder of abstraction than a corporation itself. Preston noted that corporate social performance was intended to suggest a broad concern with the impact of business behavior on society. The concern is with ultimate outcomes or results, not simply with policies or intentions; moreover there is some implication that these outcomes are to be evaluated, not simply described.

(1988: xii) There has been general agreement with this definition of SSP and the objective, but the underlying assumptions have not been questioned rigorously.It has been assumed that, because there is a "broad concern," it would therefore be possible to evaluate the impact of business on society. The impact of a business or corporation on society is a different matter from the impact of business in general on society as a whole. Wood (1991 : 691) observed that "the concept of corporate social performance has received serious theoretical and empirical attention, .

But the concept's theoretical framework and impact have not moved significantly beyond Warwick and Cochrane (1985) articulation. " The principal reason for this failure has been the lack of clarity about the appropriate level of analysis.This failure, together with the confusion and misunderstanding about the definition and meaning of corporate social responsibility, corporate social responsiveness, and corporate social performance, is a direct result of the inclusive and vague meaning of the word social. Friedman (1970) took advantage of this ambiguity and semantic confusion in his criticism of "those who speak eloquently about the 'social responsibilities of business' in a free-enterprise system. " He continued: The discussions of the "social responsibilities of business" are notable for their analytical looseness and lack of rigor The first step towards clarity in examining the doctrine of the social responsibility of business is to ask precisely what it implies for whom.Friedman chose to interpret social issues and social responsibilities to mean enviousness issues and enviousness responsibilities.

He, like so many neoclassical economists, separated business from society, which enabled him to maintain that "the business separate compartments, Friedman (1970) was able to deny the necessity, or even the validity, of the concept of CAR, decrying it as "a fundamentally subversive doctrine": [Businessmen who believe that] business has a "social conscience" and takes seriously its responsibilities for providing employment, eliminating discrimination, avoiding pollution are preaching pure and unadulterated socialism.The move from the innocuousness of social to the taint of socialism was made skillfully by this aster of rhetoric. Neither business in general nor specific corporations in particular can properly be made responsible for dealing with all social issues. Before responsibilities can be assigned and before corporations and their managers can be held accountable for the results of their actions, it is necessary to develop a systematic method of determining what is and what is not a social issue for a corporation.

From the data in the field studies of corporate performance, an inventory of issues was developed. These issues have been identified as typical stakeholder issues rather than as typical social issues.The reason for this distinction is that all these issues are of concern to one or more stakeholder groups, although these issues are not necessarily of concern to society as a whole. The positions being advanced here are: A particular society (municipal, state, or national) determines, usually over an extended period of time, what is a social issue, and, when it is considered necessary, the relevant polity enacts legislation and regulation.

When there is no such legislation or regulation, an issue may be a stakeholder issue, but it is not necessarily a social issue. A test of whether an issue has become a social issue s the presence or absence of legislation or regulation. In Table 1, 20 different issues are shown under the stakeholder heading of employees.Several, but by no means all, of these issues have been of sufficient concern to society as a whole, in the United States and Canada, that legislation and regulations have been enacted. Occupational health and safety and employment equity and discrimination are such social issues. (It is interesting to note in this context that some opposition to the North American free trade agreement [NONFAT] appears to have occurred because these are not social issues in Mexico.

No legislation has yet been enacted concerning the majority of the employee issues, such as employee assistance programs and career planning. But each can be identified as a stakeholder issue, when the level of analysis is the corporation itself.Defining the appropriate level of analysis is important, as Wood (1991 : 695) has shown: Once these three levels of analysis are distinguished (institutional, organizational, and individual) then several formerly competing concepts can be melded together to explain three corresponding principles of corporate social responsibility. Using the same levels of analysis--institutional, organizational, and individual--Table 2 proposes a framework that is different from Wood's and is grounded in the data of the corporate case studies.

The level of business and society is shown as the institutional level, the level that is appropriate for discussions of ISRC and ISRC. The organizational level is identified as that of the corporation and its stakeholder groups, the level appropriate for analysis and evaluation of SSP.The individual level is shown as that of managers who manage stakeholder issues and relationships with stakeholders, the level appropriate for analyzing and evaluating management performance. Confusion arises when terms issues of employee assistance plans and career planning are social issues is a question that should properly be discussed and answered at the level of society.

Corporate managers certainly should be cognizant of such discussions and concerns in society, but the position being advanced here is that a particular society and its polity determine what is a social issue, and, when it is considered necessary, legislation and regulations are enacted.By applying this analytic approach, it becomes evident that managers of a corporation cannot be expected to accept the lain that they have a social responsibility to institute an employee assistance plan or career planning or to provide day care, although an interesting discussion could take place about whether they have any responsibility to their stakeholders to implement such programs. Employee assistance plans, career planning, and day care are stakeholder issues at the corporate level of analysis and management issues at the level of stakeholder issues and relationships. It is the responsibility of the corporation's managers to determine whether policies and programs will be implemented to manage these issues. Whether these are social issues is not relevant in this context.

This approach makes it clear that all social issues are not necessarily stakeholder issues, Just as all stakeholder issues are not necessarily social issues. A company and its management are free to decide the extent to which they will acknowledge, recognize, or pursue obligations and responsibilities to their stakeholders concerning the issues shown in Table 1, and, of course, any additional issues identified by the corporation or its stakeholders. Their performance can then be evaluated in terms of the READ Scale as reactive, defensive, accommodative, or reactive. Clearly, there are legal requirements regarding certain social issues, as defined previously.

Social issues, such as occupational health and safety, shareholder rights, and product safety, have generated significant regulation, but there are no legal requirements for a company to assume any responsibilities to its employees for training and development or career planning, or to its customers for communications and complaints. An outside observer, a financial analyst, or an academic researcher might consider such programs to be socially desirable or socially responsible on the art of a corporation, but these are in fact matters of policy and choice for each corporation to decide. Such corporate decisions are usually made on the basis of market forces, for example, employee productivity or customer satisfaction, not necessarily because they are socially desirable. Managers are interested in results, first and foremost. Performance is what counts.

Performance can be measured and evaluated.Whether a corporation and its management are motivated by enlightened self-interest, common sense, or high standards of ethical behavior cannot be determined by the empirical methodologies available today. These are not questions that can be answered by economists, sociologists, psychologists, or any other kind of social scientist. They are interesting questions, but they are not relevant when it comes to evaluating a company's performance in managing its relationships with its stakeholder groups. Defining Stakeholders and Stakeholder Groups The definitions of stakeholders and primary and secondary stakeholders that are proposed here are straightforward.

Freeman's (1984) landmark work provided a solid and lasting foundation for many continuing efforts to define and to build stakeholder