Exploring some biases in the ethical judgement of Managers: An empirical study^*

Explorando algunos sesgos en el juicio ético de los Directivos: Un estudio empírico

Explorando alguns preconceitos no julgamento ético dos Gerentes: um estudo empírico

Corporate scandals during last years of 20th century, were the spark that increased the interest of business schools on behavioral ethics within firms. Since long ago (Kohlberg 1981), an even more now, research has been focusing on understanding moral development of people (Martynov 2009) and how people resolve ethical dilemmas and make ethical decisions within the business environment (Messick and Tenbrunsel 1996, Banaji et al. 2003). Therefore, it resulted interesting to decipher psychological tendencies that lead even good people to use information and make decisions unethically, even when they would not expect to behave that way in advance (Bazerman and Gino 2012).

The study of systematic and predictable ways in which managers make ethical decisions and judge their own ethical decisions and those of others, are at odds with intuition (Bazerman and Gino 2012). Bazerman et al. (1998) studied why people constantly struggle between a conflict of two behaviors, the want self and the should self, concluding that at the time of the decision, one's behavior tend to be dominated by the thinking of the want self. Interestingly enough, Messick et al., (1985) reached the conclusion that there is a human tendency that leads people to think that they are fairer than others.

Moreover, Jones (1991) has claimed that people intuitively tend to care more about people who are close to them -socially, culturally, psychologically, or physically- than they do for people who are distant. Banaji et al. (2003) suggested that managers are exposed to different contextual biases, among which, have noted, there is a tendency to favor -in moral judgments- those individuals who are close to oneself (in-group members) compared to people that one doesn't know (out-group members).

Accordingly, the aim of this paper is to explore, by empirical means, possible biases in the moral judgment of managers. We, likewise, believe there is a tendency within managers to judge harder the moral behavior of the average employee, compared to people they know or love, and furthermore, compared to their own moral behavior as executives. The paper will be structured as follows. In the first section, we explore the mainstream literature regarding moral decision making biases within the business context. In the second section, we explain the methodology used and the main empirical findings. In the last section, we discuss results and conclude with some recommendations and possible further research.

To identify some possible biases in ethical judgement presented in literature, it is important to define, first, what is understood by scholars as ethical judgment. Justice, fairness, acceptability and promises, are all related with ethical judgement (Carroll and Buchholtz 2014). Studies in this field have historically focused on recreating questionable ethical situations to evaluate ethical judgment of people. Nevertheless, It has been demanded by scholars to evolve from the mere analysis of the reaction of the person (behavior) to try to answer why that person had that reaction (Reidenbach and Robin 1990). Consequently, a possible way to overcome this shortfall is to try to explain the causes of these reactions by a close analysis of the possible biases in the ethical judgment of the agent.

Even though ethical judgement has been a widely studied concept, scholars haven't arrived to a generalized definition of it (Sparks and Pan 2010). It seems to be a consensus, based on the existing literature, on the multidimensional nature of ethical judgement. This multidimensionality has created disagreement in three critical points.

The first point of discussion is the rationality of the ethical judgement process. There are scholars who think that it can be based on choices (rational choice implies rational judgement) or based on goal fulfillment (a judgement is rational when helps to fulfill a goal) (Baron 2000). The second source of disagreement, is the information process in ethical judgement, which from the point of view of some scholars, can be conscious, logical and analytical (Chaiken 1980, Epstein et al. 1996), but for others, can be more experimental, based on simple decisions rules (Chaiken 1980, Petty and Cacioppo 1986). The last source of discrepancy, in the definition of ethical judgement, is the multiple options scenario in ethical judgment. An ethical judgment, according to some scholars, can be singular, judging ethicality each option alone, without considering the other alternatives involved (Kardes 2004, Chaiken 1980, Schwarz 1990) or can be comparative, weighting the ethicality of two or more options (Sparks and Pan 2010).

As a matter of fact, these three different sources of discrepancy within scholars, explain why the multidimensionality nature of ethical judgement has generated two main trends in the literature that aim to define the term. Some scholars (Hunt and Vitell 1986, DuPont and Craig 1996, Honeycutt jr et al. 2001) affirm that when ethically judging alternatives, to follow the most ethical one, there are, actually, different alternatives that arise and have different levels of ethicality. Thus the process of selecting the most ethical one, leads to a comparison between alternatives. In contrast, some other scholars (Rest 1986, Schwepker Jr l999,Valentine and Rittenburg 2004) suggest that actions are good or bad by themselves regardless of possible comparisons. Then, there would not be a scale of ethicality within both ends. This conception suggests that actions are ethical or unethical.

In an attempt to reconcile these two broad views mentioned, Sparks and Pan (2010) affirm that ethical judgement is the "individual's personal evaluation of the degree to which some behavior or course of action is ethical or unethical". Following this, and considering that the goal of the research in ethical management is to predict, explain and control ethical behavior of people in charge of organizations (Flory et al. 1993), we intend to study the main ethical judgement biases in management.

Several scholars have studied possible biases in the moral judgment of people (Batson et al. 1997, Batson et al. 2002, Rustichini and Villeval 2014) and more specifically, in the moral judgment of managers (Tenbrunsel et al. 2010, Bazerman and Tenbrunsel 20ll, Kannan-Narasimhan and Lawrence 2012, Trevino et al. 1998). According to Carroll (1987), moral managers are hard to find and play a major role in shaping organizational culture. A moral manager's main concern regarding strategic decision making, is figuring out whether his or her actions are fair or not (Carroll l989). Before a managerial strategic action is performed, a manager generally performs a moral judgment of the possible actions he or she might execute. As a matter of fact, people habitually judge ethicality of others' behaviors (Gino et al. 2010).

In fact, when a person judges the behavior of others and its own behavior, there is a general tendency or bias described by Gino, Moore and Bazerman (2010) who affirm that it is common to be more critical of others' ethical choices than of their own (Figure 1). In terms of Messick et al. (1985) "Feeling fairer or holier than others". More specifically, two of the main ethical judgement biases in management which arise from this general bias and the ones we intend to specifically study, are: the want/should self-conflict or bias (Figure 2), and, the group affiliation bias (Figure 3).

Figure 1
General Bias in the Ethical Judgment of Own and Others' Behavior
Source: The authors

Figure 2
Want/Should Self Conflict or Bias
Source: The authors

Figure 3
Group Affiliation Bias
Source: The authors

It's common to find in academic literature the scene (Bazerman et al. 1998, Ainslie 1975) described by Homer's The Odyssey, as an example of an intrapersonal dilemma, where Ulysses decided to confront sirens -females that enchantresses sailors with songs, attracting them to an island to then kill them- being conscious he will soon encounter them. Ulysses decided to put wax in the ears of his men to block songs, but didn't block his own ears. Ulysses asked his team to bind him with ropes while he listens the song and feel the pleasure of the call, but not being able to heed the call.

What Ulysses did, was to confront two opposite possible behaviors, in one side, the possibility to respond immediately to the suggestive songs of sirens without thinking on consequences (simply heed the call). On the other side, the possibility to be reflexive assessing possible consequences and avoiding the call (wax on his ears). This two possible decision paths, with which he was challenged, have been called by scholars as the multiple selves problem (Strotz l955, Ainslie l975). Research performed by Schelling (l984), found that in multiple scenarios different selves may emerge, influenced by the short-term thinker and by the long-term thinker.

This multiple selves' problem, emerges in literature as a possible bias in ethical judgment. Historically there has been an evolution of the understanding of this human tendency. For instant, Parfit (1984) and Walsh (1996) postulated a conflict between the present-self and the future-self, where immediate benefit is against a future greater benefit. These research also corresponds with theory developed by Thaler (1980) who identified two opposite selves, the planner and the doer. The planner is able to see further from immediate benefits and tends to look for greater benefits obviating time, while the doer is highly influenced by time and immediate rewards.

In contrast, Stigler and Becker (l977) carried out a study pretending to prove that tastes are not dynamic systems, but rather stable during time. They found evidence that it is not possible to assign to one person multiple selves, because tastes don't change during time. Letting time (as important variable) aside, Higgins and colleagues (l990) distinguished between two "actual" selves, the self the person thinks he/she is, and the self the person believes others think he/she is. These two selves, for authors, have an impact on the behavior of people, making them act different in diverse circumstances, depending on the self they want to be. For Loewenstein (l996), the multiple selves problem was a matter of situational factors (hunger, sleep, temper etc.) that provoke visceral responses at odds with long-term interests.

Until now, research suggests that humans tend to thrive between two opposite behaviors, one that is more related with immediate benefits and tastes driven (impulsive, present, doer, immediate etc.), and, the other one, more rational and driven by a higher benefit (future, planner, best option etc.). Consequently, Bazerman et al. (1998) addressed the multiple selves problem from the point of view of organizations and, more important, making it compatible with decision making process. These authors conceptually proposed that the two selves' problem, could be named as the want/should self dilemma.

Bazerman et al. (1998) suggest that the want-self is reflected in preferences that are "(1) emotional rather than rational, (2) affective rather than cognitive, (3) impulsive rather than thoughtful, and (4) 'hot headed' rather than 'cool headed". According to them, the should-self is "more rational, cognitive, thoughtful, and cool headed." Furthermore, Bazerman and colleagues proposed that the want self is more influential at the decision than either before or after the decision.

In empirical terms, O'Connor et al. (2002) did a study where they found that what people wanted to do and what people think should do, were very similar at the time before and after a decision. Nevertheless, in the moment of making the decision, what people wanted to do diverged significantly of what people should have done, results that go aligned with those of Bazerman and Grino (2012).

These studies conclude that the temporal dimension is vital for the want/should dilemma. It means that the want self emerges when making the decision (Mitchell et al. 1997), while the should self emerges both, before and in the recollection phase of the decision making process (Epley and Dunning 2000). Based on the want/should dilemma, Milkman et al. (2010) concluded that it's easier for people to make ethical decisions based on the should self. Specially, when the time between the decision (e.g. a person decides to buy vegetables instead of groceries for next week) and the action (e.g. during next week when the person is cooking vegetables for meal) is longer.

Tenbrunsel et al. (2010) argue that "in ethical decision making, the 'want/should' theory is useful for understanding differences between intentions, behavior and interpretations of past behavior". They also suggest that most managerial decisions are guided by visceral factors, not dependent on the moral development of the decision maker. For these authors, this bias can be found also in managers, which in the end, means that they are biased judgers of their self-ethical behavior and of the others (Messick et al. 1985, Epley and Caruso 2004, Epley and Dunning 2000). Given that managers are responsible not only for their own actions, but also for the actions of others, it becomes interesting to study this specific bias in their moral judgment.

Research has also found that people don't judge everybody with the same severity. As it was mentioned beforehand, people tend to judge harder other people than themselves. Jones (l99l) have claimed that people intuitively tend to care more about people who are close to them -socially, culturally, psychologically, or physically- than they do for people who are distant.

Tajfel and Turner (2004) affirm that intergroup relationships are biased by a human tendency that leads individuals to favor the in-group over the out-group members in evaluations and behavior. This favoritism goes over a conflict of interests, a simple perception of belonging to two distinct groups is sufficient to trigger intergroup discrimination (Tajfel 1974, Brewer 1979).

Interestingly, this bias seems to be present also within organizations and between managers. Banaji et al. (2003) suggested that managers are exposed to different contextual biases, among which, have noted, there is the same tendency to favor -in moral judgments- those individuals who are close to themselves (in-group members) compared to people that they don't know (out-group members). This bias has been labeled by scholars as the group affiliation bias.

Group affiliation bias within organizations has been already studied. For instance, Wright (2011) studied the relationship of consumers and marketers. He analyzed the response of consumers towards an unethical action of the seller. His findings show that people not only focuses on the nature of the act (good or bad), but also on the relationship they have with the executor, tending to judge more harshly out-group marketers, than in-group sellers.

In summary, research has arrived to the conclusion that not only people tend to be more benevolent judging themselves than others, but also, that people tend to be harsher judging behaviors of out-group members compared to in-group members. Given the responsibility managers have in decision making, it results interesting to study if this bias appears in the ethical judgement of their own behavior and in the behavior of other members of their organizational context.

Within the realm of moral reasoning and ethical judgment, several scholars have suggested that some demographic characteristics of the individuals play an important role on how they assess their own past and future behavior and the behavior of others. Ford and Richardson (l994) reviewed the empirical literature in order to assess which variables were influencing ethical beliefs and decision making. They found as influential: nationality, religion, sex, age, education, employment, personality, referent groups, rewards and sanctions, codes of conduct, type of ethical conflict, organization effects, industry, and business competitiveness.

For instance, in terms of gender, age and professional development, Eweje & Brunton (2010) found that females are more ethically aware than their male counterparts. Furthermore, their study suggests that age is a factor that does impact on ethical judgement. These results, additionally, indicate that there is a difference in ethical judgement related to work experience. In the same line Weeks et al. (1999) found that females tend to adopt a more strict ethical stance than their male counterparts. In their empirical study they show that there is a significant difference in ethical judgement across career stages.

Not only gender and age, but also level of study, have shown significant differences in managers' ethical judgment. Deshpande (l997), in an empirical study found that female managers perceived the acceptance of gifts and favors in exchange for preferential treatment, significantly more unethical than male managers. He also found that older managers (40 plus) perceived five practices significantly more unethical than younger managers. According to this study, graduate degrees also showed significant differences in ethical behavior. For instance, the practice of padding expense account by over 10% was reported to be significantly more unethical by managers with a degree.

Acknowledging the impact demographics may have we will also study variables such as gender, age, level of education, profession and company type and industrial sector, in order to assess possible differences in the ethical judgment of the surveyed managers. There we will not assess biases, but explicitly and directly the ethical/unethical behavior of managers past and future self-behavior and the perception of managers regarding the ethical/unethical behavior of others.

Based on the literature reviewed performed, we aim to study whether or not and, up to what extent, managers tend to be more benevolent or indulgent with themselves than they are with others (Figure 1). Specifically, we believe that managers' want/should bias is also present in ethical judgment as they might be more benevolent with their own future behavior than with their own past behavior (Figure 2). We, as well anticipate, that the biases also imply that managers tend to be harsher or more severe with the people they do not know (out-group members) than with the people they know (in-group members), as shown in Figure 3.

The theoretical framework we follow and, hence, the one that shows the results we would expect to obtain in case these biases are present, and actually playing a role in the ethical judgment of the managers is shown in Figure 4. In relation to the demographic variables, what we will assess is not biases, but explicitly and directly the ethical behavior of managers, and their perception of the ethical behavior of known people and of an average employee (Figure 5).

Figure 4
Theoretical Framework
Source: The authors

Figure 5
Ethical Behavior
Source: The authors

We surveyed managers of different levels, who have recently participated in executive programs of INALDE Business School in Colombia. The total number of managers invited to participate in the study was 671. We invited them, via e-mail, to take part in an online survey. Anonymity and confidentiality were guaranteed. We collected responses from 181 managers, giving us a response rate of 26,97%. We discarded 28 surveys with important missing information out of the l8l surveys collected. As a consequence, in the present study the total number of managers included is l53. The sample covers a variety of economic sectors, professions, education levels and ages.

Table 1 shows the most relevant demographic characteristics of the surveyed managers. The most representative economic sectors in our sample are manufacturing (16,99%) and consultancy (11,76%). The two main professions of the managers included are: Engineering (49,02%) and Business Administration (15,69%). In addition, 42,48% of respondents are women and 57,52% are men, which seems to be a balanced sample in terms of gender. The majority of our sample is made of mature people as 52,29% of it are people aged over forty years old. Most of our managers come from local companies as only 24,18% work in foreign firms. It is interesting to see that 56,2l% of the included managers has a very high level of education with Masters or Ph.D. degrees.

Table 1
Demographic Characteristics of the Managers Surveyed

Source: The authors

Following the Defining Issues Test (DIT) settled by Rest (1986, 1989), and the Moral Judgement Interview (Colby and Kohlberg 1987), we adapted and re-designed a new questionnaire for the business environment. The final research instrument we applied, is made out of 30 questions, cases and hypothetical situations. With the 30 items that the questionnaire has, we composed four main categories of ethical decision making in which managers and employees tend to be involved within the organizations. These categories are: veracity, use of working time, money management and, use of corporate assets. The 30 questions of our research were mandatory, hence, we didn't allow respondents the option NR/NA.

The questionnaire was divided in 4 main sections as it is shown in the Exhibit 1. The first section of the questionnaire is made out of ten questions regarding the perception of the managers about the ethical behavior of an average employee (out-group member) at work. The second section includes six cases where managers have to judge the ethical behavior of related people at work (in-group members). The third section of it, describes four hypothetical situations where managers forecast their own ethical future behavior at work. Finally, the fourth section of our instrument is composed of ten questions regarding managers' self-judgment of their own past behavior.

The four sections of our instrument were divided in the same four suggested categories of ethical decision making. Only the second section of the instrument was divided in three categories, leaving the veracity category outside, as we believe that cases are not prompt to assess it.

All the questions of our survey had 5 response options. In the case of the second and third section of the instrument (cases and hypothetical situations), we classified the response options according to the benevolence / harshness level in the ethical judgment. A grade of 1, means an indulgent or benevolent attitude in the judgment of the respondent. A grade of 5, means a severe or harsh attitude in the judgement of the respondent. The criteria or scale used to classify response options, for these two sections of the questionnaire are explained in detailed in the Exhibit 1. This exhibit also explains how we graded the response options for the first and the fourth sections of our questionnaire, where the response options were directly related to the 1 to 5 scale.

As the purpose of our study was to try to assess significant differences on how respondents ethically judge the suggested behaviors, the main statistical tool we used in the present study was two sample t-tests with equal variance (Wooldridge 2009).

We present our results in five sets of Tables (Tables 2 to 5). Tables 2 show the main results of the study. There, we present the mean differences in the four categories explored according to several types of behaviors without any type of filter or segmentation.

Tables 2
Mean Differences in the Categories Explored according to Type of Behavior.

Source: The Authors

Tables 3
Mean Differences in the Categories Explored and Type of Behavior according to Gender

Source: The Authors

Table 4
Mean Differences in the Categories Explored and Type of Behavior according to Age Behavior of Known People

Source: The Authors

Tables 5
Mean Differences in the Categories Explored and Type of Behavior according to Level of Study

Source: The Authors

Tables 2 are divided in four different tables comparing: own past behavior with own future behavior; own past behavior with average employee behavior; own past behavior with behavior of known people; and, average employee behavior with the behavior of known people.

Tables 3 to 5 we present the mean differences in the categories explored and the types of behavior according to several demographic variables, i.e: gender (Tables 3), age (Table 4), and level of study (Tables 5). According to the aim of our study, we explored several combinations and categories looking for possible ethical judgment differences. Here, we present only the results for which we found those significant differences in the ethical judgment of managers.

The analysis was performed using R's "t-test" module. We indicate the significance level of the mean differences according to the following notation f: significant at the 10% (0,1) level; * significant at the 5% (0,05) level; ** significant at the 1% (0,01) level, and *** significant at the 0.1% (0,001) level.

The main findings of this exploratory study, including means, standard deviations and mean differences in the compared behaviors, categories and demographic variables used to segment or filter in our research, are shown in Tables 2 to 5. The most general results are shown in Tables 2. In Tables 3 we show gender differences. Age differences are presented in Table 4. Finally, in Tables 5 we summarize differences according to the level of study of surveyed managers.

According to theory studied, we expected to have results in line with our proposed model (Image 4). Indeed, there were several results supporting our claim. However, there were also mixed results. Ethical Judgment means didn't constantly, varied according to the benevolence/harshness scale that we proposed. In what follows, we will describe those behaviors where we found statistically significant differences, showing up to what extent there are biases in the ethical judgment of managers.

The initial step we took was to compare the whole sample of managers, regardless demographic characteristics, such as age, gender or level of study. Accordingly, in Tables 2, we show the main findings regarding the ethical judgment of the managers comparing different types of behaviors for the four suggested categories. Interestingly enough, we found significant differences in three of the four categories explored, i.e. use of working time, money management and, use of corporate assets. Results were not conclusive only for the veracity category, for which there was a significant difference merely when we compared managers' own past behavior with the behavior of the average employee.

Tables 2.2 & 2.3 allow us to assess the broader bias in ethical judgment, which according to theory implies that people tend to judge harsher others than themselves (Image 1). According to our results three of the four categories, i.e: veracity, use of working time and money management did not only present significant differences, but also, support I the idea that managers tend to be more severe with others than with themselves. All mean differences found were significant at the 0.001 level. The lower mean implied a 0,24 difference and the highest a 1,06 difference.

However, in the case of the use of corporate assets, Tables 2.2 & 2.3, it showed significant differences also at the 0.001 level, but these mean differences are counterintuitive with the general bias theory (Image 1). It is also important to note that the results reported in Tables 2.2 & 2.3, although are in line with the general bias in ethical judgment (Image 1), do not satisfy what we expected to find according to the proposed model on Image 4, as the harshness level reported is somehow higher for the people manager know than for the average employee.

Regarding the want/should conflict or bias, we found that there were significant differences for three of the four categories assessed, i.e. use of working time, money management and use of corporate assets. There was not a significant difference for the category veracity, as it is shown in Table 2.1. The theory suggests that the ethical judgment of self-past behavior tends to be less benevolent than the judgement of self-future behavior (Image 2). Our results support theory in the case of the use of corporate assets as the reported mean is higher 0,92 at the 0.001 level for the own past behavior of the managers surveyed. Nevertheless, the categories of use of working time and money management, have significant mean differences at the 0.001 level, but they go the other way around pf what theory suggests.

In relation to group affiliation bias, in Table 2.4, results consistently showed significant differences between the ethical judgment of the average employee and the people the managers know, for the three categories assessed, i.e. use of working time, money management and use of corporate assets. Nevertheless, according to theory, as reported on Image 3, the expected means of severity in the ethical judgment should have been higher for the average employee (out-group members), but in our case, consistently they were higher for the behavior of the people manager know (in-group members). The lowest mean difference we found was 0,23 for the money management category and the highest mean difference was 0,56 for the use of corporate assets category.

Having compared the main variables of the study and having assessed biases in the ethical judgment of managers, we also applied t-tests for demographic variables, such as gender, age and level of study, in order to assess possible significant differences in the ethical behavior of the managers surveyed and in their perception of the ethical behavior of related people at work and of the average employee (Image 5).

Concerning gender, Tables 3 report the main findings. Regarding the managers' own past behavior there were significant differences within males and females. Consistently with what theory suggests, we found that females tend to be more ethical towards money management (mean of: 1,123) and the use of corporate assets (mean of: 2.246), compared to males (1,239 & 2.542) respectively at the level of 0.1 (Table 3.1). Regarding the managers' own future behavior our results show a significant difference of 0,36 at the level of 0.05 showing a better ethical behavior for female managers in the use of corporate assets (Table 3.2). We also found a significant difference at the level of 0.05 in the ethical judgment of the people managers know in relation to money management. Female managers reported a mean of 1.544, whereas male managers reported a mean of 1.750 (Table 3.3).

Age, according to our results, seemed to be the less influential factor. It surprisingly didn't have an important effect across the categories studied. We found significant differences only when assessing the ethical behavior of known people, specifically in the category: use of the working time. Managers over forty years reported a mean of 2.288, in contrast to a mean of 2.021 reported by younger managers with a significance level of 0.01.This result implies that older managers consider the ethical behavior of the people they know to be less upright (Table 4).

Finally, the level of education, surprisingly, according to our results showed a somehow negative effect towards the ethical behavior of managers and towards their perception of the ethical behavior of related people at work. As shown in Table 5.1, we found a significant difference in the manager's future behavior in relation to the use of working time. The most educated managers reported a higher mean of 2.291 (less ethical behavior) compared to the the managers without a master's degree who reported a mean of 1.761 with a significance level of 0.01. The results in relation to the use of corporate assess are similar as there was a mean difference of 0,25 at the level of 0.1 suggesting a more ethical behavior for less educated managers. As reported in Table 5.2 we found a significant difference of 0,21 with a significance level of 0.05 in the money management category. Those managers with masters or Ph.D. degrees consider the ethical behavior of the people they know to be less ethical.

As we presented in the introduction of the paper -according to literature-, human beings are exposed to three main biases in ethical judgment. The first one, is the general tendency to feel more ethical than others. The second one, is called want/should self conflict or bias, which appears especially in the predictions of the future own behavior. It implies that one tends to be more benevolent with the actions to be performed and harsher regarding the past actions. The last bias is called group affiliation bias, which suggests that the agent is influenced by the proximity that he or she has with the person that is being judged. It implies that often people tend to be harsher with the unknown people (out-group members).

In our study we found that Colombian managers are more severe judging others (known and unknown people) than judging their own past behavior. These results are consistent with the general bias theory of Gino et al. (2010) and Messick et al. (1985), for whom managers tend to feel holier than the average employee or, even more, they feel more ethical than people that are closely related to them. In our study this general bias was present in all categories and type of behaviors studied, as it is shown by the very significant differences we found.

Our results were consistent with theory for three of the four categories studied, i.e. veracity, use of working time and money management, in the sense that our respondents are more benevolent with themselves than with others. However, still with a significant difference, we found that managers in the case of the use of corporate assets, were more benevolent judging others than themselves. A possible explanation to this finding, is that managers regarding the use of corporate assets as cellphones, corporate computers, or even company cars, feel a certain right to use them for personal purposes, as in their positions, very often, the company allows them to do so, but still, they might consider that behavior to be wrong.

In relation to the want/should self conflict or bias, literature reviewed argue that it's easier for people to make ethical decisions based on the should self, thinking in the future rather than in the past (Milkman et al. 2010, Bazerman et al. 1998, Gino et al. 2010). In our study, similarly to the results of O'Connor et al. (2002), we found very significant differences in three of the four categories explored, i.e. use or working time, money management and use of corporate assets, meaning that this bias seems to to be really present in our sample.

Even though we found the want/should bias present in our sample, we found mixed results. For the categories of: use of working time and money management, our results show that managers think they could behave in the future with lower ethical standards than the standards with which they have behaved in the past. This results go against theory that suggests that people tend to be more benevolent towards future action. Only in the case of the use of corporate assets, our results would support the fact that managers are more benevolent towards their own future actions. This result goes aligned with the one found in relation to the general bias, as it means, that managers think they could do better in relation to the use of corporate assets.

Regarding the group affiliation bias (Jones 1991, Tajfel and Turner 2004) we found very significant differences for all categories explored, meaning that in our sample this bias is present. Studies until now, present the group affiliation bias as the tendency to favor -in moral judgments- those individuals who are close to oneself (in-group members) compared to people they don't know (out-group members) (Banaji et al. 2003,Wright et al. 2011).

However, our findings are somehow surprising as in the three categories explored, i.e. use of working time, money management and use of corporate assets, we found results which consistently are at odds with current literature. The managers surveyed were steadily harsher with people they know than with the average employee that they would not know.

Our group affiliation findings would suggest a very enriching line of research, because generally one would expect a direct relationship between ethical judgment and severity of punishment, whereas the more unethical perceived the behavior, the more severe the punishment (Wright et al. 2011). Hence, if managers tend to judge harder ethical behaviors of near people, we could expect to have a more severe punishment for an in-group member than for an out-group member, in similar ethical situations. This must be taken into consideration by organizations, in the design of human resources management practices, in order to avoid this bias in ethical decision judgment.

Regarding the impact of some demographic characteristic on the ethical judgment, we mainly analyzed, gender, age and level of study. These three variables revealed to have an impact, as we found significant differences for the three of them. In terms of age, our results show evidence supporting I the effect of age in the ethical behavior of the people managers know. This results also highlight the relevance of age in ethical judgment (Deshpande 1997, Eweje and Brunton 2010). We found that, younger managers tend to think that, the people they know. Behave better than people known by older managers, in relation to the use of working time. These results would suggest two implications. First, the fact that managers get used to practices that are unethical. There are practices that at the beginning they considered to be wrong but, with the pass of time, they got used to them. Second, our results suggest the opportunity of allowing young people to take responsibility sooner, as they could be less greedy and more ethical.

In terms of level of study, Deshpande (1997) suggests that the better the education the better the ethical behavior. Our results, show also an impact of the level of education, but they go the other way around. In our sample, the less educated managers considered they would have a better future behavior compared to the more educated managers, in the categories of use of working time and use of corporate assets. We obtained similar results in the case of money management, in the ethical judgment of known people.Younger managers were more benevolent than older managers in this assessment. These results, suggest to us two implications. On the one hand, the fact that business schools should increase the awareness level of their students, in order to have managers with higher moral standards. But, on the other hand, the fact that perhaps, undergrad schools might be doing a better job, than graduate schools, in terms of the ethical training of their students.

Gender in literature has been considered as an important factor in ethical judgment (Eweje and Brunton 2010, Weeks et al. 1999, Deshpande 1997). Our results showed this important effect. We obtained significant differences, especially in the use of corporate assets, both when assessing own past and future behavior. In our sample, aligned with the results of Eweje & Brunton (2010) and Deshpande (1997), we found that female managers tend to be more ethical than male managers. The category of money management, revealed also a significant difference in the ethical judgment of own past behavior. Female managers, according to our results, behave better than male managers.

Weeks et al. (1999), suggest that female managers tend to be more severe in the judgment of the ethical behavior of others, compared to male managers. According to our results, for instance, in the case of money management, we obtained a significant difference, but this difference implies that female managers are more benevolent towards the people they know, compared to the result obtained for male managers. In our view, the results obtained suggest the importance of giving more space to women in management positions of organizations. Women, according to our results, are not only more ethical, but also more sensitive and, thus, more benevolent towards others. Having more women in management positions would help to achieve organizational cultures where care is promoted and perhaps to reach more humane organizations.

Up to here, we have shown that, indeed, there are some biases that affect the ethical judgment of people. We believe that it is interesting enough that we have achieved so within the organizational context, especially with a sample of managers. However, we believe there is still a lot of work to do in order to assess the precise effect, and the possible causes of these managerial biases.

We believe the main contribution of our research is to show, by empirical means, the existence of these three biases within the organizational context. Hence, as our results showed, there still seems to be a lot to explore, in order to better assess how these biases work. One thing is to know there is a bias, and another thing would be to get to know the exact role and effect this bias plays.

We have shown some of the possible effects of these three biases and the effect of some demographic characteristics on the ethical judgment of managers. However, there are still alive the questions of why these biases are present and of how to overcome them. We believe this paper highlights the importance of carrying out further research for that purpose.

Further research might consider why it is the case that the level of affiliation of a person influences the ethical judgment of managers and even more how to overcome this bias in order to be more objective and prudent. Also, it would be very interesting to study why is it, that a person tends to believe that his or her future behavior would be better than his or her past behavior. Even more, some line of research, could try to investigate why there is a general tendency or bias of judging more severe others than oneself. All these three lines of research, up to our knowledge, haven't been explored within managerial positions, and we believe the findings of this paper, open a door that would be interesting enough to explore.

Ethical judgment is a daily task of managers. Understanding possible biases in the way they arrive to their conclusions and the factors that shape their decisions is relevant, as their actions affect not only themselves but, mainly, other persons. Very often, managers make biased decisions, not because of a bad intention, but because they do not realize that they are affected by some biases. The first step for managers to produce better decisions, is to increase their level of awareness about these biases and about the effect that some demographic characteristics might also have on their ethical judgement.

In the present study, we have shown, by empirical means, that there are three specific biases in the ethical judgment of managers, namely: a general bias, a want/should bias and a group affiliation bias, which play a role. Although, contrasted with current theory, within the realm of ethical judgment, we obtained somehow mixed results, we believe the contribution we make with this research is strong enough as it shows the real existence of these biases.

Our results, which are valuable also because of the sample, show that depending on the bias analyzed and on the demographic characteristic studied, the significant effect varies in each of the four categories of ethical judgment that we considered, namely: veracity, use of working time, money management and use of corporate assets. Still, we consider further research initiatives could focus on understanding the causes of the effects we have shown and on trying to figure out possible ways to overcome these biases. We hope this paper helps managers to increase their level of consciousness about these ethical judgment biases, and therefore -even without further research-, we hope they start figuring out ways in which they may overcome these biases.