Sep 22, 2011 - In the last decade, a series of major business failures triggered a ... letter of the law rather than with the spirit, or engaging in ... the Harvard Law Review (HLR) (2003) suggested precisely this ...... Listed Company Manual.
1
The cut and paste society: Isomorphism in Codes of Ethics
BY
LORI HOLDER-WEBB ASSOCIATE PROFESSOR WESTERN NEW ENGLAND UNIVERSITY JEFFREY COHEN PROFESSOR, ERNST AND YOUNG RESEARCH FELLOW BOSTON COLLEGE
September 22, 2011 Forthcoming in Journal of Business Ethics
We would like to thank the section editor Muel Kaptein, two anonymous reviewers, Yves Gendron, Theresa Hammond, B.K. Cohen, David Sharp and Greg Trompeter for helpful comments and suggestions. An earlier version of this paper was presented at UC Dublin, Boston College, Bentley University, American Accounting Association Annual Meeting and the Third Alternative Perspectives in Accounting Research Conference at University of Laval, Quebec.
Electronic copy available at: http://ssrn.com/abstract=1932442
2
THE CUT AND PASTE SOCIETY: ISOMORPHISM IN CODES OF ETHICS Abstract: Regulatory responses to the business failures of 1998-2001 framed them as a general failure of governance and ethics rather than as firm-specific problems. Among the regulatory responses are Section 406 of Sarbanes-Oxley Act, SEC, and exchange requirements to provide a Code of Ethics. However, institutional pressures surrounding this regulation suggest the potential for symbolic responses and decoupling of response from organizational action. In this paper, we examine Codes of Ethics for a stratified sample of 75 U.S. firms across five distinct industries and find that content and language converge across organizations in ways undesired by the regulators, and that language is used to minimize the effects of the Code on constraining organizational behavior. There is, however, a noteworthy exception in the sections of the Codes dedicated the ethics of financial reporting. Although this material still contains legalistic boilerplate information it does offer concrete guidance and emphatic language pertaining to the need to maintain the integrity of reporting practices. This suggests that the corporate understanding of the source of the failures is one of fraudulent financial reporting. Aside from the matter of financial reporting, the vague and stylized content of the Codes was a predicted response and constitutes a rational response to the regulation. The regulation, however, clearly states the belief that Codes should vary from firm to firm and that individual firms should determine the specific content of a Code. Aside from financial reporting matters, the observed result suggests that regulatory efforts may have failed to instigate corporate change in attitudes towards and enforcement of higher ethical standards by corporate actors.
Key words: Code of ethics; Sarbanes-Oxley; Financial Reporting; Corporate Governance
JEL: M14, M48, D02
Electronic copy available at: http://ssrn.com/abstract=1932442
3
THE CUT AND PASTE SOCIETY: ISOMORPHISM IN CODES OF ETHICS 1. INTRODUCTION In the last decade, a series of major business failures triggered a legitimacy crisis for several American economic institutions: the markets, the regulation of business reporting, and those tasked with the societal obligation to protect the public interest – auditors, regulators, and the legislative bodies (Bealing and Baker, 2006). Assertive action to re-establish the legitimacy of these institutions was supplied in the form of the Sarbanes-Oxley Act of 2002 (SOX), an act intended to enhance the credibility of financial reporting provided by any company publicly traded in the U.S. (Bealing and Baker, 2006). Legislators and regulators chose to frame the business failures as ethical and governance problems; therefore the Act includes numerous provisions intended to increase governance quality and to constrain managers and organizations from unethical conduct. In this latter category is included Section 406, which requires publicly-traded firms to disclose whether they have a Code of Ethics (or Conduct) or to discuss why that disclosure is not provided. The implication of this provision is that having a Code of Ethics (a Code) is assumed likely to constrain unethical behavior (Krawiec, 2005). The assertive and public framing of the legitimacy crisis in ethical terms gave rise to an increase in ethicsoriented rhetoric in the public arena (Krawiec, 2003). In this study, we use the term “rhetoric” to refer to the public espousal of values that are considered necessary in order to be seen as a legitimate leader, consistent with Kallio (2007).The battle cry for “better ethics” reaches from the boardroom to the classroom, and suggestions have been made that it is now functionally a taboo to call these demands into question (Kallio, 2007). In this paper, we suggest that the Codes provided in response to Section 406 are consistent with symbolic responses to the legitimacy-driven regulation that may be unlikely to supply significant useful ethical guidance or to constrain organizational behavior – despite the voiced intentions of the regulatory bodies that require the provision of the Codes (Bealing and Baker 2006). We also suggest that this response was both predictable and predicted and question the usefulness of regulatory actions that are more likely to stimulate ritualistic responses rather than to induce substantive change. The SOX Code requirement triggered an avalanche of regulation as the Securities and Exchange Commission (SEC) and regulators of the major U.S. exchanges jumped onto the legitimation bandwagon by issuing additional restrictive policies pertaining to the disclosure of Codes of Ethics. 1 SOX Section 406 requires companies to disclose publicly a code of ethics or to explain why they are not disclosing a code. The New York Stock Exchange (NYSE 2004) and the National Association of Securities Dealers Automated Quotations (NASDAQ 2003) both subsequently issued regulations mandating that companies listed on these exchanges must have and provide to the public a code of ethics starting in fiscal year 2004. The rhetorical line discernable throughout the public discourse and the ensuing regulation is that the business failures were caused by inadequate attention to ethical matters and governance failures; the implication of the regulation is that having a Code of Ethics will act as a substantive deterrent to unethical behavior on the part of managers and employees (Verschoor, 2002, NYSE, 2004). Academics joined in the rhetorical discourse with numerous articles dedicated to establishing and maximizing the effectiveness of Codes of Ethics (Stevens 2008). Section 406 of Sarbanes-Oxley speaks directly to the subject of original content in a Code of Ethics in B.2.c.:
4
“We continue to believe that ethics codes do, and should, vary from company to company and that decisions as to the specific provisions of the code, compliance procedures and disciplinary measures for ethical breaches are best left to the company…We strongly encourage companies to adopt codes that are broader and more comprehensive than necessary to meet the new disclosure requirements.” Regulatory intent notwithstanding, prior research finds a lack of firm- or industry-specific content in Codes (Murphy 2005). This suggests a divide between the widely espoused rhetoric in support of “better” ethics and “more effective” Codes, and the actual practice of drafting and deploying a Code. The effectiveness of a code could be a function of how strongly management is committed to the code and how the document is emphasized in the training and socialization of employees. For example, Stevens (2008) notes that Codes can be effective if integrated into the corporate culture where it is seamlessly integrated within the tone of the top (Cohen and Hanno 2000). However, it is also possible that managers neither intend nor desire to promote substantive organizational or behavioral change through the development and promulgation of a Code of Ethics (Cohen et al. 2008). For example, the Hampel Report (Hampel 1998) suggests a typical managerial response to regulatory actions: complying with the letter of the law rather than with the spirit, or engaging in “box-ticking” behaviors where it is regarded as more important to produce a compliant disclosure than it is to produce an informative one. For example, Cohen et al. (2008) argue that although audit committees comply with all regulations concerning independence, in fact they may lack the will to be effective monitors over management. In the period between the passage of Sarbanes-Oxley and the implementation period for its provisions, the Harvard Law Review (HLR) (2003) suggested precisely this response, arguing also that the expected convergence upon the rules would create a disincentive to managers who wished to provide a firm-specific Code incorporating unique substantive guidance. While it is possible that the regulators also anticipated the potential for a box-ticking response, it is unlikely that they anticipated that this would be a widespread response given that their primary stated goal in issuing the regulations was to reinforce the legitimacy of the markets and the corporate structure by instigating visible improvements in the ethical climate of the corporations.2 To be effective in restoring legitimacy to the various institutions involved, the regulation needed to convince stakeholders both that the regulators had accurately addressed the cause of the failures and that the companies had altered their practices accordingly (in such a way that further failures were unlikely). The effectiveness of the regulation can thus be seen as a function of the degree to which managers comply with the spirit (drafting meaningful Codes that are substantively designed to improve the particular firm’s ethical behavior) versus ticking the box (providing a Code tailored to satisfying the regulations only and are there for “form” only). 3 To the extent that companies provide Codes that are nebulous, unlikely to constrain behavior, fail to take into account the potential for industry- or firm-specific ethical matters and functionally converge upon the “boxes” provided by the regulators is thus suggestive of failed regulation – that which accomplishes only superficial or cosmetic alterations, neither encouraging positive change nor discouraging the undesirable behavior considered to be at fault for the rash of business failures. Thus, codes may be just a sop to a public outcry over ethical failures but in fact the codes are so generic in nature so that in effect they become a meaningless or at best ineffective tool for promoting more ethical
5
behavior within a company. In essence, “form” triumphs over “substance” of this potential governance mechanism (Cohen et al. 2008). In this paper, we employ institutional theory to help us evaluate Codes of Ethics from a size- and industrystratified sample of publicly-traded U.S. firms in order to determine to what degree firm- or industry-specific ethical matters are addressed as well as to assess the prevalence of box-ticking behaviors predicted by HLR (2003). Institutional theory provides an explanation for the observed convergence in the content of Codes of Ethics that appear to have been drafted for form rather than for substance. 4 We provide empirical evidence on the convergence in content and language that is consistent with a response to institutional pressures and conclude with a discussion of the implications of the study for regulation and for academic research. 2. THEORETICAL DEVELOPMENT AND RESEARCH QUESTIONS Legitimacy is a key element in the organization’s access to necessary resources (Elsbach 1994; Meyer and Rowan 1977; Pfeffer and Salancik, 1978) and is conferred upon organizations that are effective in demonstrating dedication to social norms, values, or beliefs (Handelman and Arnold 1999; Suchman 1995). Actions that demonstrate the organizations’ support for these norms or acts of cultural allegiance are rewarded with access to financial or human capital (Elsbach 1994) as well as other forms of stakeholder support. Legitimacy, therefore, represents a strategic element necessary for the success of the organization. Isomorphism and the Code of Ethics DiMaggio and Powell (1983) develop a theory whereby entire organizational fields may obtain legitimacy by engaging in isomorphic (convergent) behavior. 5 They define organizational fields as recognized elements in institutional life, such as suppliers, consumers, and regulators, and state that organizational fields are those that feature both “connectedness” and “structural equivalence.” DiMaggio (1983) further identifies a set of criteria by which institutional definition occurs: increasing interaction between organizations within the field, sharp or defined patterns of domination and/or coalition between members of the field, increasing information load, and a mutual awareness among participants in the field that they are engaged in a common enterprise. The population of publiclytraded firms in the U.S. exhibits all of these characteristics, from increasing ties between firms through mergers and acquisitions to increasing competition for financial resources in the equity markets, but of particular interest for the current study is the increasing information load (through expanded information channels, growth in the volume and variety of media, and through increasing regulatory disclosure burdens). This is also confirmed by the arguments of Greenwood and Hinings (1996 1026) who state, “The focus of neo-institutional theory is thus not upon the individual organization but upon a category or network of organizations.” Therefore, for the purposes of this study, we define the population of publicly-traded firms in the U.S. as an organizational field. DiMaggio and Powell (1991 8) argue that the new institutionalism in organization and theory and sociology is “a rejection of rational-actor models” which for this study implies that to understand the potential effectiveness of codes we must go beyond the so called objective analysis suggested by traditional agency theory (Cohen and Holder-Webb 2006). Within organizational fields, DiMaggio and Powell (1991) suggest, there exists a process of institutional isomorphism that leads over time to a convergence, and ultimately, to a homogeneity in behaviors among organizations within the field. As Kostova et al. (2008 997) argue, “The neoinstitutional model essentially
6
holds that organizational survival is determined by the extent of alignment with the institutional environment; hence organizations have to comply with external institutional pressures.” As suggested by Meyer and Rowan (1977), this isomorphism has as a goal the function of conferring power, legitimacy, and access to resources for the isomorphic organization. DiMaggio and Powell (1983, 1991) identify three pressures that lead to isomorphic behavior: coercive pressures (formal or informal pressures on the organization provided through regulation, mandate, or focused cultural expectations), normative pressures (collective inculcation of the belief in a professional field through education that a particular convergent behavior is desirable), and mimetic pressures (uncertainty in the environment combines with the presence of a visible first-mover that provides an action template for other organizations to follow).6 Consistent with the ceremonial and ritualistic arguments advanced by Meyer and Rowan (1977), isomorphic actions are undertaken with the primary goal of attaining legitimacy rather than with the primary goal of enhancing organizational efficiency or control. To the degree that these actions are perceived as being rational and effective or convergent with other legitimate institutions – or to the extent that compliance with legal mandates confers a type of legitimacy in itself – they are likely to be effective in gaining legitimacy. 7 To the extent that the legitimacy crisis played out on multiple organizational levels, including the level of the individual firm needing to obtain capital from market participants at the lowest possible cost, the use of isomorphic content that visibly complies with the content of the regulatory mandate is a rational response on the part of the individual company. However, the legitimacy of the market and economic processes as a whole (rather than the legitimacy of any particular constituent firm) was the primary focus of the regulatory effort; in that context, it is less clear that widespread isomorphic and symbolic behavior was likely to confer the legitimacy sought by regulators. 8 The passage of SOX represented a seismic shift with a regulatory response towards codes in that it required publicly traded corporations to provide a code of ethics (or to explain why they would not provide one). SOX defines a Code as “such standards as are reasonably necessary to promote – (1) honest and ethical conduct, including the ethical handling of actual or apparent conflicts of interest between personal and professional relationships; (2) full, fair, accurate, timely, and understandable disclosure in the periodic reports required to be filed by the issuer; and (3) compliance with applicable governmental rules and regulations.”9 The requirement for a corporation to follow the laws yields no uncertainty (and arguably should not require explicitly stated organizational commitment); however, items (1) and (2) above yield considerable uncertainty. For example, Sims and Brinkman (2003), in an evaluation of the massive fraud at Enron, concluded that the firm’s culture and maniacal drive to be powerful swamped any potential benefits that may have accrued from its code of ethics. Ethical conduct is a construct that persistently defies definition by philosophers and educational institutions (Becker and Becker 2001), while defining full, fair, and understandable disclosure has been a matter of major regulatory concern by the SEC and accounting regulators for nearly a century. Thus the effect of Section 406 is to create both coercive pressures (arising from the passage of the regulation) and mimetic pressures (as Codes were required but significant uncertainty existed as to what they should contain and how much they should constrain behavior). 10 As noted above, the Sarbanes-Oxley Act specifically states that each company should develop their own Code, and that because of the unique ethical
7
challenges that face any given firm, there should be considerable variance between the content of the Codes across firms. The SEC promptly elaborated upon SOX 406 by expanding the rule to require disclosure of the internal reporting of violations of the Code to designated parties and accountability for adherence to the Code (SEC 2003). However, the SEC did not dictate the content, language, procedures, or other details of an acceptable Code, instead encouraging each organization to determine the specifics of its Code (Harvard Law Review [HLR] 2003). The New York Stock Exchange (NYSE 2004) and the National Association of Securities Dealers Automated Quotations (NASDAQ 2003) both subsequently issued regulations mandating that companies listed on these exchanges must have and provide to the public a Code of Ethics starting in fiscal year 2004. Verschoor (2002) indicates that the intent of the NYSE rules is that codes should be value-based, not simple or legalistic prescriptions of prohibited acts. The exchange regulators also elaborated upon the SEC action by providing a list of minimum coverage for a Code (discussed below). The SEC echoes the expectation of Sarbanes-Oxley that Codes will vary from one firm to another. Contemporaneous public information releases suggest that this intent was well understood among the professional class likely to be consulted for guidance on the Code content. Numerous professional (accounting, consulting, and law) firms provided implementation guidance pertaining to the new regulations to provide a Code at the time the rules were passed. It is clear from this guidance that the general understanding among this professional community was that the SEC intended that the Codes did and should vary from firm to firm and that the specific provisions should be left up to each firm to establish for itself (Crow 2003; Deloitte & Touche 2002; Haynes and Boone 2003; Holmes Roberts & Owen 2003; Locke Liddell & Sapp 2003). Despite this general understanding, there was contemporaneous speculation that the intent of the regulators would be contravened. HLR (2003) suggests that in the ideal world, as investors become more accustomed to seeing Codes, they (the investors) will become more skilled at identifying firms with effective Codes and will create market pressures that induce firms to provide more substantive Codes. Compliance-based deterrents (Krawiec 2003) and isomorphic forces, however, suggest a “window-dressing” response that offers legitimacy with the market and the possibility of reduced exposure to legal liability without actually deterring wrong-doing. Regulation requires that a Code must be provided, and provides guidance but does not mandate specific content. Coercive pressures exist that lead to the provision of a Code and may also influence the content of the Codes; mimetic forces arising from uncertainty over implementation (content) of the Codes may also lead to convergence both in content and in the language that is used to convey the content. However, when the regulatory bodies conveyed guidance on the minimum required content, they did so with the expectation that this would provide a basis for drafting Codes rather than an exhaustive and inclusive list of what must be included (SarbanesOxley Act 2002; Verschoor 2002). The regulators clearly indicated that the preference is for compliance with the spirit, rather than with the letter, of the rule and dictated that Codes should be informed with insider understanding of specific organizational ethical risks (for example, conflicts of interest between pharmaceuticals and physicians, fiduciary duties for financial professionals, environmental protection issues for extractive industries, animal testing issues for pharmaceuticals, employee safety for manufacturing industries etc.).
8
Nevertheless, Edelman et al. (1999) suggest that even the box-ticking may ultimately confer legitimacy to the extent that convergence with the stated regulatory guidelines becomes a legally-supportable managerial response. Edelman et al. (1999) also indicate that if the uniform response is perceived as legally supportable, it becomes a matter of efficiency: if market participants consider the guidance to be authoritative in nature, they will expect Codes of ethics to comply with the suggested content whether or not that content is appropriate and/or applicable to any given individual organization. Convergence thus promotes efficiency; the acquisition of legal authority combined with the distaste for authentic constraints on behavior and decoupling in the face of institutional pressures (Meyer and Rowan 1977) implies that Codes potentially may incorporate not only Hempel’s box-ticking but the vague and nebulous language suggested by HLR (2003).11 This suggests that despite regulatory preferences and the availability of likely firm- and industry-specific ethical quandaries that could effectively be addressed through a Code of Ethics, Code content is likely to converge on the minimum standards promulgated by the regulators and is unlikely to differ substantively across organizations. 12 This is examined in Research Question 1: RESEARCH QUESTION 1: Does the content of Codes of Ethics exhibits convergence across organizations consistent with isomorphic pressures? Ambiguity in the law leads to a state of uncertainty that can foster mimetic pressures (DiMaggio and Powell 1983) as well as providing space for managerial construction of law and of the meaning of compliance with laws and rules (Edelman et al. 1999). It is thus associated with a convergence in behavior that takes on authority as the convergent behavior becomes seen as sufficiently compliant with the rules. Section 406 of the Sarbanes-Oxley Act creates an air of uncertainty with respect to activity (firms must provide a Code or a reason why they don’t have one) and with respect to goals: given that firms such as Enron possessed an extensive Code and still exhibited catastrophic ethical failures (Wee 2002). This environment is also characterized by constraining pressures: firms are obliged to deliver Codes, but they are also operating in a highly litigious environment (in general) - a litigation environment now itself characterized by an emphatic focus on ethical failures. 13 The expressed preferences of the regulators were for firms to provide Codes that reflect organizational consideration of ethical matters (Sarbanes-Oxley Act 2002; SEC 2003; Verschoor 2002). However, DiMaggio and Powell (1983) suggest that when considerable goal uncertainty exists, firms will model themselves on other firms that they perceive to be successful. In light of the significant litigation-based constraints faced by U.S. firms, “successfulness” in this context must be defined not in terms of how “ethical” the corporation has become, but how well the corporation has been able to avoid lawsuits and minimize litigation-related costs. HLR (2003), in reviewing the usefulness of regulation pertaining to Codes of Ethics, notes that strong compliance programs may produce “incriminating information” that could lead to civil or criminal liability. They conclude that significant incentives exist to draft Codes that are full of vague or nebulous provisions, couched in legalese language, that are difficult to understand and impossible to enforce. This notion is consistent with Meyer and Rowan (1977), who indicate that the use of externally legitimated accounts enables management to argue rationally that failures did not occur because of negligence; adherence to legitimated myths provides evidence that management behaved in accordance with the definitions of prudence, rationality, and due care. Legal risk, as a function of language as much as of intent, suggests the rise of a uniform
9
format for Codes of Ethics that is designed to minimize litigation risk while still complying with the regulations. This prediction is also implied by the work of Edelman et al. (1999) with respect to their discussion on how managerial response to regulatory action assumes the force of law through judicial interpretation (i.e., the “everyone is doing it this way” argument). Further, in a study of a convenience sample of seven firms located in the “Atlantic Canada” area, Long and Driscoll (2008) found that there was evidence of convergence within their sample that they attribute to isomorphism. Moreover, Long and Driscoll (2008 186) argue that their results “suggest the possibility that the strategic self-interest embodied within codes of ethics is gaining cognitive legitimacy.” We are extending this issue to a US context illustrating the types of isomorphism which is at work. This gives rise to Research Question 2: RESEARCH QUESTION 2: Does the language of Codes of Ethics exhibits convergence across organizations consistent with isomorphic pressures? However, Meyer and Rowan (1977) note several difficulties with the type of convergence implied by isomorphic forces and the managerialization of law. First, responses driven by broad institutional pressures may differ from those built on efficiency, and efforts to pursue both simultaneously may lead to conflict within the organization (situations where Code-based constraints on behavior conflict with the mandate to maximize shareholder wealth). Second, actions driven by forces without the boundaries of the firm may create conflict between institutionalized rules (such as the pressure to provide a substantive Code competing with the pressure to minimize litigation risk). Third, institutionalized rules (such as the mandate to provide a Code) are produced with a very high level of generality, whereas organization-specific activities take place within unique conditions. In all situations, institutionalized rules must be reconciled with potentially incompatible technical activities. Meyer and Rowan (1977) describe a method for resolving these conflicts through decoupling structure from activity by reducing formal evaluation, inspection and control of activities and reassigning coordination to the informal organizational structure. This may be done through couching goals in ambiguous terms: when goals are not measurable or otherwise subject to evaluation, inspection and evaluation procedures are rendered ceremonial. This theoretical perspective is consistent with the concerns articulated by HLR (2003) in respect to the content of Codes. Thus, organizations that pursue an isomorphic strategy have incentives to discourage direct evaluation of organizational practices. They prefer instead to take refuge in the assertion that adherence to environmentally legitimated forms constitutes a necessary degree of care, that they are behaving in a proper and adequate manner and that this behavior protects the organization from questions about its legitimacy (that is, formal evaluation procedures become unnecessary). Rules are couched in deliberately vague language, and explicit goals are not included. HLR (2003) address this from a pragmatic legal standpoint, and suggest that despite the expressed preferences of the regulators, managers will create vague documents loaded with nebulous, legalistic language that contravene the intentions of the authors of SOX and the regulators mandating disclosure of Codes. HLR (2003, 2140) predict that “[c]ompanies will include only the bare minimum needed to comply with the SEC’s suggested topics for a Code, and the public filing of Codes will not matter because investors will be unable to distinguish one vague, boilerplate Code from another”.
10
While the theory and pragmatic commentary suggest that the Codes will be constructed so as to minimize the constraint placed upon the organization’s activities, it is difficult to construct a comprehensive theoretical rubric for evaluation of “vagueness”. We therefore focus on whether the Codes appear to provide a good basis for enforcement in our final Research Question: RESEARCH QUESTION 3: Do Codes of Ethics substitute decoupling language for concrete language leading to enforceable provisions? Thus, this study suggests that the content of Codes of ethics is convergent across organizations and that the convergence is on vague, nebulous, and legalistic language common to many or all participants in the organizational field. The use of this boilerplate is presumed to confer a degree of legitimacy on organizational participants that stems from its widespread use and thus constitutes a reduction in a firm’s risk of being accused of negligent or imprudent conduct. In keeping with established organizational theory and findings (DiMaggio and Powell 1983; Elsbach 1994; Krawiec 2003; Meyer and Rowan 1977; and Suchman 1995) this is of benefit in that it permits the firm to access necessary resources through the enhanced stakeholder confidence in the legitimacy of the organization. The next section describes the sample and the research method, after which the empirical findings are discussed. 3. DATA The Sample Firms This study employs a stratified random sampling design based on size and industry in order to gain an understanding of a cross-section of the population of publicly-traded firms in the U.S. and to permit analysis of Code content within industries. Stratified random sampling permits a greater amount of efficiency in statistical tests and a higher degree of confidence in results of tests than is available with comparable sample sizes under the simple random sampling technique frequently employed in archival business research (Johnson and Bhattacharya 2006; Schaeffer et al. 1990; ). It is the preferred method to use when there is reason to believe that sub-groups of the overall population are likely to exhibit considerable variability on the characteristics of interest and/or researchers wish to compare sub-groups on the characteristic of interest. In addition, because a stratified random sampling technique ensures that the sample will represent the overall population rather than risking over-sampling from some sub-groups of the population while other sub-groups are missed entirely, it permits the use of much smaller sample sizes to attain a given degree of statistical efficiency and confidence than would be required under a simple random sampling technique (Johnson and Bhattacharya 2006; Scheaffer et al. 1990). The temporal sampling frame is 2004, which was the year that the regulatory requirements for all publiclytraded firms to provide a Code became effective. The industry sampling frame was limited to publicly-traded U.S. firms listed on Compustat, excluding those engaged primarily in providing financial services, investment funds, and trusts.14 Based on the general SIC and NAICS categories represented by these firms, the largest general industry sector is manufacturing (39%), followed by firms engaged in production of intellectual property, including software and publishing (13%), those working with extractive natural resources (12%), and those engaged primarily in sales, including wholesalers and retailers (11%). No other general industry sector yielded more than 5% of the sampling frame during 2004. Due to the significant share of the market represented by manufacturing enterprises, the category was subdivided into those engaged primarily in straightforward manufacturing and those engaged in
11
manufacturing with significant research and development activities. This process yielded five industry sectors: manufacturing (without significant R&D), manufacturing (R&D intensive), extractive natural resources, intellectual property generation, and sales. Each of these categories was further divided into homogenous groups defined by 4-digit SIC codes. The largest group of similar industries within each major industry group was chosen as the representative sample (thus permitting random sampling from within size strata). This process yielded the five industries examined in this paper: manufacturers of surgical equipment; (MFG), pharmaceutical preparation and manufacturing firms (PHARMA); crude petroleum and natural gas extractors (OIL); software publishers (SOFTWARE); and supermarkets and other grocery stores (RETAIL). As discussed in the development of the Research Questions, at least three of these industry sectors possess industry-specific common ethical issues (conflicts of interest and animal testing for pharmaceuticals, environmental protection issues for extractive, and worker safety for manufacturing). Thus, the sampling technique also yields a population of firms that have industry- or firm-specific ethical issues available for discussion in their Codes.15 All firms within each industry were ranked by size (total assets and sales) and stratified into quintiles according to their relative position in the rankings by these measures.16 Three firms were randomly chosen from each industry-based size quintile, for a total sample size of 75 firms. Table 1 provides descriptive statistics pertaining to the sample firms. All firms were traded on either the NYSE or the NASDAQ and thus fell under the requirements to provide a Code of Ethics in 2004. [Insert Table 1 here] The Codes The NYSE (2004) and the NASDAQ (2003) both required that listed firms must have and provide to the public a Code starting in fiscal year 2004. Despite this, six of the sample firms either did not provide a Code or provided an incorrect address for the Code; thus, Codes were not available for these firms. For another three of the firms that did provide a Code, the electronic document provided consisted of graphic scans of typewritten documents instead of PDFs converted from word-processed documents, and thus were not machine-readable and impossible to analyze with the software. Therefore, the final sample of available, machine-readable Codes consisted of 66 firms. The Analysis Technique The study makes use of multiple approaches to content analysis, an accepted method for exploring narrative disclosures (see Gray et al. 1995; Deegan and Gordon 1996; Milne and Adler 1999; Campbell 2000). Research Question 1 asks about convergence in content, irrespective of the language used to convey that content, and requires an evaluation of the structure of each document in order to ascertain the degree to which similarities exist across firms. Research Question 2 asks about convergence in the language used to convey content, and requires an evaluation not of the structure of the documents, but of the volume of unoriginal content and language. Generally, content analysis requires development of a rubric by which content will be evaluated and measured. Research Question 2, however, requires a substantially different approach in that the primary question to be answered is not “what is the content?” but rather “how much unoriginal content does this document contain, and
12
what is it?” The issue of unoriginal content is one that is faced by every educator who assigns written work to students: in the classroom setting, the incorporation of unoriginal material into a document is known as plagiarism. Therefore, the research makes use of content analysis software designed to discriminate between original and unoriginal material in a document. The program used was MyDropbox (2010), a commercial web-based artificial intelligence that interprets text both for the use of direct quotations and for poorly paraphrased material. 17 The system defines poor paraphrases as the use of text wherein the original author’s method of expression and sentence structure have been retained and minimal word substitution has been provided, where large portions of text have been retained and rearranged in a patchwork style, or where material has been directly quoted and rearranged by clause switching, or simple substitution of common synonyms. The assessment of documents is based on a comparison of the given document with a database comprised of internet documents indexed by MSN Search Index, the body of literature indexed by ProQuest/ABI Inform, the FindArticles database and internet paper mills. The sample firm Codes were thus compared against all other Codes indexed by these sources, as well as the Codes that were submitted for all other sample firms. The software returns a summary evaluation of the percentage of unoriginal content included in the document based on the direct and paraphrased material. It also returns a copy of the document that is marked and cross-referenced so that the “suspect” content can be instantly compared with the “match” located in the service’s databases. Only one match is identified per block of text, regardless of how many matches may have been available. This software was used primarily for analysis related to Research Question 2. The 66 machine-readable Codes provide the basis for the following analyses. Company names, company contact information, and names of individuals or positions responsible for compliance reporting were disregarded for purposes of establishing the originality or lack thereof in the content. The drafters of these Codes generally subdivide the documents into sections covering major topical areas. For the analysis of Research Question 1, each topical section was scrutinized individually by MyDropbox (2009) for the number of unique concepts, thoughts, or statements contained in the section, using the sentence as the unit of analysis. Through this process, a listing of all unique ideas was developed for each section of each document. Listings were combined across documents to derive frequency counts of topical coverage by major section as well as itemized coverage within each section. For the tests of Research Question 2, words, phrases, and sentences form the units of analysis. Each document was passed through the content analysis software. Where entire sections appeared to have been copied, the section provides an additional unit of analysis. Each piece of text in each document was then classified as original or unoriginal. Word counts for the entire document were obtained and used to determine the density of copied material. Research Question 3 involves assessing the Code content for language suggesting a decoupling strategy, whereby the external response to the regulation is being separated from the internal structures used to enforce conduct, or where the external response is so generic as to fail to provide useful direction for employees seeking guidance for right action. Answering the question of how prevalent this type of decoupling is requires scrutinizing the major sections of the text as defined for the analysis of Research Question 1 (above) for language and phrasing that suggests that either the topical area is vague; that enforcement structures are poorly defined, inaccessible, or not available equally throughout the organization; or that areas that should offer concrete guidance do not. A further
13
issue is the accessibility of the Code itself; Codes that are sufficiently comprised of legalistic language may not be accessible to employees throughout the organization at all, thus guaranteeing a decoupling between the form and substance of the response. As discussed, this assessment constitutes an informed impression of the content rather than a robust assessment of the content and may offer interesting insights rather than persuasive evidence. The two aspects of institutional isomorphism – coercive and mimetic – are expected to drive similarities in the Codes on different levels. The existence of regulatory requirements pertaining to minimum coverage represents a coercive element and should lead to convergence in content. The mimetic elements associated with the uncertainty in the organizational environment and litigation risks should lead to convergence in language.18 To preserve clarity of exposition, the content and language will be discussed as separate matters through the remainder of the paper. However, because of the co-determination it is difficult to entirely separate the matters; therefore, the discussion will necessarily refer to other sections of the paper from time to time. 4. DISCUSSION OF RESULTS Convergence in Content Arising from 2002-2004 Regulation The minimum coverage suggested by the NYSE (2004) and the NASDAQ (2003) includes several categories: honest and ethical conduct including ethical handling of conflicts of interest; full, fair, accurate, timely, and understandable disclosure; compliance with applicable laws, rules, and regulations; reporting of violations of the Code. The regulators intended managers to extend beyond these minimum requirements in their development of a Code (Sarbanes-Oxley 2002; SEC 2003; NYSE 2004; Verschoor 2002) ; however, institutional theory suggests that the regulatory origin of the list of minimum topics covered infuses the list of basic requirements with legitimacy and thus provides an anchoring point for organizations concerned primarily with establishing legitimacy. That is, isomorphic pressures suggest a convergence on the regulator’s list rather than deviations – even if the deviations are made in the interest of developing an effective and authentic Code. An examination of the codes suggests that there are two basic templates (hereafter, Template 1 and Template 2 for convenience). Nine firms provided Template 1 Codes. Template 1 is featured in Appendix 1 and is an extremely brief document that advances the minimum required material as a cursory checklist. Documents in this category are no more than a single page long. In some documents, multiple items are subsumed into a single section of the Code (e.g., treating business opportunities or confidential information as a type of asset and including these and the physical assets of the firm as items that must be protected and not used for personal gain). All but two of these Template 1 Codes forthrightly acknowledge – within the content of the Code itself – that the Code is provided for no other reason than compliance with the regulation. Of the nine sample firms that have chosen this route, one is an oil firm, two are pharmaceuticals, three are software publishers, and three are grocers. Three of the firms are in the smallest quintile within their industry, one each in the second and third quintiles, two in the fourth, and two in the largest industry quintiles. There are no industry or size trends evident in terms of which organizations choose this route. The text that appears in Appendix 1 is the entire contents of that firm’s Code of Ethics –provided in explicit response to the regulatory requirements – a document that the company (a grocer) indicates applies generally to the members of the firm. While an extreme example, the Code in Appendix A is identical to the Codes of at least 50 other publicly-traded U.S. companies (based on a Google text search conducted in the summer of 2008), and cannot
14
be considered to represent an unusual phenomenon. It offers the regulators’ minimum content as a checklist and makes no effort to elaborate upon the elements, to provide any substantive guidance, or to tailor the document to the specific needs of the firm. Indeed, we may understand from this document that the stocking clerks, the cashiers, and the shopping-cart valets have guaranteed to the public that they will provide us with full, fair, accurate, timely, and understandable [financial] information and make every effort to avoid conflicts of interest with competing grocers. Even should this firm, as several others do, restrict the use of this Code format and content to apply only to the very senior executives, the lack of local consideration is not in the least consistent with the expressed desire of the regulators that firms pursue the substance, rather than the form, of the Code. In many cases, the company does not provide a statement about which employees are covered by the Code. Only five Codes indicate that they pertain exclusively to the top executives and those with financial management responsibilities. Template 2 (represented in Appendix 2) is the standardized approach employed by the remaining sample firms. Of the 57 firms that provided more than a checklist of items in compliance with the minimum standards, 53 provided Codes that complied with all minimum requirements. Of the four Codes (7% of Template 2 firms) not in compliance with minimum standards, three neglected to discuss confidentiality of company information, two did not discuss fair dealing, and one did not provide any discussion about how compliance with the Code was to be enforced (including a discussion of how violations of the Code were to be reported). Of particular interest with respect to the use of Codes to address the contemporary legitimacy crisis is the pervasive presence of material directly related to factors implicated in the recent failures: law-breaking and a lack of full, fair, honest, complete, and accessible financial reporting. All 66 firms covered these items in their Code, suggesting support for Research Question 1 (that Code content converges across organizations in a manner consistent with the presence of institutional isomorphic forces). Convergence in Content Arising from Other Institutional Pressures Regulatory minimum standards and Section 406 of SOX are not the only significant institutional pressures likely to generate convergence. An incentive to provide a Code prior to the recent mandate was the availability of reduced criminal penalties available to firms under the Federal Sentencing Guidelines for Organizations for violations of the Foreign Corrupt Practices Act, in particular, the payment of bribes to foreign officials (HLR 2003). The primary crimes for which organizations can be convicted through the actions of their employees are bribery, insider trading, and trade violations. Consequently, the convergence in content should extend to coverage of these three topics for firms susceptible to legal risks. The risk-management angle of this isomorphic behavior suggests isomorphism by risk group rather than across the entire sample. Bribery, corruption, and insider trading are risk factors for all sample firms. Trade restrictions would pertain only to firms that engage in global trade, and members of certain industries for whom there exist special trade restrictions (i.e., software publishers [Toedt 1997] and producers of medical equipment and drugs [FDA 1996]). There is no evident size trend; while more firms that include this information come from the largest 2 quintiles within their respective industries, nearly half of them are located in the smaller size strata. Anti-bribery provisions and insider trading prohibitions were provided by 37 (65%) and 42 (74%) of the Template 2 firms, respectively. Global trade restrictions are mentioned by a smaller portion of the sample (16 or 28% of Template 2 firms). The global trade restrictions are mentioned more frequently by
15
software publishers (4 firms) and medical equipment and drug manufacturers (8 firms) more frequently than by any other industry. Firms including this material tended to be in the larger than the average firm within their respective industries. To the extent that recent ethics-oriented rhetoric and regulation jointly influences the content of the Codes, earlier diversity-oriented rhetoric and regulation from the passage of equal employment legislation may also still be evidence in the Code content.19 Recent years also feature a new interest in corporate social responsibility that may exert normative pressures on the Code content.20 The Template 2 Codes demonstrate evidence of response to these forces, in that they feature discussion of diversity, harassment, and related human resources content (35 firms, or 61% of the Template 2 Codes) and health, safety, and environment matters (25, or 44% of the Template 2 Codes). A total of 42 of the Template 2 Codes (74%) feature content driven by these other institutional pressures. The number of Template 2 Codes that contain only content driven by these institutional pressures (i.e., they do not have any content that cannot be directly traced to one or more of these institutional pressures) is 32, or 56% of the sample firms. The total number of Codes (both Templates) that contain only institutional content is 41, or 62% of the firms supplying usable Codes. This degree of convergence is consistent with RQ1 (that content will converge across organizations) and suggests that the content of the Codes is driven for the majority of firms strictly by institutional pressures. Ethics-Oriented Rhetorical Content All Codes specifically advance ethics-oriented rhetoric, generally in the introductory material. The consistency and manner in which they do so is remarkable. Table 2 displays a representative sample from this introductory material in which the firms advance their commitment to the “highest” standards (and the number of sample firms using each of the quoted lines of text in their Code). Of the firms advancing claims to adherence to the “highest standards” of ethics, morality, and integrity, 22 of them limit their Codes to content driven strictly by institutional pressures. Furthermore, many Codes include additional rhetoric pertaining to the Company’s dedication to behavior that exceeds minimum standards, statements such as “Obeying the law, both in letter and spirit, is the foundation upon which this Company’s ethical standards are built” and “…in many instances this Code goes beyond the law”. Assertions of this type were made by 40 of the sample firms, including two that provided the checklist-approach of Template 1 and 20 Template 2 firms that provided no coverage beyond that stimulated by institutional pressures (3 of which provided nothing beyond the regulatory minimum content). These inflated claims are consistent with attempts to gain legitimacy by proclaiming the firm’s adherence to the popular ethics rhetoric. This also is consistent with Research Question 1, in that the institutional responses to the legitimacy crisis framed the origins of that crisis as ethical failures and made extensive use of ethics rhetoric to support and reinforce this proposition. [Insert Table 2 about here] Firm- and Industry-Specific Content Thirty-four firms provided some firm-specific or industry-specific content in their Codes. The volume of text dedicated to original content varied greatly, with nine of these firms providing original content comprising less
16
than 1% of the document, ten firms providing between 1% and 10% original content, eight firms providing between 10% and 30% original content, and seven firms providing more than 30% original content. Original content included original ethics rhetoric (11 firms), non-standard examples of gifts and/or industry-specific potential for conflicts of interest (10 firms), discussions of the industry’s regulatory environment (7 firms) ethics of marketing (7 firms) and work place safety (6 firms). Detailed information about compliance programs and/or ethics reviews was provided by only five firms. Industry-specific competitive factors, product safety and quality control issues, and foreign business practices were discussed (with original text) four times each within the sample. Details of insider trading plans, industry-specific environmental issues, and harassment and diversity issues were discussed with original text three times each within the sample. The integrity of the scientific research process, training programs for ethical issues, human rights matters, ethical duties to customers, and industry-specific external Codes of Conduct were discussed twice within the sample. The majority of the items on the preceding list were yielded by seven of the sample firms (who provided a minimum of 37% original content in the Code of Ethics). Most items on the list were comprised of short blocks of text one or two sentences in length. Substantial industry-specific content was provided by five firms (two pharmaceutical firms and three medical instrument manufacturing firms). One pharmaceutical expressed concern for the ethical climate surrounding the research process: “Research integrity is fundamental to scientific progress and to Genelabs' ability to discover and develop novel products. All of Genelabs' research and development must be conducted according to all applicable laws and regulations and to the generally accepted ethical standards of the scientific community. Scientific misconduct, such as fabrication, falsification or plagiarism in proposing, conducting or reporting research disregards the intellectual contributions and property of others, impedes the progress of research and corrupts the scientific record. All such activities are strictly prohibited.” (Genelabs) While the other pharmaceutical expressed more varied concerns based on the industry, including drug marketing practices, compliance with specific healthcare laws, patient privacy matters, and mentioned that the firm had adopted a specific industry-wide Code of Conduct, along with various other pharmaceutical- and R&D-intensive ethical concerns: “Communicating Honestly: It is in Everyone's Best Interest. Trust is a significant part of our business. Medical professionals trust our research and results. Consumers trust the quality of our products. Shareholders trust we will continue to build value for them. That is why honest communication about product experience is absolutely essential. You are required to inform the Company of any adverse reactions to products when you become aware of them...Product Experience Disclosure. Pfizer has a worldwide practice of keeping medical and veterinary professionals fully informed of the uses, safety, contraindications, and side effects of our products and, where appropriate, their operational requirements and characteristics. We provide this information using: package inserts, mailings to physicians and other healthcare professionals, educational and/or promotional materials, and presentations by our service representatives. The information provided must be consistent with the worldwide body of scientific knowledge pertaining to the relevant products and must comply with local requirements of good medical practice and government regulation.” (Pfizer) Of the medical technology manufacturers providing industry-specific content, one described a position of compliance with Medicare anti-fraud provisions, indicated that the Company subscribed to an industry-wide Code
17
(the AdvaMed Code) but did not provide that Code or a means by which that Code could be obtained by interested parties, and made a commitment to maintaining patient privacy. Another mentioned the Company’s adoption of the AdvaMed Code and provided a link to the AdvaMed web page. This company also provided text pertaining to quality control issues pertinent to the industry: “QUALITY AND REGULATION OF MEDICAL DEVICES Many of the products that BD makes and sells are used to diagnose and treat patients all over the world. To help successful diagnosis and treatment, these products must be medically safe and effective for their intended use. The Food and Drug Administration in the U.S. and similar agencies in other countries and regions have established requirements to ensure the safety and effectiveness of medical products. These requirements have the force of law in most countries. The regulatory agencies that enforce them are recognized by the public for protecting the health of the population. BD has always believed that quality management and product quality are among our most important values, along with the prevention of regulatory violations. We help to ensure product quality by: adhering to regulatory and generally accepted good manufacturing and laboratory practices, and quality system requirements; conducting product clinical trials in accordance with regulatory and ethical standards; making accurate product claims, supporting them with product testing and clinical trials where appropriate; properly registering all products by submitting true and complete information; properly labeling, advertising, and promoting our products; and responding to complaints and other indicators of potential problems and taking timely and appropriate corrective action. Paying attention to quality is good for the patient, good for BD, and good for our personal satisfaction in our work. Failure to meet product quality requirements can expose the patient to possible harm, and can expose BD, and you, to serious civil and criminal fines, and even imprisonment. If you have any questions or issues in this area, please contact Corporate Quality Management or Corporate Regulatory Affairs or, if applicable, your regional Regulatory Affairs associate.” (Becton Dickinson) The third medical equipment manufacturer provided a review of general principles of bioethics: “Bioethics Baxter believes that biotechnology can provide significant benefits to life and that these technologies carry with them responsibilities. Any Baxter decision to commercialize a biotechnology-derived product or therapy will receive careful consideration of the risks and the benefits of the technology in light of the information available. The ability to perform an activity will not automatically justify the activity. As a company, these principles and processes govern our efforts to research, develop, and provide products for critical therapies for patients worldwide.” (Baxter International) The Code content provided above yields a picture of the type of variability between firms – the inclusion of content specific to the particular firm’s ethical issues that provides concrete guidance and transparency as to the internal processes to be followed in settling ethical issues – that presumably was envisioned by the framers of SOX when coming up with the code requirement. Despite the existence of a small number of firms providing significant industry-specific content, the majority of firms yielded little or none of this type of material. Furthermore, the industry-specific content is concentrated in a limited industrial group (pharmaceuticals and biomedical device manufacturing firms), yet other industry sectors should also possess driving ethical issues warranting coverage in the Code of Ethics (for example, environmental matters for the extractive firms). The presence of the type of disclosures requested by the regulators among a very few firms cannot be considered sufficient evidence to reject Research Question 1 (institutionally-driven isomorphic content); if anything, it merely highlights the “blandness” of the content of the typical Code.
18
Even companies that provide industry- or firm-specific addenda to the coverage stimulated by institutional pressures adhere closely to the mandatory listing requirements. Nor is there any statistical evidence to suggest industry or size trends in terms of which type of firm includes which type of information. 21 The lack of discernable pattern suggests, for example, that grocery retailers are just as concerned with issues of confidentiality as R&Dintensive pharmaceuticals and software developers, while software manufacturers are concerned with the same ethical issues as major oil and gas firms. This approach does not strongly suggest a desire on the part of the drafters of these Codes to create a document that is highly tailored to the individual needs of the organization and certainly flies in the face of the regulators expressed intent that the Code content should vary from firm to firm.
Language Suggesting a Decoupling Dynamic Research Question 1 recalls that the response to institutional forces frequently requires a type of decoupling of organizational processes from response in order to reduce the frictions arising from a lack of alignment between the behavior preferred by the organization and that implied by the symbolic response. Numerous firms provided language indicative of a decoupling dynamic, separating the symbolic response from the true internal processes responsible for determining the ethics of any given action and enforcing ethical conduct within the organization. Decoupling internalizes the compliance and enforcement functions, rending them opaque to parties who are external to the organization and possibly hiding the process even from members of the organization. The following statement is an example of language that suggests decoupling and hidden processes, in that it is clear that waivers of the Code may be obtained, but not how, when, or under what circumstances this will happen: “The Corporation will waive application of the policies set forth in this Code only when circumstances warrant granting a waiver…” (Med Gen, Inc.) The following text appears to provide more information. Careful reading, however, yields the information that the process of enforcing the Code (conducting investigations and punishing wrong-doers) is still fundamentally hidden: “The [Board of Directors/Audit Committee] shall determine, or designate appropriate persons to determine, appropriate actions to be taken in the event of violations of [the Code]…Such actions shall be reasonably designed to deter wrongdoing and to promote accountability for adherence to the Code…” (Antares Pharma, Inc., Rochester Medical Corporation, Rockwell Medical Corporation) This type of nebulous and legalistic language is the type of response predicted by HLR (2003). It is so vague as to be unenforceable and leaves both external parties and members of the organization looking to the Code for guidance uncertain what the penalties for violation will be, or how they will be determined. As a contrast, consider the following text – which does not specify the investigation process, but provides a range of likely penalties to be imposed upon violators: “Failure to comply with the standards outlined in this Code will result in disciplinary action including reprimands, warnings, probation or suspension without pay, demotions, reductions in salary, discharge and restitution. Certain violations of this Code may require the Company to refer the matter to the appropriate governmental or regulatory authorities for investigation or prosecution. Moreover, any supervisor who directs or approves of any conduct in violation of this
19
Code, or who has knowledge of such conduct and does not immediately report it, also will be subject to disciplinary action, up to and including discharge.” (Art Technology Group, Inc.) Decoupling language in other Codes renders the boundary of the Code itself a matter for internal interpretation: “While some of the policies contained in this Code must be strictly adhered to and no exceptions can be allowed, in other cases exceptions may be possible.” (Art Technology Group, Inc., MRO Software, Inc., Icagen, Inc.) 22 Contrast the preceding language with the following text (from a different firm) emphasizing the degree to which the boundaries of the Code are not considered flexible by the organization: “OUR BASIC POLICY: "WE DO WHAT IS RIGHT" It is [the Company’s] policy to comply with all laws, rules and regulations pertaining to its businesses around the world and beyond this to act in an honest and ethical manner at all times. You should never "assume" or "read between the lines" that [the Company] ever wants you to violate a law or regulation, or to act unethically in your work even if asked or ordered to do so by your supervisor…Respecting and following the principles of this Guide is a condition of your employment with [the Company]. You will never advance your career with [the Company] by violating these principles – but you could end it.” (Becton, Dickinson and Company) This approach pertains to individual elements of the Code, not just to the overall tone or enforcement policies. Many of the Template 2 Codes (37 Codes) include sections on the ethics of accepting gifts. These sections come in one of two forms: one type consists of a generic injunction against accepting gifts that would be likely to cause a conflict of interest: “The purpose of business entertainment and gifts in a commercial setting is to create good will and sound working relationships, not to gain an unfair advantage with customers. No gift or entertainment should ever be offered, given, provided or accepted by any Company employee, officer, director, family member of an employee, officer, director or agent unless it: [1] is not a cash gift, [2] is consistent with customary business practices, [3] is not excessive in value, [4] cannot be construed as a bribe or payoff and [5] does not violate any laws or regulations.” (Antares Pharma, Inc., Ariba, Inc., Goodrich Petroleum Corp., Merit Medical Systems, Inc., Ultimate Software Group, Inc., W&T Offshore, Inc.) The second type provides concrete guidance as to what gifts are acceptable vs. unacceptable and items employees are likely to encounter that may not immediately be recognized as a “gift”, as well as restrictions on acceptable gifts and instructions on what to do if a vendor provides a gift that is unacceptable under company policy: “Team Members and directors should not give anything of value to anyone, or accept with anything of value from anyone, when doing so might compromise or appear to compromise the objectivity of business decisions…Some gifts and entertainment are allowed as follows: (1) Gifts with an established value of $25 or less are generally allowed. (2) Business-related meals of nominal value are allowed, subject to specific requirements in [named internal policy]. (3) Gift baskets or flowers may be accepted within reason, but they must be made available for sharing with everyone…(4) Promotional items, such as those bearing a vendor's logo may be accepted up to total estimated value of $25. (5) Existing Team Members may accept samples of new or reformulated products, and new Team Members may accept samples of existing products (one time only). It is not acceptable for Team Members to receive for their personal use multiple samples of the same product from a vendor. (6) Store-level Team Members may accept a vendorpaid trip made for the sole purpose of education and training. The vendor may pay for all expenses
20
including airfare, accommodations and meals. There is a one-time limit on vendor- paid trips unless there is a significant change in products, programs or business practices. Global and regional Team Members may not have any expenses for a trip paid by the vendor. If someone tries to give you a prohibited gift, you should also tell your Team Leader. Then, either return the gift or personally reimburse the giver of the gift for its full value.” (Whole Foods Market) The inclusion of concrete guidance renders this Code both enforceable and binding, and constrains the ability of employees or management to exploit loopholes. At the highest levels, the use of this type of concrete language may present an impediment to organizational flexibility. The evidence provided in analysis of Research Question 1 suggests that organizations make at least limited use of decoupling language, consistent with the predictions of Meyer and Rowan (1977). If an organization is not interested in having the behavior of its management constrained by the Code it is likely to use vague decoupling language, or to directly state that a decoupling process of some type exists (consigning to internal organizational processes the determination of boundaries or the definition of compliance and enforcement mechanisms). Explicit decoupling language internalizing the organizational processes pertaining to the management of corporate ethics was included in 51 of the 66 usable codes (77% of the total pool of Codes). This type of behavior is a distinctive characteristic of the symbolic responses generated by institutional pressures for conformity (Meyer and Rowan, 1977). Decoupling Language and Financial Reporting Practices One section of the Codes that is notably not characterized by the use of decoupling language is the material dealing with financial reporting and record maintenance; this material is generally very concrete in nature, with explicit boundaries, firm guidelines, and detailed prescriptions for appropriate behavior. It is not clear whether the lack of vague language in these sections is a result of a generally-accepted vocabulary and set of guidelines for appropriate treatment (i.e., GAAP), or if it is a result of increased sensitivity to these issues in the wake of the massive financial reporting frauds that triggered the Code regulation in the first place. Either way, this section of the Codes is unusually concrete and appears intended by the authors to provide binding constraints on organizational behavior: “Our records are the basis of our earnings statements, financial reports and other disclosures to the public and are the source of essential data that guides our business decision-making and strategic planning. Company records include booking information, payroll, timecards, travel and expense reports, e-mails, accounting and financial data, clinical records and data, measurement and performance records, electronic data files and all other records maintained in the ordinary course of our business. All Company records must be complete, accurate and reliable in all material respects. There is never a reason to make false or misleading entries. Undisclosed or unrecorded funds, payments or receipts are inconsistent with our business practices and are prohibited. You are responsible for understanding and complying with our record keeping policy…Inaccurate, incomplete or untimely reporting will not be tolerated and can severely damage the Company and cause legal liability. Employees should promptly report evidence of improper financial reporting. Examples of evidence that should be reported include: financial results that seem inconsistent with the performance of underlying business transactions, inaccurate Company records, such as overstated expense reports, or erroneous time sheets or invoices, transactions that do not seem to have a good business purpose, and requests to circumvent ordinary review and approval procedures.” (Conceptus, Inc.)
21
Fifty-four of the Template 2 firms provided a separate section on record-keeping and financial reporting, generally with the level of detail provided in the excerpt above. The consistency with which detailed, binding guidance was given in this area is remarkable, and perhaps best explained by the introductory text for the reporting section of one of the sample firm’s Codes: “Remember Enron? Everybody knows about them – they set up their financials to look a lot better than they actually were. Well, this policy exists to always make sure that [the Company’s] name never ever is tied to a financial scandal.” (Wild Oats Markets, Inc.) The list of items covered in the record-keeping and reporting section (see Appendix 2) reads like a laundry list of the elements of the Enron fraud. This material appears to be a direct and unambiguous response to the legitimacy crisis provoked by the major accounting frauds and likely consists of a robust attempt at rebuilding legitimacy through providing assurances of the quality and credibility of financial reports and the accounting process. Consistent with this, five firms provide prohibitions against intimidating, threatening, or lying to the company’s outside auditors, while 21 firms provide a section of the Code specifically requiring employees to cooperate with any federal, state, or other criminal investigation procedures. The convergence in the coverage of this section of the Codes provides a clear depiction of isomorphic behavior in response to coercive forces, and – distinctively – a response to these forces that is designed for substance rather than for form. Furthermore, the material upon which the Codes converge with respect to ethical issues surrounding financial reporting is, unlike the other material upon which the Codes converge, distinguished by the use of concrete prohibitions and binding constraints. This area, unlike all other areas of the Code, is notable for the lack of decoupling language.
Convergence in Language Research Question 2 suggests convergence in the language in which the Codes are framed. Table 3 provides, for reference, descriptive statistics about the raw data, the length (in words) of the Codes included in the study. [Insert Table 3 about here] Table 4 displays the associated percentages of unoriginal content in the entire document, based on the reports from the analysis software. The statistics displayed in Table 4 indicate that the sample Codes are primarily comprised of unoriginal material. Many Codes (nearly 40% of the viable sample) contain no original material at all, other than the name of the company, the date of the Code, and the firm-specific details involved in reporting violations (management titles, phone numbers, mailing addresses). Another 18 of the firms (27%) provide Codes with less than 5 percent original content. Only nine Codes provide Codes that contain less than 80% boilerplate. The numerical evidence provides support for Research Question 2. 23 [Insert Table 4 about here] In many cases, there has been no effort at providing a Code that appears to have been authored by a member of the firm or by the firm’s Board of Directors. Table 5 provides statistics on the use of boilerplate by major Code section. As shown in Table 5, the median unoriginal (boilerplate) content for all major sections of the Codes is 100% - or in other words, at the median, zero percent of the content for any given section constitutes firm-specific guidance to employees.
22
[Insert Table 5 about here] A typical example of the type of boilerplate suggested by the statistics in Table 5 is provided in Table 6. These excerpts are from the Gift policies of the Codes of four firms in different size quintiles and industries. The text is virtually identical from one excerpt to the next; differences involve minor changes to the wording, punctuation choice, and formatting of the text. [Insert Table 6 about here] The section on competition and fair dealing is characterized by exact copying rather than poorly paraphrased material, as seen in this text used by nine of the sample firms (verbatim), and by an additional 14 firms (with very minor word substitutions): “[w]e seek to outperform our competition fairly and honestly. Stealing proprietary information, possessing trade secret information that was obtained without the owner's consent, or inducing such disclosures by past or present employees of other companies is prohibited. Each employee should endeavor to respect the rights of and deal fairly with the Company's customers, suppliers, competitors and employees. No employee should take unfair advantage of anyone through manipulation, concealment, abuse of privileged information, misrepresentation of material facts, or any other intentional unfair-dealing practice”
In a like manner, no firm includes any original material in the discussion of corporate opportunities, as seen in the use of the following text (verbatim) by 19 of the Template 2 firms: “[e]mployees, officers and directors are prohibited from taking for themselves personally opportunities that are discovered through the use of corporate property, information or their positions without the consent of the Board of Directors. No employee may use corporate property, information, or his or her position for improper personal gain, and no employee may compete with the Company directly or indirectly. Employees, officers and directors owe a duty to the Company to advance the Company's legitimate interests when the opportunity to do so arises” The use of similar language throughout the sample is consistent with the notion of mimetic isomorphism. The environment surrounding the development of these Codes is characterized both by an explicit need to rebuild legitimacy and by significant uncertainties arising jointly from vagueness in the regulatory environment and aggressive legal liability and litigation regimes (Seetharaman et al. 2002; Baginski et al. 2002; and Macey 2007). These are circumstances under which mimetic isomorphism is thought to flourish (DiMaggio and Powell 1983). For example, symbolic responses become invested with authority through an endogenous process involving the legislators, the organizations, and the enforcement bodies. This suggests that the use of boilerplate text in Codes is a rational response to the uncertainty, and one that may provide legal protections through the widespread adoption of the text. That is, organizations may garner protection from organizational threats arising from the publication of binding constraints by adopting a “herd mentality” and taking refuge in a standardized response.24 Edelman et al.’s (1999) theory suggests that some sections of the Codes should involve more extensive boilerplate than others pursuant to different levels of exposure to liability. Even firms that have clearly made an effort to tailor the Code content and language to their organization rely heavily on boilerplate in certain sections. The nine firms that provide Codes that contain at least 20 percent original material generally devolve into boilerplate in sections discussing insider trading, conflicts of interest, competition and fair dealing, employment matters,
23
reporting and disclosure matters, proper use of company assets, confidential information, gifts, compliance with laws, antitrust, and bribery – all areas of particular legal risk. With respect to the content sections on recordkeeping, reporting, and disclosure, only four of the 66 sample firms include any original material at all, and the use of original material in those firms is limited to conversational prefaces and slides promptly into boilerplate text for the main body of the content in that section. Only six of the sample firms include original material in their employment and human resource coverage (employment, discrimination, and harassment). This degree of convergence in language is strongly supportive of Research Question 2, and of a rational managerial response to legal risks. In particular, the extensive use of boilerplate text in regard to corporate reporting reflects the heightened levels of risk generated by the preceding major and public business failures that led to the imposition of regulatory requirements for providing a Code.
Additional Analysis Even areas that one would expect to contain original material in many instances do not. Eleven of the Codes are prefaced by letters from the CEO to the Company’s employees. Several letters offer details about support resources, additional supplemental material available from the Company, and means of reporting violations or suspected violations of the Code. They also assign responsibility for the contents of the Code to the employees. However, for the most part these letters consist of extended streams of ethics-oriented rhetoric: “The reputation and integrity of the Company are valuable assets that are vital to our success.” (Auxilium, Inc.) “The Company has always held itself to the highest standards of business conduct. Our commitment extends beyond compliance with the law to include a firm belief that the best way to be a great company and to deliver value to our customers, associates, shareholders and communities is to be fair, honest and ethical in our business practices and personal behavior at work.” (Becton, Dickinson and Company) “Personal and corporate integrity has been one of the Company’s guiding principles since the day our company was founded. It enters into everything we do and is a central part of our daily lives.” (Boston Scientific Corporation) While the content of the letters is generally original, one of the letters is approximately 50% unoriginal in content, and two of them are completely unoriginal. The three unoriginal letters are distinctive through their evident generality and lack of any industry- or firm-specific content, including the additional resources discussed above. The following letter was advanced as a personally signed letter from the CEO yet is completely unoriginal except for the company’s name: “The Company is dedicated to conducting its business consistent with the highest standards of business ethics. We have an obligation to our employees, stockholders, customers, suppliers, community representatives and other business contacts to be honest, fair and forthright in all of our business activities. As an employee of the Company, you are faced every day with a number of business decisions. It is your personal responsibility to uphold the Company's high standards of business ethics in each and every one of these situations. It is not possible for our Code of Business Conduct and Ethics (the "Code") to address every situation that you may face. If you use your good business judgment and experience, the majority of your business decisions are not likely to raise ethical issues. When
24
you are faced with an ethical issue, we hope that this Code will serve as a guide to help you make the right choice. We encourage you to take this opportunity to review our policies and to discuss any questions you may have with your supervisor, an HR representative, the Company's legal counsel or with any member of the Company's management team directly. The guidelines set out in this Code are to be followed at all levels of this organization by our directors, officers, employees and agents. We rely on you to uphold our core values and conduct our business honestly, fairly and with integrity.” (Conceptus, Inc) The additional analysis yields the finding that even in areas where institutional pressures are not significant – in material not indicated by the regulation as obligatory to cover – isomorphic responses are still observed. The President’s/CEO’s letter shown above is replicated in full by a large number of non-sample firms including WNS Global Services, Mellanox Technologies, First Consulting Group, Compass Minerals International, and Nile Therapeutics, Inc., each of which advance it as a personal letter signed by one or more executives, and is provided in boilerplate format on onecle.com. The lack of originality in this area is more difficult to explain with the legal riskmanagement argument; the introductory material is entirely optional and presumably intended to convey a degree of sincerity about the “tone at the top”, or management’s personal commitment to ethical behavior. The use of formulaic approaches and unoriginal copy in this area may suggest a profound commitment to complying with the form of the laws rather than seriously considering or adhering to the substance of them.
5. CONCLUSIONS U.S. governmental and regulatory bodies advanced a requirement in SOX (2002) for publicly-traded companies to have and to provide a Code as a response to the legitimacy crisis triggered by business failures around the turn of the century. Regulators expressed the expectation that Codes of Ethics would vary from one firm to another based on internal organizational consideration of ethical issues unique to each firm, yet this study provides evidence that the structure, content, and language of the Codes is instead consistent with a widespread response to isomorphic pressures. Far from varying from firm to firm, companies instead appear to have made generous use of boilerplate copy that is generally vague and lacking in substantive binding constraints on organizational behavior. Codes that are ostensibly provided to guide the daily decision-making activities of employees are instead generic carbon-copies of one another, with little or no attempt for most firms to provide any sort of guidance tailored to the ethical pressures unique to the organization or to the industry. As anticipated by the HLR (2003) critique of the pending regulation, most Codes are now boilerplate documents couched in nebulous and legalistic terms, unlikely to constrain actions, and difficult to enforce. The content of the Codes suggests a rational yet symbolic response to the regulation, in that they provide for decoupling of the organization’s public response from the internal workings of the firm in order to manage frictions between the necessary content of the symbolic gesture and the underlying reality of the organization. When law is ambiguous or vague – as in the case of Code regulations – the possibility for managers to construct the meaning of compliance with the law is great (Edelman 1992). Hampel (1998) discusses the responsiveness of organizations to the law – and to lawmakers’ understanding and anticipation of those responses – and notes that both parties understand the likelihood of compliance with the letter and simultaneous flouting of the spirit, but also notes that no acceptable or effective solutions to this quandary have been advanced. The relative
25
freedom of managers to interpret the law, the predilection for box-ticking behavior, and the understanding on all sides that box-ticking is a likely response to regulation all lead to a scenario ripe for ritualistic responses that emphasize form over substance. If managers emphasize only the form of codes and if auditors also merely use a check box to evaluate the code (Cohen et al. 2008), then the existence of a code may not significantly improve the ethical behavior of the firm. Although Lindsay et al. (1996) suggests that codes may be effective if properly implemented; in the end they conclude that codes are primarily mere lip service. While some Codes display evidence of the cynical response to these frictions that was projected by Meyer and Rowan (1977), most of them provide the appearance of an authentic effort to articulate organizational ideals and ethics that is characteristic of the decoupling response also projected by Myers and Rowan (1977). However, because most codes use deliberately vague or nebulous language in most sections it appears that many, if not most, Codes are designed to minimize the likelihood of being used as part of a formal evaluation or investigation. A distinctive exception to the general finding that specific prescriptions or prohibitions are not provided or that language is sufficiently vague as to render it unenforceable, is the section of the Codes pertaining to the responsibility to maintain financial records and to provide full, fair, accurate, and timely disclosures. This area is one of particular import given that the impetus for requiring the Codes in the first place was the rash of catastrophic financial reporting frauds. Any effort to repair the legitimacy of underlying economic institutions would be obliged to direct attention to this area, given that the failures and frauds were framed as a function of ethical failures. SOX and the ensuing ethics-directed regulation from the SEC and the stock exchanges is directed at addressing this aspect of the failures. With respect to the remaining portions of the Codes, the appearance of authenticity is superficial; the vast majority of the Codes are built around boilerplate text that is essentially copied from other firms. 25 The lack of firmor industry-specific content of the Codes suggests a response to mimetic pressures to conform; it is possible that the use of a highly-formulaic approach is a camouflage strategy within the litigation environment, a deployment of a rational myth. To the extent that a given firm can argue in a court of law that other organizations have successfully deployed the formulaic strategy, the mimetic “sameness” may provide a degree of protective coloration. Or, as Edelman et al. (1999) suggest, the story of how a firm should respond to the law determines both the organization’s actions and the court’s perceptions of the legitimacy of those actions. This may be especially important given the highly litigious environment found in the United States. One direction for future research arises from consideration of the contents of Template 2. It is striking how predominant legal and regulatory, not ethical, items are featured in the Codes. Whether the Code would feature this legal compliance orientation in the absence of regulation is unclear; what is certain is that the disclosure of these items is completely in conformity with the regulatory requirements. Perhaps the temporal proximity of the data (2004) to the enactment of SOX (2002) may have caused more emphasis on financial reporting and recordkeeping issues. A future study could track changes in codes based on external political and economic pressures. For example, in light of the credit and liquidity crisis of 2008 and 2009 did firms adopt risk management clauses in their codes? Further, effective codes could be a function of the corporate culture where management can attempt to integrate the code within the tone at the top (Stevens 2008; Cohen and Hanno 2000). A future study could examine if the
26
effectiveness of the code varies between those companies where the code is integrated within the training and socialization of employees and those companies where the codes are not integrated within the training and socialization process. A related feature of these Codes of Ethics requires further investigation: uniformly, even among the companies that have an image of public-spirit, the Codes revolve around attempts to protect the firm and to extend the power the firm enjoys over its environment. Minimal amount of attention is paid to the ethical obligations of the firm; the question of whether the firm has an ethical obligation does not, in fact, arise – let alone an attempt to address what that obligation might be. Every element of these Codes represents a unilateral attempt to constrain the behavior or employees in ways that may have nothing to do with ethical matters. Employees in exchange give up the right to use their intellectual property as they wish; they give up the right to speak freely during and after employment; they give up the right to privacy of their messages; and in some cases, they give up the right to behave as they wish when not on the job. And in exchange – what do they receive? The promise of an uncertain paycheck, with obligations – but not benefits – that extend beyond the term of employment. Several companies address this directly, offering reminders that nothing in their Code should be construed as a promise of employment and that employment is at-will. Future studies should carefully consider the extent to which these Codes are used as a means to stretch the boundaries of the firm and to expand the power of the organization into arenas not traditionally considered the province of the employer. In this study, the focus is on the observable output of internal organizational processes, and the results suggest that these outputs are consistent with processes theorized within the institutional literature. However, there is no direct examination of the organizational processes that give rise to these outputs. Several of the sample firms were contacted with questions about how the Codes were developed, and either did not respond to inquiries or indicated that this information is not available to the public. This study also does not yield information on the means by which the isomorphic information is transmitted between organizations, although other research (Brivot 2008) and the widespread dissemination of these materials over the internet suggest attractive possibilities. It is of general interest to determine the routes of communication between organizations and the growing role of the professional service firm in facilitating knowledge transfer of the type required in order to generate isomorphic behavior; future research could focus on the evident dissemination of Codes of Ethics to determine how this information is transferred. Further, recent research in neoinstitutional theory (Greenwood and Suddaby 2006) suggests that multidisciplinary professional practice firms such as elite accounting firms often initiate changes that may be emulated by others. A future study could examine if certain types of firms took some leadership in developing a type of code and what led others to emulate that particular code. The difficulty with the rational response observed is that the regulators provided an entirely different set of expectations with respect to management behavior. Given managers’ known propensity for box-ticking and formulaic responses to reporting requirements involving a checklist, one must ask whether the regulators were aware of the likely outcome of the regulation, and if not, why not. And if they were aware of the likely outcome, what then is the purpose of the regulation – other than to provide the appearance of taking action in a crisis? The difficulty with the current situation lies in part with managers, who respond to institutional pressures by engaging in herd
27
behaviors, and in part with regulators, who disregard historical evidence that providing managers with a checklist is akin to painting a bulls-eye on an archery target. Once the checklist has been delivered, the result is a narrow focus on the list items to the exclusion of other factors. The losers in this situation appear to be the investing public and other stakeholders of the firm, who may take comfort in the knowledge that the management of their company has written a Code of Ethics, without arriving at the awareness of the limitations of the content of their Code. Moreover, even with the emphasis on financial matters in the code it appears that the codes may be overly concerned with the investment community at the expense of other stakeholders. The narrowness of the code suggests that we have lost an opportunity to have an external signal that companies should go beyond narrow self-interest and also be concerned with a wide array of external stakeholders.
28
REFERENCES Baginski, S., J. Hassell, and M. Kimbrough. 2002. The effect of legal environment on preemptive disclosure: Evidence from management earnings forecasts issued in U.S. and Canadian markets. The Accounting Review 77(January): 25-50. Becker, L. C. and Becker, C. B. eds. (2001). Encyclopedia of Ethics. Garland Publishing: New York. Bealing, W. E. Jr. and R. L. Baker. 2006. The Sarbanes-Oxley Act: Have we seen it all before? The Journal of Business and Economic Studies 12(2): 1-12. Brivot, M. 2008. Exploring the control dimension of knowledge management systems and its implications for the “ethos of partnership” in professional service firms. Presented at the Third Alternative Perspectives in Accounting Research Conference at Universite Laval, Quebec (April 11). Campbell, D. J. 2000. Legitimacy theory or managerial reality construction? Corporate social disclosure in Marks and Spencer Plc corporate reports, 1969-1997. Accounting Forum 24(1): 80-100 Cohen, J. and D. Hanno. 2000. Auditors’ consideration of corporate governance and management control philosophy in audit planning and client acceptance decisions. Auditing: A Journal of Practice and Theory (Fall): 133-146. Cohen, J. and L. Holder-Webb. 2006. Rethinking the influence of agency theory on the accounting academy. Issues in Accounting Education (February): 17-30. Cohen, J., G. Krishnamoorthy, and A. Wright. 2008. Form versus substance: The implications for audit practice and research of alternative perspectives of corporate governance. Auditing: A Journal of Practice and Theory (November): 181-198. Crow, K. S. 2003. Codes of ethics, complaint channels, and whistleblowers – Applying the new rules. Corporate Counsel (May): A7-A8. Deegan, C. and B. Gordon. 1996. A study of the environmental disclosure practices of Australian corporations. Accounting and Business Research 26(3): 187-199. Deloitte & Touche. 2002. Guidelines for Writing a Code of Ethics/Conduct. (February). Deloitte & Touche Corporate Governance Services. DiMaggio, P. J. and W. W. Powell. 1983. The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review 48 (April): 147-160. DiMaggio, P. J. and W. W. Powell. 1991. Introduction. In W. W. Powell, & P. J. DiMaggio (Eds.), The new institutionalism in organizational analysis (pp. 1-38). Chicago, Il: The University of Chicago Press. Edelman, L. 1992. Legal ambiguity and symbolic structures: Organizational mediation of Civil Rights Law. The American Journal of Sociology 97(6): 1531-1576. Edelman, L., C. Uggen, and H. S. Erlanger. 1999. The endogeneity of legal regulation: Grievance procedures as rational myth. The American Journal of Sociology, 105(2): 406-454. Elsbach, K. D. 1994. Managing organizational legitimacy in the California cattle industry: The construction and effectiveness of verbal accounts. Administrative Science Quarterly 39(March): 57-88. Food and Drug Administration (FDA). 1996. FDA Export Reform and Enhancement Act of 1996. Public Law 104134. Forster, M., T. Loughran, and B. McDonald. 2009. Commonality in codes of ethics. Journal of Business Ethics 90: 129-139. Gray, R., R. Kouhy, and S. Lavers. 1995. Corporate social and environmental reporting: A review of the literature and a longitudinal study of UK disclosure. Accounting, Auditing & Accountability Journal 8(2): 47-77. Greenwood, R. and C. R. Hinings. 1996. Understanding radical organizational change: Bringing together the old and the new institutionalism. Academy of Management Review 21(4): 1022-1054. Greenwood, R. and Suddaby R. 2006. Institutional entrepreneurship in mature fields. The big five accounting firms. Academy of Management Journal 49 (1): 29-47. Hampel, Sir R. 1998. Committee on Corporate Governance, Final Report. London: Gee Publishing. Handelman, J. M. and S. J. Arnold. 1999. The role of marketing actions with a social dimension: Appeals to the institutional environment. Journal of Marketing 63(3): 33-48. Harvard Law Review. 2003. The good, the bad, and their corporate codes of ethics: Enron, Sarbanes-Oxley, and the problems with legislating good behavior. Harvard Law Review 116(7): 2123-2141. Haynes and Boone. 2003. SEC Adopts Code of Ethics Disclosure Rules. (January 24). Haynes and Boone LLP: Dallas. Holmes Roberts & Owen. 2003. Final Rules: Code of Ethics. (February 7). Holmes Roberts & Owen LLP: Boulder, CO. Johnson, R. A. and Bhattacharya, G. K. 2006. Statistics: Principles and Methods. Hoboken, NJ: John Wiley & Sons.
29
Kallio, T. J. 2007. Taboos in corporate social responsibility discourse. Journal of Business Ethics 74: 165-175. Kerkow, U., J. Martens, and T. Schmitt. 2003. The limits of voluntarism: Corporate self-regulation, multistakeholder initiatives and the role of civil society. World Economy, Ecology and Development Association. Kostova, T., Roth, K., and M. T. Dacin. 2008. Institutional theory in the study of multinational corporations: A critique and new directions. Academy of Management Review 33(4): 994-1006. Krawiec, K. 2003. Cosmetic compliance and the failure of negotiated governance. Washington University Law Quarterly 81: 487-544. Krawiec, K. 2005. Organizational misconduct: Beyond the principal-agent model. Florida State University Law Review 32: 571-592. Lindsay, R. M., L. M. Lindsay, and V. B. Irvine. 1996. Instilling ethical behavior in organizations: A survey of Canadian companies. Journal of Business Ethics 15: 393-407. Locke Liddell & Sapp. 2003. Corporate Governance Alert: SEC Adopts Final Code of Ethics Rules. Locke Liddell & Sapp LLC. Long and Driscoll. 2008. Codes of ethics and the pursuit of organizational legitimacy: Theoretical and empirical contributions. Journal of Business Ethics 77: 173-189. Macey, J. 2007. What Sarbox wrought. The Wall Street Journal (April 7): A9. Meyer, J. W. and B. Rowan. 1977. Institutionalized organizations: Formal structure as myth and ceremony. American Journal of Sociology 83(2): 340-363. Milne, M. J. and R. W. Adler. 1999. Exploring the reliability of social and environmental disclosures content analysis. Accounting, Auditing & Accountability Journal 12(2): 237-256. Murphy, P. E. 2005. Developing, communicating and promoting corporate ethics statements: A longitudinal analysis. Journal of Business Ethics 62: 183-189. MyDropbox. 2009. http://www.mydropbox.com/. Blackboard, Inc. NASDAQ. 2003. SEC Approves NASDAQ Corporate Governance Rules. (November 3) New York: NASDAQ. NYSE. 2004. Listed Company Manual. New York: NYSE. OECD. 2001. Codes of Conduct: Exploring their Economic Significance (OECD, Paris). Pfeffer, J. and G. Salancik. 1978. The External Control of Organizations: A Resource Dependence Perspective. New York: Harper and Row. Sarbanes, P., and M. Oxley (SOX). 2002. Sarbanes-Oxley Act of 2002. Washington, DC: U.S. Congress. Schaeffer, R. L., Mendenhall, W. and L. Ott. 1990. Elementary Survey Sampling. Boston: PWS-Kent. Securities Exchange Commission (2003). 17 CFR Parts 228, 229, and 249. [Release Nos. 33-8177; 34-47235; File No. S7-40-02]. RIN 3235-AI66. Disclosure Required by Sections 406 and 407 of the Sarbanes-Oxley Act of 2002. January 24, 2003. Seetharaman, A., F. Gul, and S. Lynn. 2002. Litigation risk and audit fees: Evidence from U.K. firms cross-listed on U.S. markets. Journal of Accounting and Economics 33(February): 91-115. Sims, R. R., and J. Brinkman. 2003. Enron ethics(Or: Culture matters more than codes). Journal of Business Ethics 45: 243-256. Stevens, B. 2008. Corporate ethical codes: Effective instruments for influencing behavior. Journal of Business Ethics 78: 601-609. Suchman, M. C. 1995. Managing legitimacy: Strategic and institutional approaches. Academy of Management Review 20(3): 571-610. Supreme Court of the United States. 2009. Free Enterprise Fund et al. v. Public Company Accounting Oversight Board et al. No. 08-861, argued December 7 2009, decided June 28, 2010. Toedt, D. C. III. 1997. Law and Business of Computer Software. Clark Boardman Callaghan. Verschoor, C. C. 2002. New governance initiatives have ethics component. Strategic Finance 84(5): 22-23. Wee, H. 2002. Corporate ethics: Might makes right. Business Week (April 11).
30
TABLE 1. DESCRIPTIVE STATISTICS FOR THE SAMPLE FIRMS Panel A. Entire Sample Total Assets* Sales*
Mean 5650.53 5094.25
Median 107.86 72.03
Minimum
Maximum 123684.00 56434.00
0.61 0.86
Panel B. By Industry Industry OIL PHARMA SOFTWARE GROCERY MFG
Total Assets* Mean Median 5253.01 566.47 17736.16 49.59 140.43 32.39 5721.65 2618.89 1909.54 34.37
Sales* Mean 2071.13 10004.33 66.83 16029.21 1365.17
Median 172.07 15.89 25.70 10632.85 32.69
Panel C. By Aggregate Size Quintile** Total Assets* Sales* Quintile Mean Median Mean Median 1 6.96 7.12 7.89 3.96 2 39.90 36.29 22.65 23.35 3 141.43 133.93 86.38 73.99 4 1219.02 933.89 3100.56 1048.16 5 27871.71 13577.40 23180.35 11368.00 * in millions **Aggregate size quintiles identified by ranking the entire sample by total assets in order from smallest to largest and dividing into quintiles.
31
TABLE 2. ETHICS-ORIENTED RHETORIC EMBEDDED IN CODES OF ETHICS Text from Code “This Code is intended to deter wrongdoing and to promote the conduct of all Company business in accordance with high standards of integrity and in compliance with all applicable laws and regulations.” “[The Company] is, and since its inception, has been committed to the highest standards of honesty ethics and integrity in its business dealings.” “[The Company] is committed to the highest standards of legal and ethical business conduct…” “[The Company] is dedicated to conducting its business consistent with the highest standards of business ethics.” “Each employee of the Company is responsible for conducting the Company's business in a manner that demonstrates a commitment to the highest standards of integrity.” “This Code of Business Conduct and Ethics (the ``Code'' ) is intended to be a guide for applying legal and ethical practices to your everyday work and to explain the types of behavior that will help our Company meets its commitment to operate on the highest standards of ethical conduct.” “This policy demands adherence to the highest standard of business ethics and conduct.” “[The Company] … is committed to the highest standards of business and ethical conduct.” “…all employees are expected to conduct business according to the highest ethical standards of conduct.” “…we should each be personally committed to demonstrating the highest standards of ethical business conduct.” “Our Company is built upon the highest standards of business integrity and corporate morality.” “[The Company] has established a strong reputation for integrity in our business. To maintain and enhance that reputation, it is important for each of us to adhere to the highest moral ethical and legal standards.” *Represents the number of Codes from the 75 sample firms that used the exact, verbatim text.
No. of Sample Firms Using Text* 3
3 2 2 2 2
2 1 1 1 1 1
32
TABLE 3. CODE LENGTH DESCRIPTIVE STATISTICS Panel A. Word Count Descriptive Statistics Mean Median 4120.25 3461.00 Panel B. Code Length in Words, by Industry Industry Mean OIL 3812.50 PHARMA 4550.90 SOFTWARE 2942.82 GROCERY 5191.00 MFG 4300.33
Minimum 261
Median 3629.50 4203.50 3265.00 4779.00 2999.00
Maximum 11180
Minimum
Maximum 616 261 506 282 1136
7998 10825 4456 9011 11180
Panel C. Code Length in Words, by Aggregate Size Quintile** Quintile Mean Median Minimum Maximum 1 2426.50 2149.50 506 6088 2 3825.00 3484.00 1397 7339 3 2964.50 3313.00 282 4604 4 5098.64 4200.00 261 9011 5 6134.91 4779.00 416 11180 *Aggregate size quintiles identified by ranking the entire sample by total assets in order from smallest to largest and dividing into quintiles.
33
TABLE 4. BOILERPLATE DESCRIPTIVE STATISTICS FOR ENTIRE CODE* Panel A. Percentage of Code Comprised of Boilerplate, Entire Sample Mean Median Minimum 90.60 98.10 26.92 Panel D. Percentage of Code Comprised of Boilerplate, by Industry Industry Mean Median Minimum OIL 96.48 98.09 82.94 PHARMA 90.90 96.61 43.97 SOFTWARE 97.94 100.00 85.93 GROCERY 81.40 81.18 60.76 MFG 85.82 100.00 26.92
Maximum 100.00
Maximum 100.00 100.00 100.00 100.00 100.00
Panel C. Percentage of Code Comprised of Boilerplate, by Aggregate Size Quintile** Quintile Mean Median Minimum Maximum 1 97.32 100.00 85.93 100.00 2 94.72 99.40 62.59 100.00 3 97.96 98.33 90.62 100.00 4 86.99 94.58 60.76 100.00 5 75.77 86.91 26.92 100.00 *Boilerplate is defined as unoriginal material. Calculated as the number of words comprising unoriginal content, as identified by the analysis software, divided by the total number of words in the entire document
34
TABLE 5. BOILERPLATE DESCRIPTIVE STATISTICS BY MAJOR CODE SECTION
Section Conflicts of Interest Gifts Corporate Opportunities Protection and Proper Use of Company Assets Record-Keeping and Reporting Confidentiality Competition and Fair Dealing Compliance with Laws Compliance with Code and Reporting Violations Bribery Insider Trading Diversity and Human Resources Health, Safety, and Environment Waivers
Length of Section (in Words) Mean Median 511.70 436 185.88 108 69.23 62 165.77 128
Unoriginal Text as a Percentage of Section Length Mean Median 97% 100% 93% 100% 100% 100% 96% 100%
285.28 168.98 224.58 133.30 208.07
238 111 120 98 141
97% 97% 91% 96% 93%
100% 100% 100% 100% 100%
209.16 182.70 216.49 95.47 71.96
131 130 65 0 59
94% 96% 92% 93% 100%
100% 100% 100% 100% 100%
35
TABLE 6. SAMPLE BOILERPLATE CONTENT Company 1 The purpose of business entertainment and gifts in a commercial setting is to create good will and sound working relationships, not to gain unfair advantage with customers. No gift or entertainment should ever be offered, given, provided or accepted by any Company employee, family member of an employee or agent unless it (1) is not a cash gift, (2) is consistent with customary business practices, (3) is not excessive in value, (4) cannot be construed as a bribe or payoff and (5) does not violate any laws or regulations. Please discuss with your supervisor any gifts or proposed gifts which you are not certain are appropriate. Company 2 The purpose of business entertainment and gifts in a commercial setting is to create good will and sound working relationships not to gain unfair advantage with customers. No gift or entertainment should ever be offered, given, provided, or accepted by, an employee unless it: is not a cash gift, is consistent with customary business practices, is not excessive in value, cannot be construed as a bribe or payoff, and does not violate any laws or regulations. Employees should discuss with their managers or a Human Resources representative any gifts or proposed gifts if they are not certain that such gifts are appropriate. Company 3 The purpose of business entertainment and gifts in a commercial setting is to create good will and sound working relationships, not to gain unfair advantage with customers or suppliers. No gift or entertainment should ever be offered, given, provided or accepted by any Employee, family member of an Employee unless it: (1) is not a cash gift, (2) is consistent with customary business practices, (3) is not excessive in value, (4) cannot be construed as a bribe or payoff and (5) does not violate any laws or regulations. Supervisors can advise on the appropriateness of any gifts or proposed gifts. Company 4 The purpose of business entertainment and gifts in a commercial setting is to create good will and sound working relationships, not to gain an unfair advantage with customers. No gift or entertainment should ever be offered, given, provided or accepted by any Company employee, officer, director, family member of an employee, officer, director or agent unless it: (1) is not a cash gift, (2) is consistent with customary business practices, (3) is not excessive in value, (4) cannot be construed as a bribe or payoff and (5) does not violate any laws or regulations. Please discuss with your supervisor any gifts or proposed gifts that you are not certain are appropriate.
36
APPENDIX 1. CONTENT OF TEMPLATE 1 CODE OF ETHICS (N=9)* In my role as a _____________________________ of the Company, I recognize that I hold an important and elevated role in corporate governance. I am uniquely capable and empowered to ensure that stakeholders' interests are appropriately balanced, protected and preserved. Accordingly, this Code provides principles to which I am expected to adhere and advocate. The Code embodies rules regarding individual and peer responsibilities, as well as responsibilities to the company, the public and other stakeholders. I certify to you that I adhere to and advocate the following principles and responsibilities governing my professional and ethical conduct. To the best of my knowledge and ability: 1. I act with honesty and integrity, avoiding actual or apparent conflicts of interest in personal and professional relationships. 2. I provide constituents with information that is accurate, complete, objective, relevant, timely and understandable. 3. I comply with rules and regulations of federal, state, provincial and local governments, and other appropriate private and public regulatory agencies. 4. I act in good faith, responsibly, with due care, competence and diligence, without misrepresenting material facts or allowing my independent judgment to be subordinated. 5. I respect the confidentiality of information acquired in the course of my work except when authorized or otherwise legally obligated to disclose. Confidential information acquired in the course of my work is not used for personal advantage. 6. I share knowledge and maintain skills important and relevant to my constituents' needs. 7. I proactively promote ethical behavior as a responsible partner among peers in my work environment and community. 8. I achieve responsible use of and control over all assets and resources employed or entrusted to me. *This example represents a verbatim representation of the entire Code of Ethics for one of the sample firms.
37
APPENDIX 2. TEMPLATE 2 CODE OF ETHICS* (N=57) Introduction (n=55) Promotion of ethical behavior Conflicts of Interest (n=57) Definition or general description of conflicts Prohibition or recommendation to avoid and obligation to disclose Gifts, Gratuities, and Entertainment (n=37) o Restriction on size of gift that may be given or received o Statement that the purpose is to “create good will and sound working relationships” o Does not create or appear to create an obligation Corporate Opportunities (n=57) Prohibition against personally taking opportunities discovered through work for firm Prohibition against using company resources for personal gain Prohibition against competing, directly or indirectly, with company Statement of duty to advance the company’s legitimate business interests Protection and Proper Use of Company Assets (n=57) Statement that employees have an obligation to protect company assets Restriction of the use of company property to company business Caution against “theft, carelessness and waste” of assets Reminder that proprietary information is included in the protected assets Record-Keeping, Reporting, and Disclosure (n=57) o Commitment to providing “full, fair, accurate, timely, and understandable” disclosure o Requirement to maintain books that accurately represent transactions/prohibition against creating misleading accounts or causing accounts or books to become misleading o Statement of commitment to honest and accurate reporting o Requirement to comply with Generally Accepted Accounting Principles (GAAP) Confidentiality (n=54) Obligation to maintain confidentiality General definition of “confidential” information Competition and Fair Dealing (n=55) Instruction to “deal fairly” with stakeholders Prohibition against “taking unfair advantage” of stakeholders Compliance with Laws, Rules, and Regulations (n=57) Statement that applicable laws, rules, and regulations must be followed by all employees Compliance with Code and Reporting Violations (n=56) Statement of non-retaliation for good-faith reports/provision for anonymous reporting Requirement for employees to report violations or suspected violations Anti-Bribery Provisions (n=37) Statement of compliance with Foreign Corrupt Practices Act Insider Trading (n=42) Prohibition against using material, nonpublic information to make stock trades Reference to separately published insider trading policy Prohibition against providing “tips” to others Employment and Human Resources (n=35) Prohibition of harassment EEO statement Assertion that employee diversity is “valued asset” Health, Safety, and Environmental (HSE) (n=25) General commitment to “safe and healthy” work environment Assignment of responsibility for providing same to employees Prohibition against substance use on the job *The number of firms using Template 2 is 57. Specific items within each category represent those made by at least 50% of Template 2 firms.
38
1
Key provisions of the Sarbanes-Oxley Act have recently been upheld by the U.S. Supreme Court
(Supreme Court of the United States, 2009) 2
An alternative explanation is that they may have anticipated the widespread box-ticking, but did not
anticipate that the investing public would become aware of the empty compliance; implicit in this argument would be a supposition that investors would be unlikely to compare Codes between firms or possibly even to read the Codes, simply inferring information from the fact that a Code had been provided. 3
Lindsay, et al. (1996) studied a broad range of mechanisms for promoting and enforcing ethical conduct
within organizations, and concluded that the options for promotion and enforcement are wide-ranging and that most organizations pursue this end goal through multiple means. However, they also concluded that the use of a Code is “the dominant plank” in the firm’s effort to ensure ethical conduct. Despite this, they found evidence of limited use of mechanisms to enforce the Code or to reward ethical behavior, and limited efforts to communicate the Codes and educate employees about them. In addition, half of their firms had no mechanism through which the Code was to be periodically reviewed and revised. The authors conclude that there is reason to suspect that many companies may be engaging in empty gestures, or paying lip service, in promoting ethical conduct. 4
This study focuses only on an observable output that yields information about the behavior across the
entire institutional field, rather than on the processes used by any given organization to develop a specific Code or the specific means by which the isomorphic content is transmitted. The examination of these factors is of practical and academic interest and is addressed in the final section of the paper. 5
DiMaggio and Powell (1983) indicate that members of an organizational field possess characteristics of
connectedness and structural equivalence. Publicly-traded firms in the U.S. are collectively subject to a common set of regulations and are soliciting an identical resource (financial capital) from a collective pool of providers (equity investors); these elements tie the organizations together in common pursuits under common institutions, and demonstrate a degree of connectedness. They also possesses structural equivalence in that they have a collective obligation to adhere to government regulation (the same kind of ties to the same set of regulatory agencies).Therefore, for the purposes of this study, the population of publicly-traded firms in the U.S. is defined as an organizational field. 6
Isomorphic behavior refers to convergence in behavior across organizations, where one organization
begins to resemble another closely. Coercive forces are those that involve the exertion of pressure to converge behaviors that arises from forces that have some type of control over the organization. Normative forces are those that involve the societal development of desirable behaviors, upon which organizations then converge. Mimetic forces are those involving mimicry and herd behavior (imitation without planned direction). 7
In January 2006, the Ethics Officer Association changed its name to Ethics and Compliance Officers
Association to reflect the growing emphasis on compliance with ethics and conduct standards. 8
Codes were provided by some, but not many, firms prior to the advent of the regulation discussed above.
However, it was only in the wake of the decision to frame the legitimacy crisis in terms of ethical failings that a culture that promotes ethical behavior and social responsibility began to take hold and gain momentum.
39
Developments in subsequent years yielded increased emphasis on these factors through the social responsibility reporting movement, triple-bottom-line accounting, and an increase in awareness of ethical issues via the popular media and organizational training. These developments are likely to have fostered the generation of normative isomorphic pressures that may have subsequently come into play with respect to corporate ethics; however, the key focus in the realms in which normative forces are developed for the business world (MBA programs, undergraduate business education, and corporate training) did not have a strong ethics- or responsibility-orientation in the early years of the 21st century. We focus, therefore, on the two isomorphic forces known to be in effect at that time: coercive pressures arising from the new government regulation, and mimetic forces that arise in situations involving a high degree of uncertainty. 9
Survey research indicates that prior to 2002, 94 out of the 100 largest trans-national corporations had
published codes of conduct (OECD, 2001). Another study conducted prior to the issuance of SOX found that more than 500 companies in the USA adhered to some kind of codes of conduct (Kerkow et al., 2003). These studies suggest that codes were in wide but not universal use prior to the legal mandate to provide them. One potential reason for providing a code before SOX came from an earlier Congressional initiative that made having proof of an effective code of ethics a mitigating factor under the Federal Sentencing Guidelines for Organizations (HLR, 2003). However, the notion of “effectiveness” was not well-defined. 10
Prior to the passage of the Sarbanes-Oxley Act of 2002, firms were permitted to develop and provide
codes of ethics at will. The leniency under the Federal Sentencing Guidelines for Organizations available to companies who possessed codes argues for some coercive element; however, the leniency was contingent upon proof of effectiveness, a nebulous quality that promotes the sort of uncertainty that generates mimetic pressures. 11
Decoupling refers to the deliberate manufacture of discontinuities or gaps between actual organizational
practice and the formal policies about those practices (Meyer and Rowan, 1977). 12
One related paper to ours that we were not aware of when we designed and developed the study was
conducted by Forster et al. (2009). They studied codes of conduct of S&P 500 firms during the first half of 2008 as well as a sample of small firms to compare the effect of size on convergence of codes. Their study was not motivated by a particular theory and was primarily data driven. They found a tremendous amount of convergence in the codes which post hoc they attribute to isomorphism. We complement that study by examining codes closer to the adoption of SOX (codes in 2004) and by a priori using institutional theory to motivate the research questions as well as to explain the results. Further, we make extensive use of the codes themselves to illustrate how the data supports the theory. Thus, the use of a theory-driven paper with examples from the codes as well as using a sample closer in temporal proximity to the enactment of SOX adds to the value we bring to the burgeoning literature on codes of ethics in the US and in the world. 13
See Seetharaman et al. 2002, Baginski et al. 2002, and Macey 2007 for a discussion of the relative
litigation risk of the U.S.
40
14
These entities operate under a substantially different regulatory regime and may or may not have active
operations, and therefore represent a population that may be subject to different institutional pressures than those faced by operating entities. 15
This discussion is not intended to indicate that only these three industry sectors have industry- or firm-
specific matters to cover with their Codes of Ethics, but to note that the sample yields the express potential for this type of unique Code content to be provided. 16
The identification of quintiles and position within quintiles was qualitatively unaffected by the choice of
variable (total assets or sales) as proxy for firm size. 17
The Mydropbox technology has since been acquired by Blackboard Inc. and integrated into the
SafeAssign product. It is now available as part of Blackboard Enterprise edition. 18
A source who spoke under conditions of anonymity shed light on the construction of the Code at her
organization. When the Code became mandatory, the instruction given by the CEO was to “go and find out what others are doing, and then put something together that looks like that.” This pragmatic example yields the conclusion that for isomorphic firms, the content and language may well be co-determined. 19
See Edelman et al. (2001) for a discussion of the institutional pressures generated by equal-opportunity
regulation, the accompanying development of diversity rhetoric, and how these influence managerial behavior. 20
See Kallio (2007) for a discussion of social responsibility rhetoric and associated matters.
21
Based on the results of chi-square goodness-of-fit tests, not reported separately; no test attained statistical
significance at any conventional level. 22
No further information was provided in this Code as to what matters must be adhered to and which are
open for negotiation. 23
Of the excerpts provided in the previous section on content convergence, the only one that represents
original material is the one about Enron. All other excerpts are entirely unoriginal. 24
This theory is not posited as an alternative to the mimetic isomorphism argument; the two theories lead to
the same conclusion (use of boilerplate). Edelman et al.’s (1999) theory spells out the details whereby mimetic behavior is also economically rational. 25
We thank an anonymous reviewer for the observation that in many organizations, the legal department
oversees the development of the Codes; it is normal and acceptable behavior in the legal profession to take advantage of document repositories that provide uniform wording designed to standardize the discussion of various legal matters. Thus, the heavy involvement of lawyers in the construction of these Codes may explain the significant volumes of unoriginal content observed within the sample firms. However, this explanation does not materially change the conclusion that the Codes are unlikely to be intended to constrain behavior, nor does it suggest that the construction and provision of these boilerplate Codes is not functionally a mimetic response rather than an attempt to develop a distinct ethical climate for employees of a given company.
