Introduction

There are many conflicting studies that suggest similarities or differences in ethical evaluations and ethical decision-making of males and females. Early gender studies have suggested that women are more ethical than males; see a meta-analysis of 47 studies by Borkowski and Ugras (1998). Implications of such research suggest that as more women enter the workforce there should be more ethical decisions made by organizations. Over the last decade, the accounting industry has seen a myriad of earnings scandals and fraud. There is only one known study (Silver and Valentine 2000) exploring gender differences and moral intensity evaluations using Jones’ model of moral intensity. Since ethical decisions are often context specific; there is a need to continue to examine whether gender differences are stable over various types of ethical dilemmas. We explore the following research question: are there differences in ethical sensitivity, moral judgment, and/or perceptions of moral intensity between male and female accounting professionals when considering a situation involving earnings management?

This study offers several contributions to prior research. First, this study responds to suggestions to extend the application of the Jones (1991) model of moral intensity (May and Pauli 2002; Cohen and Bennie 2006) and to further explore gender differences for moral intensity items (Silver and Valentine 2000). Next, this study examines gender differences for accounting professionals’ ethical evaluations, moral judgments, and intentions to report this ethical dilemma using Jones’ (1991) model of moral intensity, and examines whether males and females evaluate moral intensity differently. Further, Bailey et al. (2010) have suggested that “ethics research in the accounting literature has focused too narrowly on Component II of Rest’s Four-Component Model (1979). There has been confusion over the purpose of the Defining Issues Test (DIT) instruments, a neglect of metrics other than P scores, and a weakness in making a connection with the broader research on ethical development within the professions. As a result, the accounting research began to falter after an enthusiastic early start. We suggest that if the ultimate goal of ethics research in accounting is to improve the ethical performance of accountants, then research must consider all four components” (Bailey et al. 2010, p. 18). This study also addresses these concerns by examining the first three steps of Rest’s four component model of ethical decision-making.

The remainder of this paper is divided into four sections. The next section reviews literature concerning ethical decision-making theory and gender differences that may explain behavior. The second section reviews the survey procedures and methodology. The third section analyzes the survey data, and the final section discusses the results and implications.

Literature Review

Ethical Decision-Making Models

There are differing theories of ethical decision-making and various models that suggest processes for ethical decisions. One well-known model is described by Rest (1986) which suggests a four-step process of ethical decision-making that includes moral sensitivity, moral judgment, moral intention, and moral behavior. The first step in the model is moral sensitivity; when an individual realizes that an ethical problem exists. In this step, a moral agent recognizes that a situation presents an ethical dilemma and that there are potential consequences that can affect others as a result of the behavior of the moral agent (Rest 1986). The second step in the model is moral judgment; when an individual evaluates whether actions are morally wrong or morally right. This second step involves evaluating various courses of action. The third step in the model is moral intention; when an individual selects a course of action. The fourth step in the model is an individual engaging in moral behavior.

Gender Differences in Ethical Decisions

The psychology literature offers several theories that attempt to explain gender differences for various types of evaluations, intentions, and behaviors. A brief explanation is provided here for competing theories related to gender differences and similarities.

A cultural explanation posits that gender identity is established through the socialization processes during childhood (the socialization theory). Socialization theory suggests that because gender identity is stable and unchanging, the different values, interests, and traits that males and females bring to the workplace should cause differences in ethical perceptions to be stable over time (Dawson 1992). For example, feminine and masculine personalities are developed in childhood and differences in moral values and ethical views will be detectable through life (Dawson 1992).

A structural explanation posits that social group differences arise from common social roles found in organizations and families (the social role theory). Social role theory suggests that men and women behave according to the stereotypes associated with the social roles they occupy (Eagly 1987). For example, males are often viewed as providers for the family and females are often viewed as homemakers. Further, females can be viewed as more communal, having concern for others, and more expressive with emotions, while males are more independent and assertive (Eagly and Wood 1991). The social role theory can be viewed as more flexible than the socialization theory, because individuals often have multiple roles and may change behaviors based on their role at the time.

Some suggest that organizational roles may override gender roles and that “male and female leaders who occupy the same organizational role should differ very little” (Eagly and Johnson 1990, p. 234). Further, self-selection theory suggests that females who choose a career in business contrary to gender stereotypes may share similar values, motivations, needs, and behavioral intent as males who choose a career in business.

Gender Differences in Moral Sensitivity

Research examining gender differences for moral sensitivity has produced mixed results. Cohen et al. (1998) found that female accounting students viewed dilemmas involving expensing personal gifts as a business expense and offering bribes to foreign officials as less ethical than male students; however, females were not more sensitive than males to other business situations involving earnings management, product safety issues, and copying software. Shawver et al. (2006) found that female accountants were more sensitive to issues involving offering bribes and unfair loan practices; but were not more sensitive to other business situations. Barnett and Brown (1994) found that female students indicated higher ethical sensitivity for many business situations. In each of the 24 scenarios, male students rated all actions as more ethical than the female students with 22 out of 24 situations as statistically different (Barnett and Brown 1994).

Dalton and Ortegren (2011) suggest that gender differences in ethical decision-making may occur not because females are more ethical than their male counterparts, but because females are more prone to respond in a socially desirable way. To further explore gender differences in moral sensitivity, we present the following hypothesis in null form:

H1

There are no gender differences in professional accountants’ moral sensitivity for earnings management.

Gender Differences in Moral Judgment

Many studies have explored gender differences in moral development outside of the accounting discipline. These studies have produced mixed results. Rest (1979) reports findings from 15 different studies that show no differences in moral reasoning ability based on gender, while Thoma (1986) reports a meta-analysis of 56 studies that suggests females have higher moral reasoning abilities than males. Studies using accounting professionals and students have suggested that males have lower moral reasoning abilities than females (Ariail et al. 2012; Etherington and Hill 1998; Bernardi and Arnold 1997; Shaub 1994; Eynon et al. 1996; Etherington and Schulting 1995; Jones and Hiltebeitel 1995).

Valentine and Rittenburg (2007) found that female business professionals have marginally higher ethical judgment than male business professionals; however, no significant differences were found for judgments of moral equity. Lund (2008) found that male and female marketing professionals differ significantly in their moral judgment. Overall, female marketing professionals indicated significantly higher judgment than their male counterparts for situations involving purchasing decisions, advertising, and pricing (Lund 2008). To further explore gender differences in moral judgment, we present the following hypothesis in null form:

H2

There are no gender differences in professional accountants’ moral judgment for earnings management.

Gender Differences in Moral Intentions

Gender differences in ethical decision-making may appear as a result of the context of the dilemma. Hoffman (1998) found a significant gender difference for intentions to market an unsafe product, moderately significant differences for issues of product misrepresentation and industrial espionage, and no statistically significant gender differences for offering bribes among various levels of managers in the study. In a meta-analysis of 47 studies, Borkowski and Ugras (1998) suggest that women recognize ethical issues in business more often than do men, and that women often select ethically preferable actions. Valentine and Rittenburg (2007) found that female business professionals indicate significantly higher ethical intentions than male business professionals. Dawson (1995) identified that it might be simplistic to expect that females’ influence on organizations will lead to higher ethical standards; however, it is more certain that females’ influence on organizations will expose differences in the way that ethical problems are perceived and resolved. Nath et al. (2013) found that “gender intersects with other identities to yield different values, experiences, and opportunities that can lead to gender-based preferences for [corporate social responsibility] information.”

Several studies have explored personal issues (as opposed to business issues). In a study of 51 practicing accountants, no gender differences were found for several personal and business dilemmas involving cheating on taxes, copying software, reporting information, racial discrimination, and freedom of speech; however, gender differences were found for insider trading (Radtke 2000). Dawson (1992) found significant gender differences in a sample of college students for cases involving ethical issues of a relational nature (such as a sales manager seeking to hire a competitor’s employees, and a sales manager interrogating a newly hired employee about a previous employer’s marketing plans) but found no significant gender differences for non-relational situations involving padding an expense account, making unauthorized long-distance calls, or violating company policy for consuming alcohol at lunch-time. Beu et al. (2003) also found that females were more likely to indicate ethical intentions than males. Cohen et al. (1998) found that female accounting students indicate that certain unethical actions would not be completed. In nearly all situations where gender was significant, female accounting students perceived that they would be less likely to perform each action that they viewed as unethical (Cohen et al. 1998). For a more exhaustive review of the empirical literature, O’Fallon and Butterfield presented a listing of 49 findings pertaining to gender, and the results of the effects of gender in those works are “somewhat mixed” (2005, p. 377).

Gender Differences in Moral Intentions to Whistleblow

Gender differences for whistleblowing intention also appear to be mixed. Bjorkelo et al. (2010) found that females were more likely than males to report wrongdoing once, and males were more likely than females to report wrongdoing two or more times. Keenan (2007) studied Chinese and American managers and their propensity to whistleblow on wrongdoing, but found no significant difference based upon gender. Sims and Keenan (1998) found that male undergraduate business students were more likely to whistleblow externally. With a sample of MBA students, Kaplan et al. (2009) found that females’ reporting intentions for an anonymous reporting channel are significantly higher than males’ reporting intentions; however, no differences in reporting intentions were found when reporting to internal audit. Mesmer-Magnus and Viswesvaran found that “females and more tenured employees appear to be slightly more likely to actually blow the whistle” (2005, p. 285). Stylianou et al. (2013) found that females are more likely to report intellectual property and privacy rights violations than males. Cassematis and Wortley (2013) explored the predictors of whistleblowing that potentially separate whistleblowers from those who are non-reporting observers and were unable to find gender to be a predictor in their study of Australian whistleblowers.

The theoretical arguments related to the role of gender on whistleblowing intentions have also been mixed (Vadera et al. 2009). Researchers suggest that women are likely to report questionable or illegal acts more frequently than men because women, on average, feel a greater public responsibility to speak against wrongdoing (Rothschild and Miethe 1999). Other researchers suggest that whistleblowing behavior is considered risky and men are more likely to report these acts while women tend to conform to a majority opinion and the majority opinion may tend to not report actions (Miceli and Near 1984). To further explore gender differences for whistleblowing intentions, we present the following hypothesis in null form:

H3

There are no gender differences in professional accountants’ moral intentions to report earnings management.

Potential Factors Which May Mitigate Gender Differences

Research supports the possibility that gender differences may be mitigated by a variety of factors. For example, gender differences may decrease due to socialization and by occupational roles (Feldberg and Glenn 1979) to the point that males and females may eventually make similar work-related decisions (Lacy et al. 1983). Also, the gender effect may disappear as a result of reward structures or training, according to Owhoso (2002), who found no gender differences in estimating fraud risk by both experienced and inexperienced auditors who had been exposed to the same training.

Robin and Babin (1997) found support for possible socialization of students into the profession because gender differences were observed between students but not between retail management professionals, suggesting that one possible explanation for this may be that occupation overrides gender differences after entering a profession. “Gender differences due to socialization are more likely to appear before individuals enter the’structure’ of the business world, but once there, differences are expected to be minimal” (Robin and Babin 1997, p. 70). Further, Terborg (1977) identifies that “new employees of either sex must be socialized properly if they are to fit with the established functioning of existing work units and if they are to develop into contributing members of organizations” (Terborg 1977, p. 651).

Kish-Gephart et al. (2010) were unable to find “that males and females differ markedly in how they puzzle through ethical dilemmas or that social expectations lead to systematic, gender-specific responses to ethical dilemmas by actors in the workplace” (2010, p. 20). Because practicing accountants are trained to be professionals and are expected to adhere to codes of ethics and industry standards in performing their duties, we posit that there should be no gender difference between the ethical evaluations, judgments or intentions of practicing accountants.

Ethical Decision-Making and Moral Intensity

Jones (1991) extends Rest’s model by suggesting that prior research has not focused on the importance of the moral issue itself and suggests that moral intensity impacts each step in Rest’s four component model (described above). The first dimension, magnitude of consequences, is described as the harm or benefit to individuals arising from an action. The second characteristic, social consensus, is described as the degree of social agreement regarding the goodness or evil of an action. The third characteristic, probability of effect, is defined as the probability that an action will occur and cause harm or benefits. The fourth characteristic, temporal immediacy, is defined as the length of time between the action and its outcomes or consequences. The fifth characteristic, proximity, measures the physical, cultural, or social association of the moral agent to those affected by an action. The sixth characteristic, concentration of effect, is the degree to which a limited number of people experience harm or benefits from the action.

Gender Differences in Moral Intensity

The components of moral intensity could be perceived differently by males and females. The component of proximity has several implications for differences in ethical perceptions. Jones (1991, p. 376) suggests that “people care more about other people who are close to them (socially, culturally, psychologically, or physically) than they do for people who are distant.” The close proximity of various people including customers, stockholders, employees, peers, and managers can effect ethical evaluations and decisions. Jones (1991) identifies that proximity plays a role in legal situations. An attorney can develop close proximate relationships with a client which can clearly be separate from those who are harmed by the client. Cohen and Bennie (2006) describes an example of high proximity as situations involving close relationships auditors and clients. These relationships may result in the auditor’s independence being compromised, as the auditor could act in the client’s best interest at the expense of the other users (Cohen and Bennie 2006).

Where ethical issues involve a personal relationship, females may feel nearer to participants, which might mean that proximity is more important to women than to men when evaluating ethical issues. Females tend to demonstrate an ‘ethics of care’ (Gilligan 1982), rather than a ‘justice’ or ‘rule-based’ reasoning that males exhibit (Kohlberg 1969). Gilligan’s (1982) work suggests that females use a “different voice” or perspective when describing ethical dilemmas. In her study, females often described resolving ethical dilemmas within the context of relationships, with a responsibility of caring for others and avoiding harm. Males often described ethical dilemmas more in terms of justice, fairness, and following the rules (Gilligan 1982). However, Derry (1989) suggests that an orientation toward justice or care may be situation specific although a majority of the interviews indicated a preference for a justice orientation.

There is only one known study that has explored gender and moral intensity evaluations using Jones’ model of moral intensity. Silver and Valentine explored the varying perceptions of moral intensity among 105 college students. They found “that gender and age were determining factors, with women perceiving greater moral intensity in marketing scenarios than men, and older students perceiving greater moral intensity than younger students” (2000, p. 309). Further, significant gender differences for all six components of moral intensity were reported. That study was limited to students, most of whom probably had not worked in the accounting environment. But we posit that once the individual becomes an experienced accountant, it is possible that the gender differences are decreased or eliminated based upon prior research reported above (Feldberg and Glenn 1979; Lacy et al. 1983; Owhoso 2002; Robin and Babin 1997; and Terborg 1977). Therefore, we present the following hypothesis in null form:

H4

There are no gender differences in professional accountants’ ratings of moral intensity for earnings management.

Methodology

A pre-test of the instrument was completed using accounting students prior to collecting the data reported in this study using accounting professionals. Minor modifications were made to the instructions and the instrument as a result of the pre-test. Individuals attending a continuing professional education seminar sponsored by a state society of certified public accountants were asked to participate in this controlled experiment. Each individual was asked to evaluate a scenario related to earnings management. The scenario used in this study is based on a scenario developed by Cohen et al. (1998) and the questions to measure the moral intensity items are based on similar questions used by Singhapakdi et al. (1996) with some modification. Izraeli (1988) found that managers, in general, rated themselves more ethical than their peers when evaluating ethical beliefs and behaviors. To eliminate possible social desirability response bias, survey questions involving an intention to act were worded in the third person. The scenario used in this study and questions to evaluate moral intensity are included in Appendix.

Of the 957 attendees, 192 agreed to participate in the study (a 20.1 % response rate). Those who did not complete the survey were eliminated, leaving 171 usable responses. Table 1 provides demographic information about the individual participants in this study. Of the 171 participants, 109 are male and 62 are female. Nearly all of the participants (168 or 98 %) identified themselves as a CPA.

Table 1 Demographics
Full size table

The participants evaluated several statements utilizing a 7-point Likert scale rated from 1, “strongly agree,” to 7, “strongly disagree.” The first statement asked participants to indicate whether the vignette involves an ethical problem indicating their moral sensitivity to the earnings management technique suggested in the scenario by evaluating the statement, “The situation above involves an ethical problem” (reverse-coded). The second statement elicits a moral judgment for whether or not the manager in the vignette should act as proposed by evaluating the statement, “The manager should not do the proposed action.” Each participant indicated several moral intentions to report the action by indicating the likelihood their peers would report the manager’s request if guaranteed anonymity (reverse-coded), guaranteed their job (reverse-coded), offered a cash reward (reverse-coded) reporting to an internal manager (reverse-coded) and reporting to an external manager (reverse-coded). Each participant evaluated all six moral intensity items suggested in Jones’ model of moral intensity. Magnitude of consequences was measured by responses to the statement, “The overall harm (if any) done as a result of approving the shipment would be small.” Societal consensus was measured by responses to the statement, “Most people would agree that approving the shipment is wrong” (reverse-coded). Probability of effect was measured by the statement, “There is a very small likelihood that approving the shipment will cause any harm.” Temporal immediacy was measured by, “Approving the shipment will not cause any harm in the immediate future.” Proximity was measured by responses to the statement, “If the manager is a personal friend, approving the shipment is wrong” (reverse-coded). Concentration of effect was measured by responses to the statement, “Approving the shipment will harm very few people if any.” Table 2 provides the means and standard deviations for all of the variables.

Table 2 ANOVA results
Full size table

Results and Discussion

The participants in this study indicated that shipping product early to meet a quarterly bonus involves an ethical problem (means closer to 7). Both males and females were morally sensitive to this issue (mean for males 5.94 and mean for females 6.16). Both groups agreed that the action should not be completed with means closer to 1 (males 2.08 and females 1.94). Both groups indicated a moral intension to whistleblow anonymously for this type of behavior with means closer to 7 (males 5.27 and females 5.72). We also explored differences in reporting channel or conditions that may encourage whistleblowing behavior. We found no significant gender differences for whether these accounting professionals would report to either an internal or external manager or whether they would report if guaranteed their job or encouraged to whistleblow for a cash reward. In all reporting options, females indicated a higher likelihood to report this action (higher mean scores); however, the mean scores were not statistically different between male and female accounting professionals. Therefore, we find no gender differences for steps 1, 2, or 3 in Rest’s four component model of ethical decision-making.

When comparing gender differences for the means of the moral intensity items, males believed the situation involved slightly higher magnitude of consequences, probability of effect, and concentration of effect. Females believed the situation involved slightly higher societal consensus, temporal immediacy, and proximity. Of the six moral intensity items, both males and females indicated that even if the manager was a personal friend (proximity), approving the shipment is wrong with the highest scores of all six moral intensity items (5.73 males, 6.16 females); followed by societal consensus as their second highest mean (5.24 females, 5.10 males); followed by temporal immediacy. The professional accountants used in this study indicated that concentration of effect and overall harm were rated lower than the other moral intensity items. This may indicate that this group of accountants recognizes that the act of shipping product early to meet a quarterly bonus is unethical and should not be completed; however, it is less likely to cause harm to a large number of individuals. An ANOVA was completed to identify statistically significant gender differences in the responses to each statement. In Table 2, we observed no significant gender differences in professional accountants’ moral sensitivity, moral judgments, moral intentions to report the action under various circumstances, or evaluations of the dimensions of moral intensity for the earnings management situation described in this study. H1, H2, H3, and H4 are all supported.

We further explored the prior hypotheses by examining the Pearson correlation coefficient for each dependent and independent variable with gender (Table 3). Gender is not correlated to moral sensitivity, moral judgment, moral intention, or any of the moral intensity variables examined in this study, further supporting all four of the hypotheses.

Table 3 Correlations of gender with moral intensity and ethical decision-making variables
Full size table

MANOVA was performed to examine the impact of gender on all six moral intensity items (magnitude of consequences, societal consensus, and probability of effect, temporal immediacy, proximity, and concentration of effect). The results of the MANOVA tests are provided in Table 4. The results provide further support for H4; there is no significant gender differences in professional accountants’ ratings of the moral intensity in the earnings management scenario used in this study.

Table 4 MANOVA results
Full size table

Of all the statements used in this study, none revealed any significant differences in responses between male and female accounting professionals, supporting previous research that found no gender differences in ethical decision-making for practicing accountants (Radtke 2000) or auditors (Owhoso 2002).

To further test the robustness of these results, the sample was split by age. Fifty-six individuals indicated that they were under the age of 50 (32 % of the sample), while the remaining participants indicated that they were over the age of 50. Neither the ANOVA test nor the MANOVA test indicated significant differences (p values exceeded .05) for these age groups for the variables examined in this study.

There is only one known study that has examined gender differences for moral intensity evaluations. Our results conflict with those of Silver and Valentine (2000) who found significant gender differences with a sample of college students for all six components of moral intensity using marketing situations. The results of this study may be explained by the structural approach that suggests that organizational roles may have eliminated traditional gender roles. Further, self-selection theory suggests that females who choose a career in business may share similar values, motivations, needs, and behavioral intent as their male counterparts who chose similar careers in business. In addition, socialization of individuals, regardless of gender, occurs over time within organizations and professions. Generally, an individual will be successful when his or her performance matches the expectations of the organization or the profession. Further, codes of conduct for accounting professionals are explicit and outline proper behavior and expectations for the profession. We find that male and female professional accountants made a similar work-related ethical evaluation, moral judgment, moral intention and evaluations of moral intensity.

Conclusions

This study explores differences in ethical sensitivity, moral judgment, moral intentions and perceptions of moral intensity between male and female accounting professionals and provides interesting insights into ethical evaluations and moral judgments of male and female accounting professionals. While there are several studies that suggest females are more ethically sensitive or make more ethical choices than males, other research has found no significant differences. The implications of this study have relevance to the conflicting theories relating to gender differences in ethical evaluations.

This study found no differences in ethical evaluations, moral judgments, moral intentions to whistleblow, or evaluations of moral intensity for male and female accounting professionals. While we cannot know the reasons for the lack of differences between male and female practicing accountants, there are several factors suggested by prior research that offer explanations. It is possible that socialization in the accounting environment has effectively eliminated the gender differences (Feldberg and Glenn 1979), and that male and female accountants eventually make the same types of decisions (Lacy et al. 1983). Our results also support the idea of Owhoso (2002) that similar training may produce similar ethical evaluation results. Finally, our results also support the idea by Robin and Babin (1997) that gender differences are minimized once an individual enters the structure of the business world.

A second possibility for our results may be a severe “file drawer problem” (Rosenthal 1979). While some researchers have likely explored the possibility of gender differences without finding significant results, as we have in this study, their findings may not be reported, or be filed away in a drawer, because of the bias favoring publication of statistically significant outcomes (Hubbard and Armstrong 1994). We believe that exploring the ethical evaluations of male and female accounting professionals is both interesting and important research, and we find it even more interesting that we did not find statistically significant differences between them. If, in fact, there are no gender differences between male and female practicing accountants’ ethical decision-making, the implications on hiring and training are profound. This suggests that hiring and training may not need to be gender-specific, and there may be no need to “balance” male and female staff in order to create and maintain an ethical atmosphere.

Future Research

Although this study contributes to the literature by exploring three out of four steps in Rest’s four component model of ethical decision-making and explores gender differences for moral intensity evaluations, there are several limitations that should be noted. Accounting professionals may not actually behave in a manner similar to the way in which they responded to the intention to act when confronted with similar issues in a business environment. Few studies have attempted to measure the fourth step in Rest’s model of ethical decision-making, actual behavior. Therefore, future research may wish to attempt to measure actual behavior. Further, future studies may wish to increase the sample size, sample a different geographic area, use different situations of earnings management, or consider additional variables that might impact ethical and moral intensity evaluations.