تأثیر اخلاق در موفقیت بازار کار : مدارک و شواهد از MBA ها

This paper empirically investigates the link between ethics, earnings and gender. Using a self-reported measure from a longitudinal survey of registrants for the Graduate Management Admission Test, we find that ethical character is negatively associated with males’ wages. For females, however, this relationship does not hold. In addition, using measures of the degree to which ethics is emphasized in business school curricula as an indicator for enhancement of individual ethical standards of graduates, we investigate variation in the returns to an MBA degree. We find that the larger the degree to which males report that business education enhanced their ethical character, the lower their wages, holding other aspects of their education constant. For females, however, enhanced ethics through business school is positively and significantly associated with returns to the MBA degree. More objective measures of ethics emphasis in business school curricula provide further support of these findings.

A considerable number of studies in the social sciences have focused on identifying differences in behavior and attitudes between men and women. Ethical behavior and reasoning is one area in which substantial differences have been found. In particular, many studies have pointed towards the conclusion that men are more selfish than women, who are relatively more socially conscious in their thinking and decision making. For example, women have been shown to take stronger ethical stances (Glover et al., 1997 and Reiss and Mitra, 1998), score higher on ‘integrity tests’ (Ones and Viswesvaran, 1998) and make more ethical decisions in business (Loe et al., 2000). They have also been shown to be more prone to helping others (Eagly and Crowley, 1986), to vote based on social issues (Goertzel, 1983), and their representation in government has been linked to lower levels of corruption (Dollar et al., 2001). In addition, gender has been shown to have an important effect on behavior and payoffs in several different experimental economic settings related to selfish versus selfless decision making.1 In ultimatum games, for example, women tend to be more cooperative, making more generous proposals and more likely accepting an offer of a given amount (Eckel and Grossman, 2001). Women also have a higher probability of cooperating in prisoner's dilemma games (Frank et al., 1993), and tend to be more generous and have an affinity towards fairness in dictator games (Eckel and Grossman, 1996, Eckel and Grossman, 1998 and Andreoni and Vesterlund, 2001). Women are also more cooperative in public goods games (Nowell and Tinkler, 1994 and Seguino et al., 1996), and are less trusting but more trustworthy than men in investment games (Buchan et al., 2004).2 Of course, differences in behavior in these games normally translate into differences in individual payoffs. Depending on the structure of the game and the composition and behavior of other participants, more socially minded behavior may be associated with higher or lower individual payoffs. Specifically, one's payoff in such games appears to not only be due to one's behavior, but also a function of perceptions or stereotypes of gender groups (Eckel, 2008). To the extent that more ethical behavior may imply increased cooperation, coordination or altruism, and to the extent that women (relative to men) are expected to exhibit these behaviors, they may receive a penalty for deviating from them. On the other hand, if engaging in non-socially efficient behavior is more acceptable for men, they may be more likely to engage in lucrative self-promoting behaviors. A primary goal of our study is to see whether these differences in attitudes, behavior and performance in experimental settings are borne out in differences in success in real-world labor market data. Specifically, we ask the questions: how do individuals’ ethical standards influence their earnings? And, does the magnitude or direction of the effect vary by gender? To our knowledge, no previous research has addressed these questions. Nonetheless, there exist reasons to believe that ethics related behavior could affect one's earnings (either positively or negatively) on the job. First, as mentioned above, experimental evidence directly links socially minded behavior to earnings, and suggests likely differences in the reward or punishment of ethical behavior across gender. Second, over the last few decades, firms have devoted increasing attention to ethical practices. Even prior to the 1980s, many businesses and corporations began to adopt codes of ethics (White and Montgomery, 1980). Government agencies have done the same (Hays and Gleissner, 1981). In addition, some professional associations have their own codes, such as the American Institute of Certified Public Accountant's (AICPA) Code of Professional Conduct.3 In accordance with this emphasis at the firm level, one might expect that employers who behave ethically will be rewarded for doing so. On the other hand, any emphasis a firm places on business ethics may just be “window-dressing”, and firms may actually reward unethical behavior, either directly or indirectly (White and Montgomery, 1980). Indeed, empirical studies of the effects of ethics codes on ethical behavior or unethical incidents provide mixed results (see Loe et al., 2000, for a review). Further, even if firms do desire ethical behavior, monitoring the ethical decisions of employees may be difficult. If only outcomes are observed, individual employees may find it optimal to engage in unethical activity on the job if employers indirectly reward such activity with increased likelihood of promotion or higher pay. A likely reason for the dearth of empirical studies addressing these questions is that characterizing the true ethical standards of individuals, or observing their relevant behavior in the workplace, is a difficult or impossible task. Further, the problem of confounding factors affecting earnings is likely to bias the estimates of interest. Ideally, after determining a specific definition of what constitutes ethical versus unethical behavior, we would observe specific incidents of socially commendable behavior or ethical misconduct on the job. Rather, in this paper, we use unique survey data and attempt to identify several different variables which likely serve as a proxy for an individual's commitment to ethics in the workplace. By including a wealth of control variables and in some cases individual fixed effects, our hope is to minimize any resulting bias that may come from the use of imperfect proxies or the endogeneity of survey responses used in this study. While the effect of ethics on individual earnings has not been investigated previously, there is a large body of literature pertaining to the effect of firm-level ethical character on performance. The most common approach has been to analyze the relationship between some general measure or non-quantitative rating of “social performance” (such as monetary contributions to the community, charities, and companies singled out in the press) on financial performance (usually measured by changes in stock price, return on equity, or return on assets). This research has not reached any convincing conclusion regarding even the direction of the effect, although many studies do find that firms that are more socially responsible perform as well as, or even better than, firms that are less socially responsible (see Griffin and Mahon, 1997 and Roman et al., 1999 for reviews, and Orlitzky et al., 2003 for a meta-analysis). Our paper can be seen as a complementary, individual-level analog of this body of research. We utilize panel data from surveys of individuals who registered to take the Graduate Management Admission Test (GMAT), an exam required by most MBA programs for admission. Individuals were subsequently surveyed in four waves over a span of about eight years, whether or not they ultimately obtained an MBA. There are several reasons why MBAs are a particularly appealing group for use in a study of business ethics. This group of people is most likely to consist of future business leaders, the people for which it is very important to maintain high ethical standards. Indeed, the ethical behavior of employees is often modeled after the behavior of managers (Stead et al., 1990). Furthermore, as the sample includes only those who have demonstrated some degree of interest in obtaining an MBA, it represents a relatively homogeneous group of men and women in terms of prior accumulated human capital and commitment to their careers. More importantly, the survey data include individual earnings, work experience, and information regarding self-reported ethical character, as well as self-perceived gains in ethical character through business school, allowing us to empirically investigate the relationship between earnings and ethical character. By linking this individual-level data to information regarding characteristics of business schools, we are also able to investigate the effect on earnings of MBA attainment from programs differing in their emphasis on ethics. Finally, a further advantage of analyzing the returns to an MBA is that it allows us to exploit the fact that most MBA students work full-time prior to obtaining an MBA.4 The presence of both pre- and post-MBA earnings observations allows us include individual fixed effects in our regressions as a method of controlling for unobserved heterogeneity in our sample.5 Career paths of MBAs (as well as those who considered but did not attend MBA programs) are likely to differ by gender. While Arcidiacono et al. (2008) find only small gender differences in estimates of short-term returns to a broad class of MBAs, other studies cite large earnings differentials of graduates. Bertrand et al. (2010), for example, cite the gender earnings differential of University of Chicago MBAs to be as high as 60% over 10 years after graduation. A large part of this differential can be attributed to preferences over work versus family, a finding that is echoed in analyses of Harvard graduates (Goldin and Katz, 2008 and Herr and Wolfram, 2009). However, a recent and growing body of research attributes some gender differences in earnings and career paths to variation in noncognitive skills or characteristics, like motivation, assertiveness, or reactions to competition (Booth, 2009, Braackmann, 2009, Fortin, 2008, Grove et al., 2011, Long, 1995, Mueller and Plug, 2006 and Thiel and Thomsen, 2009). With this gender heterogeneity in mind, the types of jobs or work environments in which women are likely to thrive are likely to differ from those for men. Furthermore, their career related goals may be different, as men are more likely to place higher priority on earnings or status than women, who may alternatively place more emphasis on personal development and intrinsic rewards (Powell and Maneiro, 1992, Russo et al., 1991 and Sturges, 1999). Gender differences in the effects of obtaining an MBA degree may be due to desired versus actual skills gained by males and females within the program. The MBA has been found to provide graduates with value added in areas such as information analysis and initiative (Kretovics, 1999), self-esteem and self-confidence (Baruch and Peiperl, 2000), and decision-making (Kakabadse and Kakabadse, 1999), but still may fall short in fostering the development of interpersonal and leadership abilities (Boyatzis and Renio, 1989) and other ‘soft skills’ associated with the actual practice of management (Mintzberg, 2004 and Pfeffer and Fong, 2002). Mirroring their differing priorities, there is some evidence that men and women gain from an MBA in diverging ways. In particular, women gain more than men in terms of intrinsic skills or traits, such as confidence, assertiveness, interpersonal skills, and communication skills (Simpson, 1995, Simpson, 1996, Simpson, 2000 and Simpson et al., 2005). To the degree that MBA programs may differ in their emphasis on alternative skills or areas of development, gender heterogeneity in the returns to an MBA may result. In this paper, we specifically investigate the role of ethics in association with business school curricula in explaining earnings, and allow this relationship to vary by gender. While virtually non-existent a few decades ago, business ethics has in one form or another found its way into the curriculum of most MBA programs. The Association to Advance Collegiate Schools of Business (AACSB) added ethics to its list of required knowledge in 1974. Since the 1990s, however, the removal of a clear AACSB mandate requiring a course in business ethics has facilitated the dropping of ethics courses from the curriculum of some business programs. The result is a wide variety in the way in which and degree to which ethics is incorporated into graduate management education. Some programs incorporate ethics into other courses. Others require specific ethics courses or service-based learning. Still others do very little in the way of incorporating ethics into their curricula (Kelley, 2003). To the extent that men and women may desire different content within an MBA program, MBA programs’ varying emphasis on ethics may shed some additional light on earnings differentials, either because of a direct effect of curriculum on the managerial capabilities of graduates, or because of an indirect effect of MBA programs sorting graduates into varying careers. Empirical analyses of the effects of ethics courses have been limited primarily to inferences based on situation or decision based questionnaires administered before and after the course. There is no clear consensus among the large body of research in this area. Some studies conclude that ethics courses positively affect the values and opinions of students and their ability to recognize ethical issues (Glen, 1992, Green and Weber, 1997, Gautschi and Jones, 1998 and Weber and Glyptis, 2000). Several others conclude that ethics courses or service-based learning have little or no effect on student attitudes (Martin, 1981–1982, Wynd and Mager, 1989 and Conroy and Emerson, 2004). Rather than investigating the effects of ethics education on an individual's values, we go beyond the previous research by looking at the effect of expressed or implied differences or changes in ethical values on an individual's wage. To the extent that higher ethical standards may positively or negatively affect earnings, the monetary return to an MBA may be positively or negatively affect by the degree to which the MBA program emphasizes ethics. Using variation in MBA programs’ emphasis on ethics and the resulting wage increases of their graduates is thus another way to determine, indirectly, the effect of ethics on labor market success. Furthermore, to the degree that males’ and females’ ethics are valued differently in the labor market, this analysis offers a possible additional explanation of male–female post-graduation earnings differentials that has not been addressed in the literature. Indeed, our results differ substantially between males and females. For males, self-reported ethical character is negatively associated with wages (−3.4%). For females, however, this relationship does not hold. Furthermore, males who report that business education greatly enhanced their ethical character saw a significantly lower return to obtaining an MBA (−6.5%), holding other aspects of their education constant. For females, however, enhanced ethics through business school is positively and significantly associated with wages (+5.5%). Using external ratings of MBA programs’ emphasis on ethics resulted in similar findings. To the extent that our measures of ethical character proxy for actual differences in ethics related behavior in the workplace, our results thus corroborate the frequent finding in the experimental economics literature of substantial differences between men and women in social versus individualistic behavior and subsequent performance. The remainder of the paper proceeds as follows. Section 2 describes the data. Our empirical methodology is described in more detail in Section 3. Results are presented in Section 4. Section 5 concludes by offering additional discussion of the results, as well as potential caveats regarding our findings.

Using data from a longitudinal survey of registrants for the Graduate Management Admission Test, this paper has investigated the relationship between a self-reported measure of individual ethical character and job market success as reflected in an individual's earnings. In addition, using several measures of the degree to which ethics is emphasized in business school curricula as an indicator for enhancement of individual ethical standards of students, variation in the returns to an MBA degree was investigated. Results differ substantially between males and females. It seems that on an individual level, having high ethical standards is bad for males’ earnings. For females, however, it may be unimportant or even good. These results have some credibility for a few reasons. First, by looking at potential MBAs only, we consider a relatively homogenous group of people. Unobserved heterogeneity is likely to be less of a problem than it would be in the case of a sample of the general population. Second, a unique data set has allowed for a large number of control variables that are not commonly available to researchers. In particular, we control for spurious tendencies in response to skill possession questions or MBA skill enhancement questions by including an index of general responses to these skills separately from the responses dealing with ethical standards. We also control for other individual attitudes and beliefs, such as the personal importance of religion, in the regressions. Third, the results generally hold up under a number of specifications, including when individual fixed effects are included, and when non-self-reported variation in ethics emphasis across graduate management curricula is used as a proxy for changes in ethical standards of graduates. In addition to potential concerns previously raised, a possible objection to a causal interpretation of the results presented here relates to selection into jobs. Several studies show that the gender wage gap is significantly diminished when one takes into account different preferences by gender over fields of employment, types of jobs, and college major (Blau and Kahn, 1997, Fields and Wolff, 1995 and Loury, 1997). Ethical character could be serving as a similar sorting mechanism. To be sure, the job related controls used in this study have been limited, and our ethics measure may be picking up differences in preferences over jobs. Even still, the ability to obtain certain types of jobs (or promotions or raises) is likely to be one way in which willingness (or unwillingness) to engage in ethical behavior manifests itself. At the very least, our results suggest that raw earnings differentials may exist not only across gender, but simultaneously across gender and ethical character. According to one specification, the wage gap between “unethical” men and women is estimated to be 7.1% (controlling for a rich set of covariates), but this gap is cut in half when comparing “ethical” individuals. As a final caveat to any interpretation of the results presented here, it should be emphasized that this study has necessarily remained agnostic as to the definition of ethical standards. Indeed, much of the evidence presented here has relied on individual self-reported ratings of ethical standards, which should be viewed, at best, as an imperfect proxy for actual behavior in the workplace. In particular, this measure certainly depends on one's perception of ethics. It is possible that what women view as being ethical differs from the views of men. Indeed, in her pioneering work, Gilligan (1982) found that men and women use fundamentally different approaches in the way in which they make decisions about morality. The male approach involves the view that individuals have certain basic rights which must be respected. The female approach involves the view that people have certain responsibilities towards others. Thus, differences in perception, rather than differences in actual behavior, may be driving the results presented here. Of course, differences in individuals’ perceptions of business ethics are also likely to result in differences in actual behavior. According to Gilligan, while the male approach to morality imposes restrictions on what one can do, the moral decisions of females are more likely to be driven by an imperative to care for others. Some of the results presented here may thus be incorporating the combined effect of differences in perception or definition of high ethical standards and actual differences in ethics related behavior. More research is needed in order to zero in on (or rule out) a true causal relationship. In the absence of workers’ behavioral data, however, this study provides a step towards gaining a better understanding of extent to which individuals in business may have financial incentives to act with more or less regard to ethical standards. It also underscores the idea that economically relevant behavioral and perceptional differences between men and women exist outside of the laboratory, even in a relatively homogeneous sample of individuals interested in obtaining an MBA. These differences need to be more carefully considered in applied work.