Journal of Business Ethics

, Volume 115, Issue 3, pp 531–553

The Bright and Dark Sides of Religiosity Among University Students: Do Gender, College Major, and Income Matter?

Authors

  • Yuh-Jia Chen
    • Rinker School of BusinessPalm Beach Atlantic University
    • Department of Management and Marketing, Jennings A. Jones College of BusinessMiddle Tennessee State University
Article

DOI: 10.1007/s10551-012-1407-2

Cite this article as:
Chen, Y. & Tang, T.L. J Bus Ethics (2013) 115: 531. doi:10.1007/s10551-012-1407-2

Abstract

We develop a theoretical model involving religiosity [intrinsic (I), extrinsic-social (Es), and extrinsic-personal (Ep), Time 1], Machiavellianism (Time 2), and propensity to engage in unethical behavior (Time 2) to investigate direct and indirect paths. We collected two-wave panel data from 359 students who had some work experiences. For the whole sample, intrinsic religiosity (I) indirectly curbed unethical intentions through the absence of Machiavellianism, the bright side of religiosity. Both extrinsic-social (Es) and extrinsic-personal (Ep) directly, while extrinsic-social (Es) indirectly, exacerbated unethical intentions, the dark side of religiosity. Multiple-group analyses across gender, college major, and income showed that the bright side existed directly for low-income students, but indirectly for males and females, business majors, and low-income students. Our novel finding showed that Ep undermined unethical intentions indirectly for females. For the dark side, Es incited unethical intentions directly for males, business students, and low-income individuals, but indirectly for females, psychology majors, and low-income people. The Machiavellianism–unethical intentions relationship was the strongest for high-income participants. Religiosity had the highest number of significant paths for low-income individuals and the strongest dark side for males and high-income students, but the highest bright outcome for females. Our novel, original findings foster theory development and testing, add new vocabulary to the conversation of religiosity and unethical intentions, and improve practice.

Keywords

ReligiosityReligious Orientation ScaleMachiavellianismUnethical intentionsTheftCorruptionDeceptionGenderMajorIncomeASPIRE

Introduction

Kish-Gephart et al. (2010) investigated ethical decision making from the perspectives of bad apples (individual), bad cases (issue), and bad barrels (environment). Individual characteristics contribute to ethical/unethical behaviors (Patwardhan et al. 2012). More than six decades ago, Culliton (1949) stated that “religion has something to offer business” (p. 265). Religiosity is defined as the faith that a person has in God (McDaniel and Burnett 1990). Although 94 % of respondents in the US express a belief in God (Sedikides 2010), very few researchers treat religiosity as a major individual difference variable in studying business ethics (Hunt and Vitell 1993; Singhapakdi et al. 2012; Tang 2012; Tracey 2012; Vitell 2009; Vitell et al. 2005, 2009; Weaver and Agle 2002).

There are many religions around the world (Gibbons 2007). Among them, from the Christian perspective, the Bible teaches us to obey the Ten Commandments which can be summarized into two key sub-constructs: “love your God” and “love your neighbor.” To further simplify, it is related to one simple idea: “love one another” or to an even higher level: “love your enemies.”1

Many researchers have attempted to operationalize constructs of religiosity in the literature. Since Hill and Hood’s review of 126 measures of religiosity in 1999, more scales have been developed (Egbert et al. 2004). Among them, Allport and Ross’ (1967) Religious Orientation Scale (ROS) has been recognized as one of the most often-used scales in the literature. ROS has three dimensions: intrinsic religiosity (I), extrinsic religiosity-social (Es), and extrinsic religiosity-personal (Ep) (Flere et al. 2008; Maltby 2002; Tiliopoulos et al. 2007). Intrinsically motivated individuals live their religion and treat religion as an end unto itself; whereas those extrinsically motivated people use their religion and consider religion as a means to some end (friendship or solace; Trimble 1997). People with high-intrinsic religiosity find their master motive in religion, internalize religious beliefs and prescriptions, find the meaning in life (Francis and Hills 2008), and are less likely to engage in unethical behavior (Den Hartog and Belschak 2012; Vitell et al. 2007). Further, intrinsic religiosity curbed unethical intentions not only directly but also indirectly through the absence of Machiavellianism (Tang and Tang 2010).

Intrinsic religiosity (I) is consistently a determinant of consumer ethical beliefs, but extrinsic religiosity is not (Vitell and Paolillo 2003; Vitell et al. 2006). Most researchers have focused on intrinsic religiosity only, because extrinsic religiosity is related to the use of religion as a means to an end. We assert that people with Es are likely to apply Machiavellianism to achieve their goals and engage in unethical behaviors. However, researchers have paid little attention to extrinsic-personal motives (Ep) (focusing on inner peace and solace) and its impacts on Machiavellianism and unethical intentions.

Most people look to the social context to determine what is ethically right and wrong, obey authority figures, and do what is rewarded (Treviño 1986). Getting people to contemplate their own ethical values by recalling the Ten Commandments or signing an honor code eliminates cheating completely, while offering poker chips to redeem for cash, a few seconds later, doubles the level of cheating (Ariely 2008; see also Aquino et al. 2009). We address additional issues related to ethical decision making, below. First, males are more unethical than females (Ritter 2006). Second, business students have more ethics courses (Evans et al. 2006) but are more unethical than psychology students (Tang and Chen 2008). Third, some suggest that high-income people are ethical (Lam and Hung 2005); others argue that it is the rich—in higher social class—who do more unethical things than those in lower social class (Piffa et al. 2012). Furthermore, one’s wealth (Gino and Pierce 2009, 2010), love of money (LOM), and pay satisfaction (Tang 2007; Wong 2008) may curb or incite unethical behaviors across various situations. Very little research has investigated all three dimensions of religiosity, Machiavellianism, and unethical intentions simultaneously across gender, college major, and income in the literature. This study fills the void and challenges the assumption that all three aspects of religiosity (I, Es, and Ep) promote ethical intentions consistently.

We develop a theoretical model involving three dimensions of religiosity (I, Es, and Ep, measured at Time 1), Machiavellianism (Time 2), and unethical intentions (Time 2) and investigate simultaneously direct paths (Religiosity → Unethical Intentions) and indirect paths (Religiosity → Machiavellianism → Unethical Intentions) (Fig. 1). We collected two-wave panel data from 359 students who had some work experiences. We posit: intrinsic religiosity (I) is negatively related to unethical intentions, whereas extrinsic religiosity (Es and Ep) is positively associated with unethical intentions, creating the bright and dark sides of religiosity, respectively. Intrinsic religiosity (I) is negatively, whereas extrinsic religiosity-social (Es) is positively related to Machiavellianism (a mediator) that, in turn, is positively related to unethical intentions. Exploring the complex boundaries in multi-group analyses across gender, major (psychology vs. business), and income offers intricate insights and discoveries. Our bright and dark sides of religiosity do provide theoretical, empirical, and practical contributions to the literature (cf. Colquitt and Zapata-Phelan 2007), add new vocabulary to the conversation of religiosity and unethical intentions in the literature, and provide important practical implications.
https://static-content.springer.com/image/art%3A10.1007%2Fs10551-012-1407-2/MediaObjects/10551_2012_1407_Fig1_HTML.gif
Fig. 1

Theoretical model

Theory and Hypotheses

Religiosity

Theory of planned behavior (TPB; Ajzen 1991) suggests that attitudes toward the behavior, subjective norm, and perceived behavioral control predict “behavioral intention” that, in turn, predicts behavior. Although researchers have examined TPB in many fields (Armitage and Conner 2001), the contribution of TPB is not as ubiquitous as most researchers once thought, with some exceptions (Conroy and Emerson 2004; Fry 2003; Singhapakdi et al. 1999; Vitell and Paolillo 2003). Borrowing from the “attitude-to-behavioral-intention aspect” of the TPB, we briefly review the literature related to good/bad apples’ intrinsic/extrinsic religiosity and ethical/unethical intentions, below.

Most religions in different cultures around the world instill values, norms, and expectations of what is right or wrong and guide people to behave ethically (Tang 2010, 2012). “Religion has strong ties to morality in that religions prescribe morality” (Geyer and Baumeister 2005, p. 413). Religiosity has three components: religious affiliation, religious activities, and religious beliefs (Bjarnason 2007), and promotes conformity and inhibits deviance by encouraging the internalization of moral values and the acceptance of social norms (Cochran et al. 1994). Participation in religious activities is a persistent and non-contingent inhibiter of adult crime (Evans et al. 1995), deviance (Kerley et al. 2011), violent behavior for Christians and Muslims (Brettfeld and Wetzels 2011), and the use of tobacco and alcohol among high-school students (Cochran et al. 1994). In a study involving 20 countries and 17,234 people, religious people consider social responsibilities more important than their non-religious counterparts (Brammer et al. 2007). Following repeated calls for using “more comprehensive measures of religiosity” (Vitell 2009, p. 158), we adopt the ROS (Allport and Ross 1967) in this study and investigate three constructs of religiosity. We turn to intrinsic religiosity, next.

Intrinsic Religiosity (I)

Individuals with intrinsic religiosity (I) are brought into harmony with the “religious beliefs and prescriptions” (Allport and Ross 1967, p. 434), embrace creed, internalize it, and follow it fully. They live their religious convictions, go to church, and pray. Religion is an end in and of itself. The focus is on religion for its more inherent, spiritual objectives (e.g., how one might serve his or her religion or community) (Vitell et al. 2009). In a meta-analytic review of religiosity and big five personality, Saroglou (2002) found that intrinsic religiosity and spirituality are related to extraversion, high-emotional stability, and low neuroticism. Frequency of personal prayer, a significant sign of intrinsic religiosity, is the dominant factor in the relationship between religiosity and psychological well-being (Maltby et al. 1999). Among business students, there is a significant relationship between the degree of religiousness and attitudes toward the economic and ethical components of corporate social responsiveness (Angelidis and Ibrahim 2004). Among American adult consumers, intrinsic religiosity (I) is a significant determinant of ethical beliefs (Vitell et al. 2007). For both Anglos and Hispanics, intrinsic religiosity is a significant predictor of the “active” and “passive” dimensions of the consumer ethics scale (Patwardhan et al. 2012). Recently, in a national sample of 205 business practitioners, Singhapakdi et al. (2012) found that intrinsic religiosity’s positive impact is stronger than extrinsic religiosity’s negative impacts on ethical decision making. Further, it should be pointed out that in their study, they employed all eight items of intrinsic religiosity and only four items (out of six items) of the extrinsic religiosity (one item of Es and three items of Ep).

Following the literature related to authentic leadership and religious values, Tang and Liu (2012) investigated the construct of ASPIRE2—perceptions of authentic supervisor’s personal integrity and character—with three sub-constructs: (1) honesty and integrity, (2) caring servant, and (3) transparent help. ASPIRE suggests that an authentic supervisor or leader (1) has honesty3 (Murphy 1993), fairness (Colquitt et al. 2001; Folger and Konovsky 1989; Tang and Sarsfield-Baldwin 1996), and integrity4 (Dineen et al. 2006; Palanski and Yammarino 2009; Simons 2002; Simons et al. 2007), (2) cares about others’ work and provides services to subordinates as a servant,5 (Avolio et al. 2004; Bass and Steidlmeier 1999; Greenleaf 1970), and (3) is friendly and trustworthy6 (Gilbert and Tang 1998) and offers transparent decision making (Milton 2009) and professional development.7 In a multiple-wave panel study of part-time workers who were also business university students, a significant interaction effect between love of money (LOM) and ASPIRE on unethical behavior intentions provides the following novel, profound, and interesting findings: People with high LOM and low ASPIRE had the highest unethical behavior intentions, whereas those with high LOM and high ASPIRE had the lowest. These findings indirectly support the notion that perceptions of their supervisors’ religious values, as measured by the ASPIRE scale, curtail their own unethical intentions.

Extrinsic Religiosity-Social (Es)

Extrinsic values are instrumental and utilitarian (Allport and Ross 1967). People turn to God, but without turning away from self. Extrinsic religiosity has two sub-dimensions: personal and social. The extrinsic-social (Es) dimension deals with the purposes of achieving mundane social or business goals—how one’s religion might serve oneself, e.g., make friends, promote one’s business interests, meet the right people, gain social standing and acceptance in the community, or sell insurance. Anecdotal evidence suggests that after church services, restaurant owners announced the grand opening of their new restaurant and invited the congregation to the celebration, providing not only the enjoyment of fellowship and excellent food but also an opportunity for business revenue due the principle of “reciprocity” (Tang et al. 2008). Social influence on religious behavior is motivated by a desire to gain the self-serving, extrinsic end of social approval. This construct may indirectly imply that people focus on how you “use your neighbor,” rather than “love your neighbor” or “love one another.” Intrinsic religiosity was unrelated to willingness to donate organs, but Es was (Ryckman et al. 2004). Es is related to manipulating social perceptions because both macro-level and micro-level culture moderated the relation between socially desirable responding and religiosity (Sedikides and Gebauer 2010). Although there is no significant difference in intrinsic religiosity between Anglos and Hispanics, Hispanics have significantly stronger extrinsic religiosity than Anglos (Patwardhan et al. 2012).

Extrinsic Religiosity-Personal (Ep)

Extrinsic-personal (Ep) deals with private personal gains (e.g., relief, protection, comfort, peace, and happiness), i.e., a coping mechanism (Laufer and Solomon 2011). Researchers have overlooked Ep or have combined it with Es and examined extrinsic religiosity as one overall construct. Very limited empirical research is available on Ep. Intrinsic religiosity plays a role in counterbalancing the negative impact of extrinsic religiosity on the internalization of moral identity and also the negative and indirect impact of extrinsic religiosity on symbolization of moral identity via self-control (Vitell et al. 2009). Frequency of personal prayer is related to psychological well-being and intangible benefits (Maltby et al. 1999)—finding internal peace.8Ep is, to some extent, less associated with Machiavellianism than Es. Ep may function like intrinsic religiosity (I) and serve an end unto itself for some people.

Machiavellianism (A Means)

Machiavellianism (Christie and Geis 1970; Machiavelli, 1513/1966) is based entirely on expediency, manipulation, exploitation, and deviousness, and is devoid of the traditional virtues of trust, honor, and decency. Machiavellianism is related to unethical behavior (Collins 2000; Jones and Kavanagh 1996). High Machs manipulate more, win more, persuade others more, have higher performance, higher strain, lower satisfaction, steal more, aggress more against others, are rejected more as social partners, are associated with antisocial behavior, and are concerned about extrinsic goal of financial success specifically than low Machs. Young people are more Machiavellian than older ones. Very few researchers have incorporated Machiavellianism as a mediator (a means) of the religiosity to unethical intentions (an end) relationship. In this study, we treat Machiavellianism as a mediator of our theoretical model.

Propensity to Engage in Unethical Behavior (PUB) (An End)

Although it is not possible to measure people’s unethical behaviors directly (except in laboratory experiments), people are willing to provide accurate information answering an anonymous paper-and-pencil survey or computer-administered questionnaire (Schoorman and Mayer 2008). There is a significant relationship between job incumbent’s self-report and the coworker’s peer-report on counterproductive work behavior toward other persons and work stressors (De Jonge, and Peeters 2009; Fox et al. 2007; Martin et al. 2007). We acknowledge the significant differences between unethical intentions and actual unethical behaviors. Self-reported behavioral intentions are adequate surrogate measures of actual unethical behaviors (Jones and Kavanagh 1996). Among counterproductive behavior (Cohen-Charash and Spector 2001), deviant behavior (Litzky et al. 2006; Robinson and Bennett 1995), bribery (Martin et al. 2007), theft (Greenberg 2002), corruption (Ashforth et al. 2008), and unethical behavior (Treviño and Youngblood 1990), we select three sub-constructs of the Propensity to Engage in Unethical Behavior Scale (PUB) (Chen and Tang 2006): theft, corruption, and deception, in this study. PUB has been tested empirically in China (Du et al. 2007), Hong Kong (Tang and Chiu 2003), Macedonia (Sardžoska and Tang 2009, 2012), the US (Piffa et al. 2012), 31 entities across six continents (Tang et al. 2011), and multiple-wave panel experiments (Tang and Chen 2008; Tang and Liu 2012; Tang and Tang 2010), and cited in influential reviews (Kish-Gephart et al. 2010) and books (e.g., Bateman and Snell 2013).

Our Theoretical Model

Direct Paths

Attitude predicts behavioral intention effectively only when there is a high correspondence between the attitude (religiosity) and intentions (unethical behavior intentions) (Tang and Baumeister 1984). Most people with intrinsic religiosity (I) live their religion, treat religion as an end unto itself, and participate in religious activities. Intrinsic religiosity is a persistent inhibiter of crime, deviance, violent behavior, and unethical intentions. They are less likely to engage in unethical behavior. Participants primed with religious words cheated significantly less on a subsequent task (Randolph-Seng and Nielsen 2007). Religiosity is associated with lower self-reported academic dishonesty or cheating among college students (Rettinger and Jordan 2005) and is a significant determinant of ethical beliefs (Vitell et al. 2006).

Those with extrinsic religiosity (Es) turn to God, but without turning away from self (Allport and Ross 1967), use their religion (friendship; Trimble 1997), and promote their business interests (Vitell et al. 2009). We assert that Es is strongly and positively related to unethical behavior intentions. However, the relationship between Ep and unethical intentions may be much weaker than Es. We test direct paths from religiosity (I, Es, and Ep) to unethical intentions, below:

Hypothesis 1a

Intrinsic religiosity (I) is negatively related to unethical behavior intentions.

Hypothesis 1b

Extrinsic religiosity-social (Es) and extrinsic religiosity-personal (Ep) are positively related to unethical behavior intentions.

Indirect Paths

We incorporated Machiavellianism as a mediator (a means) of the relationship between intrinsic religiosity and unethical behavior intentions because Machiavellianism is strongly related to unethical behavior (Collins 2000; Jones and Kavanagh 1996). Intrinsic religiosity is negatively related to Machiavellianism (Watson et al. 1989). People with high Machiavellianism tend to have low-ethical sensitivity (Jones and Kavanagh 1996) and incite aggressive, manipulative, exploitive, and devious tactics in order to achieve their goals. Further, based on 266 business students in a four-wave panel study, intrinsic religiosity deterred unethical intentions not only directly (Intrinsic Religiosity → Unethical Behavior Intentions) but also indirectly through the absence of Machiavellianism (Intrinsic Religiosity → Machiavellianism → Unethical Behavior Intentions) (Tang and Tang 2010). Thus, people with high-intrinsic religiosity are less likely to employ manipulative strategies (Machiavellianism) that lead to a low level of ethical behavior.

Following these arguments and the defining meanings associated with Es, we assert: extrinsically motivated individuals with strong social motives (Es) are likely to use religion as a means to an end and deploy Machiavellianism in order to achieve their goals. Machiavellianism, in turn, is associated with unethical intentions. However, a positive relationship between extrinsic religiosity-personal, Ep, and Machiavellianism may not be as strong as Es, or may not be significant, because Ep concentrates on their intangible personal needs—security, inner peace, and solace. In summary, the relationships between religiosity and Machiavellianism are different for intrinsic religiosity (I) and extrinsic religiosity (Es and Ep). We will not propose a formal hypothesis for Ep.

Hypothesis 2a

Intrinsic religiosity (I) is negatively related to Machiavellianism that in turn is positively related to unethical intentions.

Hypothesis 2b

Extrinsic religiosity-social (Es) is positively related to Machiavellianism that in turn is positively related to unethical intentions.

Gender

We now examine our three moderators. Regarding gender, males’ Machiavellianism scores tend to be higher than (Christie and Geis 1970), similar to (Webster and Harmon 2002), or lower than those of their female counterparts (Rayburn and Rayburn 1996). Further, male students have higher concerns about career advancement and are at least twice as likely to engage in unfair practices as their female counterparts (Betz et al. 1989; Malinowski and Berger 1996). Female managers are more ethical than their male counterparts regarding unsafe products (Hoffman 1998), accepting favors for special treatment (Deshpande 1997), or ethical reasoning (Beu et al. 2003). Ethics training may have limited effects for female students but no effect for male students (Ritter 2006). The religiosity to academic dishonesty relationship was significant for females, but not for males.

Males tended to have higher satisfaction with pay than females, whereas females tended to have higher satisfaction with co-workers than males (Tang and Talpade 1999). The religion–emotional stability relationship exists for women (Saroglou 2002). Females are more concerned about their friendship, security, and solace than their male counterparts. Thus, the relationships between extrinsic religiosity (Es and Ep) and Machiavellianism are different for males and females. Without any additional empirical findings to guide us in making predictions for our hypotheses, we speculate that Es will be more strongly related to unethical behavior intentions for males than for females. In addition, it is also plausible that females with high-extrinsic-personal motives (Ep) may have lower likelihood of using manipulate strategies (Machiavellianism) as a means to achieve their end than males. We will not propose a formal hypothesis for Ep.

Hypothesis 3a

Extrinsic religiosity-social (Es) is positively related to unethical intentions for males, but not for females.

College Major (Psychology vs. Business)

Following Treviño’s (1986) theory of ethical decision making, people do look up to the social context (Bandura 1977), obey authority figures (Milgram 1974), and do what is right and rewarded (Skinner 1972) in our societies (Victor and Cullen 1988). Higher levels of religiosity at the societal level are linked to more positive correlations between religiosity and psychological well-being (Lavric and Flere 2008). Sedikides and Gebauer (2010) found in a meta-analysis that the socially desirable responding to religiosity relationship was more positive in samples that placed higher value on religiosity across culture (in descending order: the United States, Canada, and United Kingdom) and institutions (Christian universities and secular universities). A few “unsavory individuals,” who lack in some personal quality, may be labeled as “bad apples” (Treviño and Youngblood 1990, p. 378). There are more managers labeled as bad apples in the public sectors than in the private sectors (Sardžoska and Tang 2009).

People enter the fields of psychology or business based on their aptitudes, financial resources, dispositional values (Staw et al. 1986), beliefs regarding the economic return of a college education, and the Attraction–Selection–Attrition (ASA) process. Those with a strong love-of-money orientation enter the business major (McCabe et al. 2006; Tang and Chen 2008; Tang et al. 2012), whereas those with a strong helping orientation enter the psychology major. Although business students have more ethics courses than psychology students, very little evidence, however, supports the notion that students who take ethics courses will make ethical decisions (Conroy and Emerson 2004; Ritter 2006; Traiser and Eighmy 2011). In fact, highly educated executives in recent scandals received their training at the best, elite, and privileged business schools (Merritt 2002; The Daily Record2003). Scholars argue that these scandals are not caused by executives’ lack of “intelligence” or “brains,” but rather, by their self-interests (their LOM) and lack of “wisdom,” “virtue” (Feiner 2004, p. 85), or integrity, honesty, and character.

The ethics gap found between undergraduate business students and non-business students is discouraging to researchers. Some wonder whether the business curriculum has contributed to it or failed to decrease it (Richards et al. 2002). The top business schools not only fail to improve the moral character of students but actually weaken it (Schneider and Prasso 2002). After taking a single semester of introductory economics, for example, students show a significant decline in honesty and increase in self-interest (Frank et al. 1993). Business students see cheating as more acceptable or necessary in order to get ahead than non-business students (McCabe et al. 2006). Tang and Chen (2008) investigated LOM, Machiavellianism, and unethical intentions using a two-wave panel study design among psychology and business students in a public university. They found that Machiavellianism mediates the relationship between LOM and unethical intentions for the whole sample. Furthermore, this mediating effect existed for business students (n = 198) but not for psychology students (n = 100), for male students (n = 165) but not for female students (n = 133), and for male business students (n = 128) but not for female business students (n = 70). Moreover, when examined alone, the direct effect (the LOM → Unethical Intentions) existed for business students but not for psychology students. Thus, male students are more unethical than female students and business students are more unethical than psychology students.

Hypothesis 3b

Extrinsic religiosity-social (Es) is positively related to unethical intentions for business students, but not for those psychology students.

Income

Based on the data collected from 6,382 managers in 31 geopolitical entities across six continents, Tang et al. (2011) investigated the relationship between LOM and corruption intent, and incorporated pay satisfaction at the individual/micro level (Level 1) and Corruption Perceptions Index (CPI) at the entity/macro level (Level 2) as two moderators. The significant cross-level three-way interaction effect showed that for managers with high pay satisfaction, the intensity (slope) of the LOM to corrupt intent relationship was almost identical in high- or low-CPI entities, but the former had the lowest magnitude (intercept) of corrupt intent, whereas the latter had the highest. For those with low pay satisfaction, the slope was the steepest in high-CPI entities, but was flattest in the low-CPI entities, and the difference between the two was significant. In short, the rich in rich and ethical entities have the lowest corrupt intent, whereas the rich in poor and unethical cultures have the highest.

Lam and Hung (2005) investigated the relationship between ethics and income among individuals of different religions in Hong Kong, China. They argued that sometimes an ethical person may have to pay the price for his integrity and honesty in a competitive society. Higher religious values discourage individuals from pursuing material possessions actively and may lower their income. They found, however, a simultaneous relationship between income and being ethical: An increase in income increases individuals’ likelihood of being ethical for both Christians and the people of traditional Chinese religion, but reduces it for the non-religious group. Being ethical contributes to higher income for Christians and the non-religious group, but lowers it for people of traditional Chinese religion. On the other hand, based on seven studies using experimental and naturalistic methods, Piffa et al. (2012) found that upper class individuals (the rich) behave more unethically than lower class individuals. It is plausible that intrinsic and extrinsic religiosity may have different impacts on high-income and low-income students’ unethical behavior intentions. We will examine this issue on an exploratory basis.

Perceptions of wealth based on social comparisons evoke two potential emotional reactions—envy and empathy—that, in turn, lead to illegal behavior. Envy toward wealthy customers and empathy toward those of similar economic status drive much of this illegal behavior (Gino and Pierce 2010). First, following Tang et al. (2011) and Lam and Hung’s (2005) arguments, then, Es is more highly related to unethical intentions for low-income individuals than for high-income people. Second, following the suggestion that it is the rich who behave more unethically than lower class individuals (Piffa et al. 2012), then, Es is more highly related to unethical intentions for high-income students than for their low-income counterparts. We favor the former rationale and explore these issues tentatively, below.

Hypothesis 3c

Extrinsic religiosity-social (Es) is positively related to unethical intentions for low-income students, but not for their high-income counterparts.

Panel Studies

Panel studies measure the same respondents at different points in time and may enable researchers to provide cause–effect relationships or strong external validity. We briefly searched the ISI Web of Knowledge data base using the term “panel” or “longitudinal.” As of June 20, 2012, out of 5,196 journal articles published (from 1982 to 2012) in Journal of Business Ethics, only 39 papers mentioned the term “longitudinal” and 13 articles used the term “panel.” Although we may have overlooked many articles without using these terms, panel/longitudinal studies are rare in ethics research.

Method

Procedure

Students completed surveys/activities for course credits at Time 1 and Time 2 (about 4 weeks apart) in classes confidentially with initials and the last four digits of their social security number on all surveys in order to match these two parts. This procedure avoids the possible impact of fatigue/memory, common method variance (CMV) bias, and enhances the psychological separation of predictors and criteria (Podsakoff et al. 2003). The first author collected data from 359 students (male = 216, 50.17 %; female = 139, 38.72 %; missing data = 4; return rate = 92 %) in an undergraduate statistics course that offered to psychology (n = 119) and business (n = 240) majors at two institutions (psychology/public and business/private, respectively) in the southeastern US for several years. The professor was blind regarding students’ survey results and debriefed the purposes of this study at the end of the semester.

Measures

We adopted Allport and Ross’ (1967) ROS involving 8-item intrinsic religiosity, 3-item extrinsic-social, and 3-item extrinsic-personal (Appendix) measured at Time 1, Machiavellianism (Mach IV, 4 items, two items from Tactics and two items from Views of Human Nature, Christie and Geis 1970) (measured at Time 2), and three sub-constructs (theft, corruption, and deception) of the 15-item, 5-factor PUB (Tang and Chen 2008) (measured at Time 2). For Religiosity and Machiavellianism measures, we employed a five-point scale with disagree strongly (1), disagree (2), neutral (3), agree (4), and agree strongly (5) as anchors. The unethical behavior intentions (PUB) is a measure of self-prediction with very low probability (1), low (2), average (3), high (4), and very high probability (5) as anchors and the following instructions: If you were given the opportunity in your work environment, what is the probability that you may engage in these activities (sample items): Abuse the company expense accounts and falsify accounting records; reveal company secrets for several million dollars; accept money, gifts, and kickbacks from others; and give customers “discounts” first and then secretively charge them more money later (bait and switch). We also collected demographic variables (sex, age, years of education, current job tenure, total work experience, and annual income), and many other filler items. Age varied between 18 and 44. The majority had work experiences with some missing data. We used Z score (M = $13,939.17) to classify high-income (n = 142; $43,577.43) and low-income groups (n = 217; $4,788.19).

Results

Descriptive Statistics

The means, standard deviations, Cronbach’s alpha, and correlations of variables for the whole sample are presented in Table 1. Among religiosity (I, Es, and Ep), only extrinsic-social (Es) was significantly related to extrinsic-personal (Ep). The three components of PUB—theft, corruption, and deception—were all significantly related to each other. Cronbach’s alphas were all higher than .70.
Table 1

Mean, standard deviation, Cronbach’s alpha, and correlations of major variables

Variable

M

SD

1

2

3

4

5

6

7

8

9

10

11

12

13

14

1. Age

20.48

3.45

              

2. Gender (%male)

.61

.49

.09

             

3. Education (years)

12.15

5.18

.05

.19**

            

4. Major (%Psy.)

.33

.47

.04

−.34**

.30**

           

5. Job (years)

2.16

2.28

.21**

.11

−.19**

−.15*

          

6. Work (years)

4.60

3.68

.51**

.15**

.10

−.03

.63**

         

7. Income ($)

13,939.17

61,511.57

.07

.05

.03

−.06

.22**

.10

        

8. Intrinsic (I)

3.66

.78

−.11

−.03

−.06

−.15**

−.13*

−.06

−.08

(.83)

      

9. Ext-social (Es)

1.89

.79

.05

.12*

−.14**

−.25**

.00

−.05

.08

.01

(.84)

     

10. Ext-personal (Ep)

2.92

.87

−.01

−.05

−.01

.03

.03

−.02

.02

.09

.28**

(.70)

    

11. Mach IV

2.30

.75

−.06

.22**

−.06

−.12*

.05

−.03

.01

−.20**

.11*

.04

(.80)

   

12. Theft

1.32

.58

.10

.21**

−.11

−.11*

.04

.09

.10

−.13*

.20**

.14*

.39**

(.85)

  

13. Corruption

1.42

.62

.06

.21**

−.02

−.10

.09

.06

.07

−.16**

.14*

.16**

.44**

.67**

(.74)

 

14. Deception

1.46

.63

.06

.25**

−.04

−.17**

.01

.06

.07

−.12*

.16**

.10

.43**

.63**

.70**

(.85)

NoteN = 359. Gender: Male = 1, Female = 0; Major: Psychology = 1, Business = 0; Education, current job experience, total work experience are expressed in years. Cronbach’s alpha is presented in a pair of parentheses

*p < .05, **p < .01

Age was significantly related to current job tenure, total work experience, and low intrinsic religiosity. Males had more education and total work experience, were business students, and had high levels of extrinsic-social, Machiavellianism, theft, corruption, and deception. Those psychology students had lower levels of job tenure, intrinsic religiosity, extrinsic-social, Machiavellianism, theft, and deception. Intrinsic religiosity was related to business students and low levels of job tenure, Machiavellianism, theft, corruption, and deception. Extrinsic-social was associated with gender (male), business students, a low level of education, and high levels of extrinsic-personal, Machiavellianism, theft, corruption, and deception. Extrinsic-personal was related to high levels of theft and corruption.

Measurement Model

We examined Allport and Ross’ (1967) ROS. Results suggested that the three-dimensional ROS model (I, Es, and Ep) (χ2 = 279.26, df = 74, p < .01, χ2/df = 3.77, IFI = .90, TLI = .86, CFI = .90, RMSEA = .09) was better than the two-dimensional model [I and (Es + Ep)] (χ2 = 517.04, df = 74, p < .01, χ2/df = 6.80, IFI = .79, TLI = .70, CFI = .78, RMSEA = .13). We have confidence in using the three-dimensional model and test our theoretical model in this study.

Common Method Variance

Due to our longitudinal data in our present study, common method variance (CMV) (Podsakoff et al. 2003; Spector 2006) should not be a concern. Following suggestions in the literature, we adopted Harman’s single-factor test and examined the unrotated factor solution involving 27 items of all three variables of interest in an exploratory factor analysis (EFA). We identified six factors, with eigenvalue greater than one. We listed the scale and amount of variance explained (Total = 67.27 %) below: PUB—theft, corruption, and deception (24.72 %), intrinsic religiosity (17.24 %), extrinsic-social (9.64 %), Machiavellianism (6.35 %), extrinsic-personal (5.59 %), and a factor with cross loadings (3.72 %), respectively. No single factor accounted for the majority of the covariance in the independent and criterion variables. We had separate and independent constructs.

The Whole Sample

Root mean square error of approximation (RMSEA) tends to over-reject a true model due to “small sample size” and “model complexity” (Tang et al. 2006, p. 446). In order to maintain a good sample size to item ratio and reduce model complexity for the whole sample and subsequent multiple-group analyses across subgroups of several variables, we established a parsimonious model using 15 parcels/items—3 parcels each for intrinsic religiosity, machiavellianism, and PUB (theft, corruption, and deception), and all 3 items for extrinsic-social and extrinsic-personal—instead of 27 indicators. The sample size to item ratio was 24 (359/15 = 23.93). We present our results for the whole sample below.

A Simplified Model

First of all, we examined the relationships between three dimensions of religiosity (I, Es, and Ep) and unethical behavior intentions by deleting Machiavellianism from our proposed theoretical model (χ2 = 149.25, df = 48, p < .01, χ2/df = 3.11, IFI = .94, NFI = .92, TLI = .91, CFI = .94, RMSEA = .08). Results of this simplified structural equation model (SEM) with only Paths 1, 2, and 3 of Fig. 1 suggested that intrinsic religiosity was negatively related to unethical intentions (Path 1 = −.15, p < .05), while both extrinsic-social (Path 2 = .28, p < .001) and extrinsic-personal (Path 3 = .14, p < .05) were positively related to unethical intentions. Paths 1, 2, and 3 were all significant, as predicted, supporting our Hypotheses 1a and 1b. Further, extrinsic-personal was significantly correlated with extrinsic-social (.31, p < .001) and intrinsic religiosity (.27, p < .001). However, intrinsic religiosity and extrinsic-social were not significantly related (.04). These results reported here are not unexpected and do not seem to offer anything new to the literature.

A Full Model

Results of our full model for the whole sample are presented in Fig. 22 = 232.98, df = 80, p < .01, χ2/df = 2.91, IFI = .93, TLI = .90, CFI = .93, RMSEA = .07). Figure 2 shows that with the addition of our mediator, Machiavellianism, Paths 2 and 3 were again significant, but Path 1 was not. Hypothesis 1b was supported, but Hypothesis 1a was not. Further, Paths 4, 5, and 7 were significant. Thus, extrinsic-social was directly (Path 3) and indirectly (Paths 5 and 7), while extrinsic-personal was directly, related to unethical behavior intentions. Finally, intrinsic religiosity was only indirectly related to unethical intentions (Paths 4 and 7). Results support Hypotheses 2a and 2b. Extrinsic-personal was significantly correlated with extrinsic-social and intrinsic religiosity.
https://static-content.springer.com/image/art%3A10.1007%2Fs10551-012-1407-2/MediaObjects/10551_2012_1407_Fig2_HTML.gif
Fig. 2

Results of the whole sample

Gender

Males

We investigated the same theoretical model across gender using a multi-group analysis (see Fig. 3; Table 2) and presented results for males first. An indirect path (Paths 4 and 7) contributed to the bright side of intrinsic religiosity. However, a direct path (Path 2) and an indirect path (Paths 6 and 7) facilitated the dark side of extrinsic religiosity.
https://static-content.springer.com/image/art%3A10.1007%2Fs10551-012-1407-2/MediaObjects/10551_2012_1407_Fig3_HTML.gif
Fig. 3

Results of our model across gender

Table 2

The effects of religiosity and machiavellianism on unethical intentions (PUB)

Variable

I

Es

Ep

Mach

Sum

The whole sample

 

 Standardized direct effect on Mach

−.19**

.21**

.00

  

 Standardized direct effect on PUB

−.05

.17**

.14*

.52***

 

 Standardized indirect effect on PUB

−.10

.11

.00

  

 Standardized total effect on PUB

−.15

.28

.14

.52

.79

Gender

 The male sample

  Standardized direct effect on Mach

−.21*

.01

.25*

  

  Standardized direct effect on PUB

−.08

.19*

.18

.53***

 

  Standardized indirect effect on PUB

−.11

.01

.13

  

  Standardized total effect on PUB

−.19

.20

.31

.53

.85

 The female sample

  Standardized direct effect on Mach

−.20*

.38***

−.21*

  

  Standardized direct effect on PUB

−.10

.11

.05

.36**

 

  Standardized indirect effect on PUB

−.07

.14

−.08

  

  Standardized total effect on PUB

−.17

.25

−.02

.36

.42

Major

 Psychology

  Standardized direct effect on Mach

−.16

.24*

−.16

  

  Standardized direct effect on PUB

−.16

−.06

.16

.57***

 

  Standardized indirect effect on PUB

−.09

.14

−.09

  

  Standardized total effect on PUB

−.26

.08

.07

.57

.46

 Business

  Standardized direct effect on Mach

−.21*

.08

.18

  

  Standardized direct effect on PUB

−.01

.20**

.12

.48***

 

  Standardized indirect effect on PUB

−.10

.04

.09

  

  Standardized total effect on PUB

−.11

.24

.21

.48

.82

Income

 The high-income sample

   Standardized direct effect on Mach

−.31

.12

.21

  

   Standardized direct effect on PUB

.15

.08

−.21

.74***

 

   Standardized indirect effect on PUB

−.23

.14

.16

  

   Standardized total effect on PUB

−.08

.22

−.05

.74

.83

 The low-income sample

   Standardized direct effect on Mach

−.17*

.22**

−.11

  

   Standardized direct effect on PUB

−.15*

.26***

.32***

.38***

 

   Standardized indirect effect on PUB

−.06

.08

−.04

  

   Standardized total effect on PUB

−.22

.35

.28

.38

.79

Note: Direct Effect + Indirect Effect = Total Effect (minor differences due to rounding)

*p < .05, **p < .01, ***p < .001

Females

There were several unique findings for females. First, all three indirect paths—Paths 4, 5, and 6 and Path 7—were significant. All the three direct paths (Paths 1, 2, and 3) were non-significant. Results (Path 2) supported Hypothesis 3a. Further, among the three indirect paths, two promoted the bright side; only one aggravated the dark side of religiosity. Specifically, Path 6, from extrinsic-personal (Ep) to Machiavellianism, was significant and negative, contributing to the positive impact or the bright side of religiosity. Furthermore, this was the only significant path that was “negative” (creating the bright side) for females (Path 6 = −.21, p < .05), but “positive” (the dark side) for males (Path 6 = .25, p < .05). These results may offer possible explanations for the non-significant Path 6 (.00) for the whole sample—the negative path for females and the positive path for males canceled each other out. In addition, Path 5 was significant for females, but non-significant for males. Path 2 was non-significant for females, but significant for males. The correlation between extrinsic-personal (Ep) and extrinsic-social (Es) was not significant for females, but significant for males which contributed to these novel findings for females. The Ep and Es constructs provide different meanings for males and females.

MANOVA

We compared participants’ demographic variables across gender using a multivariate analysis of variance (MANOVA) [F(6, 217) = 12.68, p < .001, Wilks’ lambda = .740, partial eta squared = .260, power = 1.00]. Males had higher education (13.00 vs. 11.62), longer job tenure (2.45 vs. 1.71), total work experience (5.23 vs. 3.86), and higher percentage in business majors (.67 vs. .27) than females. In our second MANOVA, the differences in our major variables across gender were also significant [F(7, 327) = 4.86, p < .001, Wilks’ lambda = .905, partial eta squared = .095, power = .996]. Males had higher Machiavellianism (2.43 vs. 2.09), theft (1.43 vs. 1.18), corruption (1.54 vs. 1.27), and deception (1.59 vs. 1.26) than females.

Major

Psychology

Only one indirect path (Paths 5 and 7) contributed to the dark side of religiosity (Es) (Fig. 4; Table 2). Other paths failed to reach significance. Extrinsic-personal (Ep) was significantly correlated with intrinsic religiosity (I) (.48).
https://static-content.springer.com/image/art%3A10.1007%2Fs10551-012-1407-2/MediaObjects/10551_2012_1407_Fig4_HTML.gif
Fig. 4

Results of our model across major

Business

One indirect path (Paths 4 and 7) contributed to the bright side of religiosity (I). One direct path (Path 2) worsened the dark side of extrinsic religiosity (Es). Path 5—from extrinsic-social to Machiavellianism—was significant and positive for psychology participants, but non-significant for business students. Path 2 was significant for business students, but non-significant for psychology majors, supporting Hypothesis 3b.

MANOVA

Significant demographic variables across major [F(6, 217) = 16.78, p < .001, Wilks’ lambda = .683, partial eta squared = .317, power = 1.00] showed that business students were mostly male (.76 vs. .37), had lower education (11.04 vs. 14.27), and longer job experience (2.50 vs. 1.69) than those psychology students. The differences in our major variables across major were also significant [F(7, 327) = 6.73, p < .001, Wilks’ lambda = .874, partial eta squared = .126, power = 1.00]. Business students had higher intrinsic religiosity (3.73 vs. 3.50), extrinsic-social (2.10 vs. 1.61), Machiavellianism (2.36 vs. 2.17), theft (1.37 vs. 1.24), and deception (1.53 vs. 1.31) than their psychology counterparts. Finally, results of crosstabs (gender × major) suggested that there were more males majoring in business and more females majoring in psychology (χ2 = 39.85, p < .001).

Income

High Income

Only one significant path (Path 7) contributed to the unethical behavior intentions (Fig. 5), i.e., from Machiavellianism to unethical behavior intentions. Other paths failed to reach significance. The negative path between extrinsic-personal (Ep) and unethical intentions only approached significance (−.21, p = .058), suggesting extrinsic-personal may slightly contribute to the bright side of religiosity for those participants in the high-income group.
https://static-content.springer.com/image/art%3A10.1007%2Fs10551-012-1407-2/MediaObjects/10551_2012_1407_Fig5_HTML.gif
Fig. 5

Results of our model across income

Low Income

All three direct paths (Paths 1, 2, and 3) were significant, supporting Hypothesis 3c. Among three indirect paths, two were significant (Paths 4 and 5 and Path 7) contributed to the bright and the dark sides of religiosity. Path 3 (Ep) was significant and positive—the only significant path (Path 3) among all of our subsequent analyses. Religiosity has the least impact on unethical behavior intentions for the high-income group, but the most impact for the low-income group—all, but Path 6, were significant.

MANOVA

Participants’ demographic variables across income was significant [F(6, 217) = 18.51, p < .001, Wilks’ lambda = .625, partial eta squared = .375, power = 1.00]. Students in the high-income group were older (22.87 vs. 19.72), had longer experience on the current job (3.24 vs. 1.79), total work experience (6.96 vs. 3.93), and income9 ($29,648.87 vs. $4,344.01) than those low-income participants. Most in the low-income group were working on part-time jobs. Among high-income students, 71.1 % of them were business majors, whereas 28.9 % were psychology majors. No significant differences among variables across income levels existed [F(7, 327) = 1.75, p < .097, Wilks’ lambda = .964, partial eta squared = .036, power = .709].

Impact

Table 2 shows the results of standardized direct impact, indirect impact, and total impact of I, Es, and Ep, and Machiavellianism on unethical intentions (PUB). For the whole sample, the total impact of I, Es, Ep, and Machiavellianism on unethical intentions was −.15, .28, .14, and .52, respectively. Thus, intrinsic religiosity curbs unethical intentions (−.15), whereas extrinsic-social and extrinsic-personal promote unethical intentions (.28 and .14, respectively). The effect of Machiavellianism on unethical intentions (PUB) was the strongest one among all predictors in this study (.52). The good news is that there is a significant indirect path for bright side of religiosity. The bad news is that there are two direct paths and one indirect path for the dark side of extrinsic religiosity. After balancing the bright and the dark sides of religiosity, there is an overall dark impact (total impact = .79).

We summarize our multi-group findings cross gender, major, and income, below. On the one hand, intrinsic religiosity (I) had the strongest bright and positive impact in minimizing unethical intentions for psychology students (−.26) and the weakest for students in the high-income group (−.08). On the other hand, extrinsic-social (Es) had the strongest dark and damaging impact in facilitating unethical intentions for students in the low-income group (.35) and the weakest for psychology students (.08). For extrinsic-personal (Ep), the most potent dark and detrimental impact was for male students (.31). It is interesting to note that extrinsic-personal (Ep) also had its bright and optimistic impact in reducing unethical intentions for high-income students (−.05) and females (−.02). The dark and harmful impact of Machiavellianism was the strongest for high-income students (.74) and weakest for females (.36). When intrinsic religiosity and extrinsic religiosity were combined to form one single score, the overall total impacts were all positive. The positive overall score means that religiosity, unfortunately, has a dark and negative/bleak impact on unethical intentions—facilitating it rather than impeding it. The smallest dark impact existed for females (.42) and largest for males (.85) and high-income individuals (.83). The differences between males and females (.85 vs. .42) and between psychology students and business students (.46 vs. .82) seemed to be bigger than that between high- and low-income individuals (.83 vs. .79).

Discussion

We develop a theoretical model involving religiosity [intrinsic (I), extrinsic-social (Es), and extrinsic-personal (Ep), Time 1], Machiavellianism (Time 2), and propensity to engage in unethical behavior (theft, corruption, and deception, Time 2), and investigate direct and indirect paths among constructs based on panel data collected from 359 students. This research provides the following theoretical, empirical, and practical contributions to the literature. We turn to theoretical contributions first.

Theoretical Contributions

We reveal the importance of using our new theoretical model by treating Machiavellianism as a mediator and investigating not only the whole sample but also the subsequent multi-group analyses across gender, major, and income. Results of this study provide much more insights than the direct relationships between religiosity (I, Es, and Ep) and unethical behavior intentions. For the whole sample, the bright side of religiosity in reducing unethical intentions is indirect. For the dark side of religiosity, extrinsic-social motives reveal both direct and indirect impacts, whereas extrinsic-personal has only a direct impact in promoting unethical intentions. The standardized total effect shows that religiosity facilitates unethical intentions. The dark side is stronger than the bright side.

In a sample of business students, Tang and Tang (2010) found that intrinsic religiosity deterred unethical intentions not only directly, but also indirectly through the absence of Machiavellianism. In our present study, we partially replicate the bright side of religiosity’s indirect effect in reducing unethical intentions for the whole sample. In our simplified model, when we investigate three components of religiosity without the mediator, the intrinsic religiosity is also directly related to a low level of unethical behavior intentions. These results of the present study replicated that of Randolph-Seng and Nielsen (2007), Rettinger and Jordan (2005), Singhapakdi et al. (2012), Tang and Tang (2010), Vitell et al. (2006), and Vitell et al. (2007). Due to different samples and theoretical models used in Tang and Tang (2010) and the present study (business and psychology students from different universities), it appears that intrinsic religiosity’s bright and positive impact on restraining unethical intentions is more indirect than direct. We provide additional insights based on our multi-group analyses across gender, major, and income, below.

For males, the bright side for intrinsic religiosity is indirect, whereas the dark side is direct for extrinsic-social (Es) and indirect for extrinsic-personal (Ep), creating a strong dark impact in exacerbating unethical intentions. For females, first of all, all three impacts (I, Es, and Ep) are indirect. Second, there are two bright indirect paths to low Machiavellianism: one from intrinsic religiosity (I) and the other from extrinsic-personal (Ep), but only one dark indirect impact from extrinsic-personal (Es) to high Machiavellianism. Third, among all analyses in our present study, females have the lowest level of overall unethical intentions—the least amount of dark impact from religiosity to unethical intentions. Fourth, another novel finding is that the significant path from extrinsic-personal (Ep) to Machiavellianism is “negative”—creating the bright impact for females, but “positive”—creating the dark impact for males. That is, the extrinsic-personal (Ep) to Machiavellianism relationships are exactly the opposite for males and females which contribute to the non-significant finding for the whole sample. When “peace” is with the females, extrinsic-personal (Ep) undermines unethical intentions. Fifth, the dark impact for male students (.85) is stronger than female students (.42).

Psychology students have only one indirect path from extrinsic-social (Es) to unethical intentions. For business students, the bright side of intrinsic religiosity is indirect, whereas the dark side of extrinsic-social is direct. Since most psychology students are female and business students are male, these patterns are similar to (but slightly different from) that of females and males, mentioned above. The dark impact for business students (.82) is stronger than psychology students (.46).

For high-income individuals with mostly full-time jobs, all paths are non-significant, except one—from Machiavellianism to unethical intentions. This path is the strongest path (.74) among all analyses. In addition, the non-significant path from extrinsic-personal (Ep) to Machiavellianism is “negative”—creating a minorbright impact for high-income students. For low-income students, all paths, except one (Path 6), are significant, suggesting that religiosity has the most (bright and dark) impacts on unethical intentions. Regardless of these differences in paths, the overall negative impacts of religiosity on unethical intentions are very similar for high- (.83) and low-income (.79) participants. These findings deserve researchers’ attention in future empirical studies.

Our assertion—high religiosity leads to high-unethical intents—is paradoxical. We offer our explanations briefly below to explain our findings and theoretical implications. We adopt ROS (Allport and Ross 1967) with three dimensions: intrinsic religiosity (I), extrinsic religiosity-social (Es), and extrinsic religiosity-personal (Ep) (Flere et al. 2008). These three dimensions have different patterns of impact on unethical intentions. The overall impact will be negative due to one positive force (I) and two negative forces (Es and Ep) in our theoretical model.

The good news is that ROS captures both bright and dark sides of religiosity. The bad news is that “love your neighbor,” a strong religious belief, has been turned to “use your neighbor,” unfortunately. ROS includes only one bright side of religiosity. We suspect that this is one of the major reasons why most researchers have selected only the intrinsic religiosity aspect from the ROS in conducting empirical research. In fact, intrinsic religiosity is a persistent inhibiter of crime (Evans et al. 1995), deviance (Kerley et al. 2011), violent behavior, and unethical intentions. Future researchers of religiosity may want to incorporate not only ROS but also other constructs, e.g., “love your God,” “love your neighbor,” “love one another,” “love your enemies,” and “peace” to create a more holistic, comprehensive, and complete construct of religiosity (e.g., see Hill and Hood 1999).

Empirical Contributions

Based on business and psychology students with work experiences in the southeastern US, we tap on an important topic in the business ethics literature. Our carefully selected constructs, theoretical model, samples, and data analysis strategies fit the purposes of this study and provide some interesting, novel, and counterintuitive results. Future researchers have confidence in replicating our theoretical model in other contexts. Our counterintuitive and novel discoveries (Bartunek et al. 2006) are impossible to achieve without a reasonable sample size and several demographic variables.

Practical Contributions

Public and private universities do not always have a formal code of ethics and a high dosage of religion in their academic curriculum. This may lead to a decline of traditional religious values and result in anomie—the weakening of ethical norms and increased rates of deviance (Cullen et al. 2004). These 20-year-old students and part-time/full-time workers function in a value-free world dominated by secular values, materialism, and consumption, want to do whatever they please, dig in, and fight against ethical (i.e., legitimacy crisis, Rynes et al. 2003), religious, or work-related values (Tang and Baumeister 1984), and may have the “I do not know and I do not care” attitude. They may start with something real trivial. Inch by inch, they dig deeper and deeper into a hole of which they cannot get out (Burton 2004).

In this study, students from the Bible belt are highly influenced by intrinsic religiosity. These young workers and college students will become future full-time employees and managers in business organizations. It is not the lack of intelligence or brains, but the lack of “wisdom” (Feiner 2004, p. 85), “virtue” (Wright and Goodstein 2007), or integrity and character (Tang and Liu 2012) that caused many recent scandals in the US. We need to be aware of (1) the gaps between what they know and what they actually do, between head (intelligence) and heart (wisdom or virtue) and (2) what they have done (theft, corruption, and deception) and what they have failed to do.

Ancient philosopher Socrates argued almost 2,500 years ago that ethics consists of knowing what we ought to do and that such knowledge can be taught. Can professors make undisciplined students in an education system dominated by secular values virtuous? What can we do so that today’s students will not be tomorrow’s criminals? Does business education foster critters with lopsided brains, icy hearts, and shrunken souls (Leavitt 1989)?

Students bring dispositional values (Staw et al. 1986), such as religiosity, to the university. Psychology students and females are more ethical than business students and males, due to their strong dispositions, personal values, and the ASA process in their social, academic contexts. Although we cannot change people’s religiosity and unethical behavior overnight, repetition may reinforce and crystallize religious and ethical values in their academic journey. Following the Parable of the Lost Sheep: “And when he finds it, he joyfully puts it on his shoulders and goes home” (Luke 15: pp. 4–7, emphasis added). An important implication is that the lost sheep may not return home by itself and will need shepherd’s help. Professors/executives literally need to guide these students/managers, capture the window of opportunity, and enhance ethical values in schools and organizations. The joy of teaching is to bring the lost sheep back and make it aware: I was blind, but now I see.

Managers cannot be created in a classroom and should learn from their own experiences. Professors cannot teach management to people who are not managers (Mintzberg and Gosling 2002). Some full-time MBA students are required to visit federal prisons and interview white-collar criminals who are paying their dues to society—often for cooking the books (Kercheval 2004; Merritt 2004). Universities, business deans, and AACSB-International need to revamp the business curriculum, create a strong culture of making ethical decisions, invest in ethics education, support research on ethics and religiosity, and satisfy all stakeholders in society (business, students, media, AACSB-International, and business school) (Pfeffer and Fong 2002).

What Can I (We) Do?

In the progressively globalized economy and ever-increasing competition, managers and scholars need strong courage to face grand challenges in displaying virtuous integrity, character, and honesty and making ethical decisions. Religion may be one of the possible last resorts for teaching business ethics and promoting ethical decision making. For example, it is “natural” to tell the truth and “unnatural” to tell a lie (Heney 2012). Reciting the Ten Commandments and/or starting a new day with a prayer in the morning or in a business meeting may have the potential to set an ethical tenor for the event/day, enhance corporate ethical cultures, and reduce managers’ unethical behavior intentions in organizations. A sea change of the ethical social norm in schools, organizations, and society, or ethical community-building (McCabe et al. 2006), is needed to fight against unethical behaviors.

Educators, managers, and average citizens may simply adopt the following four ways to start this grand challenge by (1) praying a little more to develop a deep conversation with our God—8 min. in the morning and 8 min. in the evening, per day, (2) studying the faith and reading the Bible more—five pages a day, (3) giving a little more of ourselves and donating generously—1 % or 2 % more than before to the church or charity, and (4) sharing the truth a little more and becoming an evangelist—to one more person a day than before (Sappenfield 2012). The first two deal with “love your God” and the latter two “love your neighbor.” Overall, it is important to “love one another” and “love your enemies.”

Although it is extremely difficult to achieve such a grand challenge, we can start with something small. First, it is like a mustard seed that, when it is sown in the ground, is the smallest of all the seeds on the earth. But once it is sown, it springs up and becomes the largest of plants.10 Second, a journey of a thousand miles begins with a single step. Most people (like Moses) “question” their own competencies and “are afraid” to take the first single step: Who am I? What am I to tell them? The Lord said to Moses and us: “Who gives one man speech and makes another deaf and dumb? Or who gives sight to one and makes another blind? Is it not I, the Lord? Go, then! It is I who will assist you in speaking and will teach you what you are to say.”11 Third, we must write our goals down,12 make it visible, have a sense of “urgency,” and do it “now” (Kotter 2008; Latham et al. 2010). From Jesus to his 12 disciples, Christianity has expanded to a third of the world’s population (2.2 billion) and is the world’s largest religion. For example, although Mother Teresa was small in stature, 4′10″ tall, her big goal was to give her life to helping the poorest of the poor. As the recipient of the 1979 Nobel Peace Prize, she refused the conventional ceremonial banquet given to laureates, but instead, she asked that the $192,000 funds be given to the poor in India. She became the fearless beacon of life and hope.

As James (1902) observed, a lot of good people are not religious. They have moral codes based on moral theories that do not include any sort of religious beliefs, god, worship, rituals, prayer, etc. Similar to the Ten Commandments in a Judeo-Christian tradition, ethical values and hyper-norms in other religions, cultures, and countries may incite comparable emotions and behavior tendencies in most people (cf. Kriger and Seng 2005; Liu and Tang 2011; Tang 2012; Tang et al. 2011). This applies to ethical behavior intentions. Future research may test these propositions empirically.

Limitation

Our student sample (with 4.60 years of total work experience) may raise the issue of external validity. Students major in business or psychology due to their self-selection, but not due to random assignment by researchers. Our convenience sample may not represent all universities or the specific (business vs. psychology) disciplines. We measure only students’ behavioral intentions, but not actual unethical behaviors which may be verified in laboratory experiments (Ariely 2008). Researchers need to be aware of the monumental gap between intent and behavior. Our multiple-wave panel data collected in one semester may not provide a strong cause-and-effect relationship. Although longitudinal research lasting for years or decades is highly desirable, the mortality issue, however, may hurt the internal validity of the study. Our well-developed instruments with proven psychometric properties, time lags between the predictor and criterion variables, confidentiality protection, and Harman’s single-factor test provide evidence that common method variance is not a concern. Researchers may include additional attitudinal, personality variables (e.g., social desirability scale and ASPIRE; Tang and Liu 2012), and qualitative data, explore students and managers in different institutions, regions, and cultures, and investigate the generalizability our theoretical model to other contexts.

Conclusion

We develop a theoretical model involving religiosity, Machiavellianism, and unethical intentions and investigate direct and indirect paths based on panel data collected from 359 students. For the whole sample, intrinsic religiosity (I) indirectly reduces unethical intentions through the absence of Machiavellianism, the bright side of religiosity. Both extrinsic-social (Es) and extrinsic-personal (Ep) directly, while extrinsic-social (Es) indirectly, exacerbate unethical intentions, the dark side of religiosity. Multiple-group analyses across gender, college major, and income show that the bright side exists directly for low-income students, but indirectly for males and females, business majors, and low-income students. Ep undermines unethical intentions indirectly for females, a novel, but not unexpected finding of this research. For the dark side, Es incites unethical intentions directly for males, business students, and low-income individuals, but indirectly for females, psychology majors, and low-income people. The Machiavellianism–unethical intentions relationship is the strongest for high-income participants. Religiosity has the highest number of significant paths for low-income individuals and the strongest dark side for males and high-income students, but the weakest for females. Our novel, original findings foster theory development and testing, add new vocabulary to the conversation of religiosity and unethical intentions, and improve practice.

Footnotes
1

Love the Lord your God with all your heart. Love your neighbor (Mark 12: 30). Love one another (John 13: 34). Love your enemies (Matthew 5: 44).

 
2

We “aspire” to please him (2 Corinthians 5: 9).

 
3

They acted with complete honesty (2 Kings 12:15).

 
4

The rash man has no integrity; but the just man, because of his faith, shall live (Habakkuk 2: 4).

 
5

Then he poured water into a basin and began to wash the disciples’ feet and dry them with the towel around his waist…. Peter said to him, “You will never wash my feet.” Jesus answered him, “Unless I wash you, you will have no inheritance with me” (John 13: 5, 8).

 
6

Trust in the LORD with all your heart (Proverbs 3: 5).

 
7

It is I who will assist you in speaking and will teach you what you are to say (Exodus 4: 12).

 
8

Peace be with you (John 20: 21).

 
9

Sample size was smaller in MANOVA due to the missing data of seven demographic variables.

 
10

Mark 4:31–32.

 
11

Exodus 4: 11–13.

 
12

Write down the vision clearly upon the tablets, so that one can read it readily (Habakkuk, 2: 2).

 

Acknowledgments

The authors would like to thank late Fr. Wiatt Funk, Fr. Mark Sappenfield (St. Rose of Lima Catholic Church, Murfreesboro, TN), and Fr. Dave Heney (St. Paschal Baylon Catholic Church, Thousand Oaks, CA) for their inspiration.

Copyright information

© Springer Science+Business Media B.V. 2012