- Email: [email protected]
Retailers' major ethical decision making constructs
Available online at www.sciencedirect.com
Journal of Business Research 61 (2008) 834 – 841
Retailers' major ethical decision making constructs John Fraedrich a,⁎, Rajesh Iyer b,1 a
Southern Illinois University at Carbondale, United States b Bradley University, United States
Received 1 June 2005; received in revised form 1 March 2006; accepted 1 August 2007
Abstract The major constructs of the Ferrell and Gresham [Ferrell OC, Gresham L. A Contingency Framework for Understanding Ethical DecisionMaking in Marketing. J Mark 1985; (Summer): 87–96.], Hunt and Vitell [Hunt SD, Vitell S. A General Theory of Marketing Ethics. J Macromark 1986; 6 (Spring): 5–16.], and Ferrell et al. [Ferrell OC, Gresham L, Fraedrich JP. A Synthesis of Ethical Decision Models for Marketing. J Macromark 1989; 9(2) (Fall): 55–64.] models were tested using LISREL. Our findings suggest that Friends, Superiors, Business Associates, and Formalization of the Organization are more significant than Moral Philosophy. A new construct Ethicalness, comprised of individual, organizational, and societal value structures yielded a tight multi-dimensional fit that supports Kerlinger's domain theory and is dissimilar to Reidenbach and Robin's (1991) multi-dimensional scale. This research is one of the first to test as many ethical decision making constructs while explaining the multi-dimensional conundrum of ethical/unethical and legal/illegal practices within retailing. © 2007 Elsevier Inc. All rights reserved. Keywords: Retailers; Ethics; Ethicalness; Moral philosophy; Significant others
1. Introduction Since the late 1980′s ethics has become an important modeling aspect. AACSB has reiterated the importance of ethics by their accreditation formula as well as a plethora of research models, books, courses, and ethics centers. The rationale is simple: “Reinforcing ethical behavior is important for improving performance and achieving success in the marketplace” (Gundlach and Murphy, 1993, p. 44). But to improve performance, academia and practitioners must determine the empirical strength of the constructs that purport to be important within ethical decision making. Marketing is the academic leader, being first to develop ethical decision making models (Ferrell and Gresham, 1985; Hunt and Vitell, 1986; Ferrell et al., 1989; Malhotra and Miller, ⁎ Corresponding author. Department of Marketing, College of Business and Administration, Southern Illinois University Carbondale, Rehn Hall 229, 1025 Lincoln Drive, Carbondale, IL 62901, United States. Tel.: +1 618 453 4341. E-mail addresses: [email protected] (J. Fraedrich), [email protected] (R. Iyer). 1 Tel.: +1 309 6772991. 0148-2963/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2007.08.006
1998). Most research within the ethics literature in other disciplines is based on these models (Trevino, 1986; Jones and Ryan, 1997; Banerjeeet al., 1998; Brass, Butterfield, and Skaggs, 1998; Upchurch, 1998). The past decade points to the constructs of opportunity in the work environment, moral philosophy/values, peers and superiors, as well as individual factors facilitating ethical/unethical determination of business decisions. Yet, no one has attempted to take these key constructs and determine their cumulative effect on a retailing sample. This article is a first attempt. A second issue within the ethics stream is quantifying the concept, “ethical/unethical.” A large percentage of research takes ethical scenarios or questionable practices and asks whether or not the action is relatively ethical, right, fair, or moral. The premise is ethics is relative to an individual or group perspective. What one defines as right/ethical therefore becomes reality. The other argument is side-stepped concerning absolute truth from a non-relativist approach. The result is a plethora of articles explaining what is ethical/unethical from the marketer's perspective. Compounding the problem is the relativity of time on perception. The solution is a measure that is multi-dimensional with an ontological perspective that does
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
not originate with the marketer. In other words, once the decision is made how does the individual, business unit and society perceive the action? Pursuant to this question, three domains are identified, tested, and overall paths and associative strengths to moral philosophy, peers and superiors, and other individual factors determined. The sections that follow identify the significant constructs and hypotheses, define ethicalness, and analyze a model using LISREL. 2. Ethics models, constructs and ethicalness Several ethics models are generally accepted (Ferrell and Gresham, 1985; Hunt and Vitell, 1986; Trevino, 1986; Ferrel and Skinner, 1988) with Vitell and Hunt (1990) and Carlson and Kacmar's (1997) being the only ones to partially test a model. In Carlson and Kacmar, moral philosophy was the best predictor on 123 students; but, construct consistency was not tested. Theories associated with ethics models are differential association, cognitive consistency, Fishbein's AIB, as well as Piaget's learning theories with key constructs being opportunity, significant others, and individual factors (See Fig. 1).
835
Opportunity (Bartels, 1967) results from “a favorable set of conditions to limit barriers or provide rewards” (Ferrell and Gresham 1985, p. 92) and is theoretically and somewhat empirically linked to ethical decision making (O'Fallon and Butterfield, 2005; Powpaka, 2002). In Powpaka's structural analysis of bribery and Thai managers, opportunity was defined relative to corporate and cultural circumstances that allow bribes to be given. Barriers to commit unethical actions (negative) or rewards (intrinsic and/or extrinsic) for not committing such actions make up the opportunity construct. The increase or decrease of opportunity within business is affected by the formalization of the organization (centralized/decentralized), the retailer's acceptance of authority within the organizational structure, organizational controls (rules), and situation familiarity. To reduce the opportunity (formalization, controls, authority acceptance, and situation familiarity) to commit unethical or illegal actions, the health care industry hired in-house police to help minimize potential fraud in billing because of Columbia/HCA Health Care Corporation's compliance problems with the Federal Government (Kirkpatrick, 1997). Associating with opportunity are formalization of the organization, acceptance of authority, organizational controls, and
Fig. 1. Hypothesized model of ethicalness.
836
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
familiarity. Formalization of the organization is the standardizing or codifying of operations and procedures within a firm. Several studies link the formalization of bureaucracy to improved ethical conduct (Fritzsche and Becker, 1983; Hunt, Chonko, and Wilcox, 1984; Ferrell and Skinner, 1988; Singhapakdi and Vitell, 1990; Akaah, 1993). Acceptance of authority is in Ferrell and Skinner's (1988) centralization construct and associated with compliance of subordinates to a superior/company/society's legitimate right to exercise control (Grimes, 1978). Compliance comes into the ethics equation regardless of whether or not the employee agrees with the superior. The acceptance of such authority should theoretically increase or play a part in the ethical decision making process. Organizational controls include target setting, measuring or monitoring, and feedback (Ouchi, 1977). Ethics controls are operationalized by setting up rules, regulations, and internal procedures and linked to ethical behavior (Ferrell and Skinner, 1988; Thorne-LeClair et al., 1998). Familiarity relates to how routinized a process/action becomes. Familiar situations should, through cognitive consistency theory (Heider, 1958), be regarded as ethical to retailers. Cognitive consistency theory states people attempt to reach a state of decision equilibrium. Familiarity may be akin to ethical sensitivity in that the sensitivity construct measures the ethical/moral value awareness of the situation that changes the weights in the decision process (Sparks and Hunt, 1998). While many support the theory, Fraedrich and Ferrell (1992) found an anomaly that some teleological types say one thing but do something else that subjects define as unethical. Within the model all four factors relate to opportunity positively. H1. Opportunity relate positively to ethicalness. H1a. Formalization of the organization relate positively to opportunity. H1b. Acceptance of Authority relate positively to opportunity. H1c. Organizational controls relate positively to opportunity. H1d. Familiarity of the situation relate positively to opportunity. Significant other is based, in part, on differential association (Sutherland and Crissey, 1970) and role-set theories (Merton, 1957). Differential association states that people are influenced more by those closest to them whether spatially or perceptually. Role-set theory refers to the complement of role relationships which focal persons have by virtue of their social status in an organization (Ferrell and Gresham, 1985, p.91). For example, differential association explains why a person's friends or family have an influence on business-related decisions. Role-set theory explains why, through authority figures such as supervisors or the firm itself, such entities influence the ethical decision process. Zey-Ferrell et al. (1979) and others (Zey-Ferrell and Ferrell, 1982; Hunt et al., 1984; Cunningham, 1991) show a link between business associates and one's superior and their effect on the ethical decision process. In essence if the superior or associates are prone to unethical behavior, one's own decision process is affected in the negative. The
adage, lie down with dogs get up with fleas, corresponds to what may be happening in the business arena. In addition, it appears that friends outside the firm as well as corporate culture may also affect the ethical decision process. H2. Significant others relate positively to ethicalness. H2a. Friends relate positively to significant others. H2b. Business Associates relates positively significant others. H2c. Superiors relate positively significant others. Individual factors represent an individual's philosophical process. The heart of ethical decision making revolves around moral philosophies; specifically deontology and teleology. Current models suggest these two philosophical approaches are used in resolving ethical dilemmas (Ferrell and Gresham, 1985; Hunt and Vitell, 1986; Trevino, 1986; Ferrell et al., 1989) and an empirical link between philosophies and ethical behavior has also been established (Fraedrich, 1988; Ferrell and Fraedrich, 1991; Fraedrich and Ferrell, 1992; 1992a; DeConinck and Lewis, 1997). Deontology relates to philosophies centered on individual rights such as Ross' prima facie duties. Teleology stresses end-states, consequences, and group action above the individual's rights. Religiosity is an individual factor not in research reports relative to ethical decision. Religiosity relates to a person's relationship with a supreme being and is recognized as a force in consumer behavior (McDaniel and Burnett, 1990; 1991; Terpstra et al., 1993; Barnett et al., 1996; Wilkes et al., 1986) and may be a legitimate construct in business ethics (Fort, 1977). Recent research suggests that an increased level of religiosity helps people recognize ethical problems, can positively affect their judgment, and religious people are less willing to engage in unethical behavior (Kennedy and Lawton, 1996; Tse and Au, 1997; Singhapakdi et al., 2000; Razzaque and Hwee, 2002). However, many measure religiosity in differently and Giacalone and Jurkiewicz (2003) report that spirituality does not correlate well with religiosity. Demographic factors such as marital status and age are associated with ethical decision making (Loe, Ferrell and Mansfield, 2000; O'Fallon and Butterfield, 2005). For example, older individuals tend to be perceived as more ethical (Tepstra et al., 1993). H3. Individual factors relate positively to ethicalness. H3a. Moral philosophy type relates positively to individual factors. H3b. Age relates positively to individual factors. H3c. Religiosity relate positively to individual factors. When doing ethics research a large problem is the definition as to what is ethical. The relativist argument from Socratic times states that whatever one defines as right or ethical therefore becomes such. Borrowing from sociology, Kerlinger (1986) answers the relativist's argument with domain theory. Domain theory argues that one needs to view the decision relative to the
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
domain, area, or level of experience that a particular society has attained. In ethics the different domains are the ethical perceptions of the individual, organization, and society. Under cognitive consistency all should be correlated. Krugman and Ferrell (1981) allude to the dimensionality of the ethics construct with their differences between what was individually perceived as ethical and what friends and business associates believed. Singer and Singer (1997) also advocate a multi-dimensional construct as the incorporation of social and ethical concerns within a utilitarian decision structure. No research to date has simultaneously measured these three domains (individual, organization, and society). 3. Method 3.1. Sample, measures and procedure Retail managers/salespeople were chosen because of their boundary-spanning role in their organization. The questionnaire was pre-tested on business personnel and word changes in some of the statements were made to reduce comprehension confusion in our retail sample. An expert panel reviewed the changes and deemed that they did not significantly alter any item meanings. The revised questionnaire was sent to a major retail conglomerate (700 managers) resulting in 186 usable returns (26.6% rate). Because of the sensitivity of the topic and restrictions from the sample corporation, it was impossible to obtain a comparison sample of non-respondents. Three validated vignettes were used (betraying a trust, bribery, and income tax evasion) (Fritzsche and Becker, 1983; Fraedrich, 1988; Fraedrich and Ferrell, 1992). Scales adapted by John (1984) measured the opportunity via formalization of the organization and organizational controls. All items were 7 point Likert-type with several reversed scored. Acceptance of authority was measured by Withey's (1965) scale measuring the extent an individual tends to obey orders and respect authority. Separate statements associated with each vignette asked if the manager had ever come across this type of problem (situation familiarity). Moral philosophy was measured using modified statements from Boyce and Jensen's (1978) Moral Content Test (MCT) across three vignettes. Retail managers were specifically asked which statement or philosophy best explained the rationale for their decision. The forced choice helps respondents distinguish between salient and non-salient factors. Due to respondent time constraints, a multi-item philosophical scale for each vignette was inappropriate, as it would have jeopardized the study. Religiosity was measured both cognitively and behaviorally using McDaniel and Burnett's (1991) three-item scale. Ethicalness was measured by asking respondents about the perceived ethicalness of their decision (individual). Next, a modified version of Ferrell and Skinner's (1988) ethical behavior scale was used for the organizational perspective (Fraedrich, 1993). Societal ethics was measured by taking each vignette and determining the legality of each decision made. For example, evading income taxes by not reporting extra income is in
837
violation of the IRS code. Hence, individuals who state they would not report the extra income (scale item N 4.0) are defined as socially unethical. The sample consisted of white (90%), married (64%), middle-aged (21–39; 81%), males (75%), with 8–15 years (57%) retailing experience. 4. Analysis 4.1. Model fit and convergent and discriminant validity Following Anderson and Gerbing (1988), the measurement quality of the indicators were evaluated and refined before testing the structural component of the model. This two-step procedure is widely used to identify a final set of items with acceptable discriminant and convergent validity, internal consistency, reliability and parsimony. Hu and Bentler (1999) suggest that both incremental and absolute fit measures be used to assess the fit of a model. CFI and IFI were the incremental fit measures. All measures have been widely reported. Generally, CFI and IFI values greater than 0.9 (Hu and Bentler, 1999) are indicators of good fit of the data to the model tested. The standardized loadings, t-values and composite reliability values in Table 1 reveal the following fit (χ2 = 138.42; d.f. = 48; RMSEA = 0.11; GFI = 0.89; NNFI = 0.63; CFI = 0.73; IFI = 0.75). To improve model fit, Anderson and Gerbing (1988) and MacCallum's (1986) method of analyzing standardized residual values and modification indices while keeping in mind the conceptual contribution of each item yields a high modification index in the Theta–Epsilon matrix. The recommended changes are to develop a path (linkage) between formalization (observed) and significant others (latent). Although not hypothesized the formalization of corporate activities provides the corporation with an identity in the mind of the employee. The data also shows the removal of the path between familiarity and opportunity. Upon closer inspection this scale does not clearly represent the item theoretically hypothesized. The data also shows the path removal between moral philosophy, age, and religiosity and individual factors. One reason for the path
Table 1 Factor loadings and t-values of the items in the measurement model Parameter
Construct Factor t-values of Average variance reliability loadings factor loadings extracted
Opportunity 0.75 Formalization Acceptance of authority Organizational controls Significant others 0.98 Friends Business associates Superiors Ethicalness 0.76 Individual Organizational Societal
0.51 0.80 0.52
7.83 5.94
0.53
6.03
0.99 0.82 0.69
16.99 12.86 10.30
0.37 0.31 0.99
4.00 3.64 5.77
0.95
0.58
838
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
change is in the way the overall construct of individual factors was specified. This path change does not point to the conclusion that philosophy, age and religiosity are not pertinent, but that they do not have sufficient explanatory power for statistical applications. This led to a significant improvement in the model fit and one that is quite acceptable (χ2 = 59.09; d.f. = 24; RMSEA = 0.08; GFI = 0.93; NNFI = 0.91; CFI = 0.92; IFI = 0.92). The next step is assessing the convergent and discriminant validity of the model. Table 1 shows all factor loadings as significantly different from zero (consistently large t-values). Fornell and Larcker (1981) assess convergent validity by the average variance extracted per factor. Accordingly, all extracted average variances exceeded the 0.50 level. To assess discriminant validity, the procedure by Bagozzi and Yi (1991) using two models to be estimated for each pair of constructs — one model in which the correlation between the two variables is allowed to be freely estimated and one model in which the correlation is constrained to be equal to zero. If the χ2 value of the unconstrained model is more than 3.84 but less than the constrained model (where 3.84 is the critical value for one degree of freedom) there is evidence of discriminant validity. In all cases the χ2 for the unconstrained model was significantly less than the constrained model suggesting that the factors exhibit discriminant validity. Finally, Bagozzi and Yi's (1991)
procedure testing the reliability of the scales via the calculation of composite reliability scores ranged from 0.75 to 0.98 and are above the cutoff of 0.6. The results suggest sufficient validity and reliability of the measures as to allow the testing of the hypotheses. 4.2. Structural model and hypotheses testing The structural component of the revised model in Fig. 2 was tested using LISREL 8.30 (Jöreskog and Sörbom, 1993). The fit of the data to the revised model (see Fig. 2) is quite good (χ2 = 61.30; d.f. = 24; RMSEA = 0.08; GFI = 0.93; NNFI = 0.90; CFI = 0.92; IFI = 0.92), but the path from Opportunity to Ethicalness (t-value = 1.19) is non-significant. Hence H1 is not supported. However, Hypothesis 1A is significant and the path between Formalization and Opportunity positive (t = 7.83). Hypothesis 1B, the path between Acceptance of Authority to Opportunity is also positive and significant (t = 5.94) as well as the path between Organization Controls to Opportunity (H1C; t = 6.03). Hypothesis 2 is significant in that the path coefficient from Significant Others to Ethicalness is −0.28 (t-value = −2.08) as well as the sub-hypotheses (H2A: t = 16.99; H2B: t-value = 12.86; H2C: t-value = 10.30). Because the LISREL analysis precludes the path between individual factors and ethicalness, Hypothesis 3 is rejected.
Fig. 2. Revised model of ethicalness.
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
5. Conclusions, limitations, and future research From the revised model significant others such as business associates and supervisors appear to have more importance than opportunity. A surprising result is that friends contributed most instead of a person's superiors or business associates, especially given that two of the three vignettes were business related. One possible explanation for this finding may be that respondents make a distinction between business associates and friends within the business or company. Role-set theory and business associates of superiors being more significant is somewhat overshadowed by an employee's friends. Partial support for this conclusion comes from the criminal literature that found a relationship between peers, perpetrators, and behavior (Agnew, 1993; Warr, 1993; Graham and Bowling, 1996; Hawkins, 1996). Another explanation for the result is because these decisions have an ethical component, employees may trust individuals with whom they have a prolonged relationship over casual associates. In addition, conformity theory within social psychology confirms that such a finding likely exists. The results of age, length of business tenure, and contribution between friend and business associates may also be useful in result explanation. Formalization is most important in explaining ethicalness within the opportunity construct. This implies that standardization of ethical codes and subsequent behaviors be stressed within organizations more than looking at an employee's acceptance of authority figures. For example, if a retailer has limited resources, more money should be spent in developing comprehensive codes with penalties and rewards for ethical behavior than if employees readily accept authority figures or legitimate power. In hiring, retailers might look for employees who understand the why behind rules and procedures instead of militarytype personalities that obey without understanding. Another significant finding is the exclusion of individual factors. Even when the individual factors construct is relaxed and moral philosophy, religiosity, and age are directly run they yield non-significant path associations. One explanation for moral philosophy not being included is the manner it was measured. The dichotomous evaluation of deontology/teleology across three situations was dissimilar to other scale-types. Additionally, the moral philosophy scale did not show the actual continuum. Rather it represents the moral justification state that an employee uses in determining their decision. Although many studies show values and moral philosophies as important, their overall contribution to ethicalness is not. From a retailing perspective, companies should place more resources into training their trainers and less on honesty or moral philosophy tests. Because when new hires join the firm, they absorb the values of the company and the people around them, thus making such pen and paper tests useless for prediction. Although important in decision making, religiosity is not a significant contributor to any of the three ethics domains even when checked independently of other individual factors. The religiosity research shows conflicting results and the research substantiates it. An alternative explanation is even though a person may have a strong moral value system that is being
839
supported by various religious activities it just doesn't become significant when ethical/unethical business decisions are made. Finally, although age is tied to antisocial behavior age does not explain a significant amount of variance. The sample was fairly homogenous; hence the findings may be biased. However, O'Fallon and Butterfield's (2005) review on age and ethics is also inconclusive. From the findings, retailers need to reassess the adage that as one gets older one becomes more ethical. This is not necessarily the case. The research limitations should be seen as markers for future investigations. For example, in order to analyze more constructs a larger sample is needed and, depending upon the model tested, either a more homogenous/heterogeneous sample chosen. Moral philosophy needs to be better operationalized as well as somehow decreasing the questionnaire length yet increasing the amount of information obtained. Increasing the number of vignettes from three to many, as well as including more theoretical constructs for testing needs to be explored. For factors that influence unethical behavior the literature suggests that significant others (Robideaux, 1989), philosophy type (Fraedrich, 1993), and perhaps corporate culture (Herndon, 1991; 1992) have an impact. While some retailers have codes of ethics, many are quite general in language and scope. Research on whether the antecedents of customers' perceptions of retail ethics include such factors as product price or retailer advertising is needed. Work on helping retailers prevent unethical behavior; creating and communicating ethical guidelines; selecting hiring and training ethical salespeople; and measuring the impact of both ethical and unethical behavior on performance is also needed. Scale development from all three perspectives is a research opportunity, as is theory development for explaining retailer ethics. Future research within the empirical testing of ethical decision making models needs to investigate the anomalies discovered as well as testing the ethicalness construct. This is important relative to the issue of situational ethics (Kellaris, Dahlstrom and Boyle, 1996; Banerjee et al., 1998). Another issue is the differences between how individuals make decisions concerning ethical issues and dilemmas. An ethical issue is one that has some synergy between the three ethics domains as it relates to right/wrong choices. An ethical dilemma is where the matrix of ethical/unethical is not synergistic between the three domains; hence, the individual believes that there are only wrong/unethical solutions to the situation. Kohlberg's vignettes are excellent examples of dilemmas where decisions involve breaking laws, death, and stealing to survive. Far too often in business, the situations are not so severe. Retailers need to become more proactive in their codes and their effectiveness relative to sales training, amount and type of supervision, and desired type of salespeople. The multinational aspect of the models being used and tested is very important. Much of the theoretical work is done with a western perspective. Many of the scales used were developed in western cultures. The question is whether researchers are missing unique factors in other cultures thus restricting generalizability. For example, would the values construct change if the retailers tested were from predominantly Muslim, Buddhist,
840
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
Hindu, or Jewish samples? Are academicians actually capturing the pertinent philosophies associated with all business people? And finally, would the ethicalness construct be the same in other cultures? A major contribution to the field is that business ethics is not just about formal rules and procedures, or how long or familiar someone is with the organization. The acceptance of authority, religion, and age are not very important relative to the ethicalness of individuals within business either. The findings suggest that friends, business associates, superiors, and business associates that are friends, contribute most to ethicalness. The findings suggest that the adage, lie down with dogs get up with fleas, is one step closer to being empirically proven within business ethics. This research does not purport to be error free. Many choices between parsimony and data richness were made. Not all constructs identified in all of the ethics models were incorporated. Yet the contribution to the literature is in testing three major and ten associated constructs together and demonstrating their relative explanatory strengths to a construct that explains the multi-faceted reality of ethics. Our research is a continuation of the many researchers involved in teasing out, bit by bit, the secrets of how ethically charged decisions are made, what causes business people to make unethical choices, and how we can help the individual, organization, and society make better decisions. References Agnew R. Why do they do it? An examination of the intervening mechanisms between social control variables and delinquency. J Res Crime Delinq 1993;30(3):245–66. Akaah IP. An organizational culture and ethical research behavior. J Acad Mark Sci 1993;21(1):59–63. Anderson JC, Gerbing DW. Structural equation modeling practice: a review and recommended two-step approach. Psychol Bull 1988;103:411–23. Bagozzi RP, Yi Y. Assessing construct validity in organizational research. Adm Sci Q 1991;36:459–89. Banarjee D, Cronan TP, Jones TW. A modeling IT ethics—a study in situational ethics. MIS Q 1998;22(1):31–60. Barnett T, Bass K, Brown G. Religiosity, ethical ideology, and intentions to report a peers' wrong doing. Journal of Business Ethics 1996;15(11): 1161–75 [November]. Bartels R. A model for ethics in marketing. J Mark 1967;31:20–6 [January]. Boyce WD, Jensen LC. Moral Reasoning: a Psychological–Philosophical Integration. Lincoln, Nebraska: University of Nebraska Press; 1978. Brass DJ, Butterfield KD, Skaggs BC. A relationship and unethical behavior—a social network perspective. Acad Manage Rev 1998;23(1):14–31. Carlson DS, Kacmar MK. A perception of ethics across situations: a view through three different lenses. J Bus Ethics 1997;16(2):147–60 [February]. Cunningham, M. Walking the thin line: a role conflict model of ethical decisionmaking behavior in the marketing research process. 1991; Ph.D. dissertation, Texas A & M University. DeConinck JB, Lewis WF. The influence of deontological and teleological considerations and ethical climate on sales. J Bus Ethics 1997;16(5): 497–507 [April]. Ferrell OC, Gresham L. A contingency framework for understanding ethical decision-making in marketing. J Mark 1985:87–96 [Summer]. Ferrell OC, Skinner SJ. Ethical behavior and bureaucratic structure in marketing research organizations. J Mark Res 1988;25:103–9 [February]. Ferrell OC, Gresham L, Fraedrich JP. A synthesis of ethical decision models for marketing. J Macromark 1989;9(2):55–64 [Fall].
Ferrell OC, Fraedrich JP. Moral philosophies and marketing managers:an empirical investigation. J Mark Manage 1991;1(1):18–27 [Fall]. Fornell C, Larcker DF. Evaluating structural equation models with unobservable variables and measurement error. J Mark Res 1981;18:39–50. Fort TL. Religion and business ethics: the lessons from political morality. J Bus Ethics 1977;16: 263–73. Fraedrich, J.P. Philosophy type interaction in the ethical decision making process of retailers. 1988; Ph.D. Dissertation, Texas A & M University. Fraedrich JP. The ethical behavior of retail managers. J Bus Ethics 1993(12): 1–12. Fraedrich JP, Ferrell OC. The impact of perceived risk and moral philosophy type on ethical decision making in business organizations. J Bus Res 1992;24(4):283–96 [June]. Fraedrich JP, Ferrell OC. Cognitive consistency of marketing managers in ethical situations. J Acad Mark Sci 1992a;20(3):245–52. Fritzsche DJ, Becker H. Ethical behavior of marketing managers. J Bus Ethics 1983;2:291–9. Giacalone RA, Jurkiewicz CL. Right from wrong: the influence of spirituality on perceptions of unethical business activities. J Bus Ethics 2003;46(1): 85–97. Graham J, Bowling B. Young people and crime, home office research study. #145. UK: Research and Statistics Department; 1996. UK Home Office. Grimes AJ. Authority, power, influence and social control: a theoretical synthesis. Acad Manage Rev 1978;3:724–35 [October]. Gundlach GT, Murphy PE. Ethical and legal foundations of relational marketing exchanges. J Mark 1993;57:35–46 [October]. Hawkins JD. Delinquency and crime: current theories. Cambridge Criminology Series. Cambridge University Press; 1996. Heider F. The psychology of interpersonal relations. New York: Wiley; 1958. Herndon NC. Commitment, Satisfaction and Turnover. The Impact of Individual Moral Value Structures when exposed to a Conflicting or Matching Organizational Sales Culture. Doctoral Dissertation, Texas A & M University; 1991. Herndon NC. Perceived corporate ethicalness and salesforce turnover: a new rationale for more ethical corporate behavior. Dev Mark Sci 1992:330–5 XV. Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Model 1999;6(1): 1–55. Hunt SD, Vitell S. A general theory of marketing ethics. J Macromark 1986;6: 5–16 [Spring]. Hunt SD, Chonko LB, Wilcox JB. Ethical problems of marketing researchers. J Mark Res 1984;21:309–24 [August]. John G. An empirical investigation of some antecedents of opportunism in a marketing channel. J Mark Res 1984;21:278–89 [August]. Jones TM, Ryan LV. The link between ethical judgment and action in organizations—a moral approbation approach. Organization Science 1997;8(6):663–80. Joreskog KG, Sorbom D. LISREL 8.30: A guide to the program and applications. Chicago: Scientific Software International; 1993. Kellaris JJ, Dahlstrom RF, Boyle BA. Contextual bias in ethical judgment of marketing practices. Psychol Mark 1996;13(7):677–94 [October]. Kennedy EJ, Lawton L. The effects of social and moral integration on ethical standards: a comparison of American and Ukrainian Business students. J Bus Ethics 1996;15(8):901–11. Kerlinger FN. Foundations of behavioral research. New York: Holt, Rinehart and Winston; 1986. Kirkpatrick DD. Investors Circle Columbia/HCA Healthcare Corp. Wall Street J 1997:B12 [October 22]. Krugman DM, Ferrell OC. The organizational ethics of advertising: corporate and agency views. J Advert 1981;10(1):21–48. Loe TW, Ferrell L, Mansfield P. A review of empirical studies assessing ethical decision making in business. J Bus Ethics 2000;25(3):185–204. MacCallum RC. Specification searches in covariance structure modeling. Psychol Bull 1986;100:107–20. Malhotra NK, Miller GL. An integrated model for ethical decisions in marketing research. J Bus Ethics 1998;17(3):263–80. McDaniel SW, Burnett JJ. Consumer religiosity and retail store evaluative criteria. J Acad Mark Sci 1990;18(2):101–12 [Spring].
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841 McDaniel SW, Burnett JJ. Targeting the evangelical market segment. J Advert Res 1991;31:26–33 [August–September]. Merton RK. The role-set. Br J Sociol 1957(8):106–20. O'Fallon MJ, Butterfield KD. A review of the ethical empirical decision-making literature: 1996–2003. J Bus Ethics 2005;50:375–413. Ouchi WG. The relationship between organizational structure and organization control. Adm Sci Q 1977(22):95–113. Powpaka S. Factors affecting managers' decision to bribe: an empirical investigation. J Bus Ethics 2002;40(3):227–46. Razzaque MA, Hwee TP. Ethics and purchasing dilemma: a Singaporean view. J Bus Ethics 2002;35(4):307–26. Reidenbach RE, Robin DP. An application and extension of a multidimensional ethics. J Acad Mark Sci 1991;9(2):83–93 [Spring]. Robideaux DR. Significant others and ethical behavior of retail managers. Doctoral Dissertation 1989; Louisiana Tech University. Singer MS, Singer AE. Observer judgments about moral agents' ethical decisions: the role of scope of justice and moral intensity. J Bus Ethics 1997;16(5):473–84 [April]. Singhapakdi A, Vitell SJ. Marketing ethics: factors influencing perceptions of ethical problems and alternatives. J Macromark 1990:4–18 [Spring]. Singhapakdi A, Marta JK, Rallapalli KC, Rao CP. Toward an understanding of religiousness and marketing ethics: an empirical study. J Bus Ethics 2000;27(4):305–19. Sparks JR, Hunt SD. Marketing researcher ethical sensitivity: conceptualization, measurement, and exploratory investigation. J Mark 1998;62(2):92–109 [April]. Sutherland E, Crissey DR. Principles of criminology. 8th edition. Chicago: J.B. Lippincott; 1970.
841
Tepstra DE, Rozell EJ, Robinson RK. The influence of personality and demographic variables on ethical decisions related to insider trading. J Psychol 1993(4):375–89. Thorne-LeClair D, Ferrell OC, Fraedrich J. Integrity management. Tampa: University of Tampa Press; 1998. Trevino LK. Ethical decision making in organizations: a person–situation interactionist model. Acad Manage Rev 1986(3):601–17. Tse ACB, Au AKM. Are New Zealand business students more unethical than non-business students? J Bus Ethics 1997;16(4):445–50. Upchurch R. A conceptual foundation for ethical decision-making: a stakeholder perspective in the lodging industry. J Bus Ethics 1998;17(12): 1349–61. Vitell SJ, Hunt SD. The general theory of marketing ethics: a partial test of the model. Res Mark 1990;10:237–65. Warr M. Parents, peers, and delinquency. Soc Forces 1993;1(72):247–64 [September]. Wilkes RE, Burnett JJ, Howell RD. On the meaning and measurement of religiosity in consumer research. J Acad Mark Sci 1986;14(Spring):47–56. Withey S. The US and USSR: a report of the public’s perspective on the United States-Russian relations in late 1961. In: Borrow D, editor. Components of Defense Policy. Chicago: Rand McNally; 1965. pp. 164–174. Zey-Ferrell MK, Weaver M, Ferrell OC. Predicting unethical behavior among marketing practitioners. Hum Relat 1979;7(32):557–69. Zey-Ferrell MK, Ferrell OC. Role-set configuration and opportunity as predictors of unethical behavior in organizations. Hum Relat 1982;7(35): 587–604.
Journal of Business Research 61 (2008) 834 – 841
Retailers' major ethical decision making constructs John Fraedrich a,⁎, Rajesh Iyer b,1 a
Southern Illinois University at Carbondale, United States b Bradley University, United States
Received 1 June 2005; received in revised form 1 March 2006; accepted 1 August 2007
Abstract The major constructs of the Ferrell and Gresham [Ferrell OC, Gresham L. A Contingency Framework for Understanding Ethical DecisionMaking in Marketing. J Mark 1985; (Summer): 87–96.], Hunt and Vitell [Hunt SD, Vitell S. A General Theory of Marketing Ethics. J Macromark 1986; 6 (Spring): 5–16.], and Ferrell et al. [Ferrell OC, Gresham L, Fraedrich JP. A Synthesis of Ethical Decision Models for Marketing. J Macromark 1989; 9(2) (Fall): 55–64.] models were tested using LISREL. Our findings suggest that Friends, Superiors, Business Associates, and Formalization of the Organization are more significant than Moral Philosophy. A new construct Ethicalness, comprised of individual, organizational, and societal value structures yielded a tight multi-dimensional fit that supports Kerlinger's domain theory and is dissimilar to Reidenbach and Robin's (1991) multi-dimensional scale. This research is one of the first to test as many ethical decision making constructs while explaining the multi-dimensional conundrum of ethical/unethical and legal/illegal practices within retailing. © 2007 Elsevier Inc. All rights reserved. Keywords: Retailers; Ethics; Ethicalness; Moral philosophy; Significant others
1. Introduction Since the late 1980′s ethics has become an important modeling aspect. AACSB has reiterated the importance of ethics by their accreditation formula as well as a plethora of research models, books, courses, and ethics centers. The rationale is simple: “Reinforcing ethical behavior is important for improving performance and achieving success in the marketplace” (Gundlach and Murphy, 1993, p. 44). But to improve performance, academia and practitioners must determine the empirical strength of the constructs that purport to be important within ethical decision making. Marketing is the academic leader, being first to develop ethical decision making models (Ferrell and Gresham, 1985; Hunt and Vitell, 1986; Ferrell et al., 1989; Malhotra and Miller, ⁎ Corresponding author. Department of Marketing, College of Business and Administration, Southern Illinois University Carbondale, Rehn Hall 229, 1025 Lincoln Drive, Carbondale, IL 62901, United States. Tel.: +1 618 453 4341. E-mail addresses: [email protected] (J. Fraedrich), [email protected] (R. Iyer). 1 Tel.: +1 309 6772991. 0148-2963/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2007.08.006
1998). Most research within the ethics literature in other disciplines is based on these models (Trevino, 1986; Jones and Ryan, 1997; Banerjeeet al., 1998; Brass, Butterfield, and Skaggs, 1998; Upchurch, 1998). The past decade points to the constructs of opportunity in the work environment, moral philosophy/values, peers and superiors, as well as individual factors facilitating ethical/unethical determination of business decisions. Yet, no one has attempted to take these key constructs and determine their cumulative effect on a retailing sample. This article is a first attempt. A second issue within the ethics stream is quantifying the concept, “ethical/unethical.” A large percentage of research takes ethical scenarios or questionable practices and asks whether or not the action is relatively ethical, right, fair, or moral. The premise is ethics is relative to an individual or group perspective. What one defines as right/ethical therefore becomes reality. The other argument is side-stepped concerning absolute truth from a non-relativist approach. The result is a plethora of articles explaining what is ethical/unethical from the marketer's perspective. Compounding the problem is the relativity of time on perception. The solution is a measure that is multi-dimensional with an ontological perspective that does
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
not originate with the marketer. In other words, once the decision is made how does the individual, business unit and society perceive the action? Pursuant to this question, three domains are identified, tested, and overall paths and associative strengths to moral philosophy, peers and superiors, and other individual factors determined. The sections that follow identify the significant constructs and hypotheses, define ethicalness, and analyze a model using LISREL. 2. Ethics models, constructs and ethicalness Several ethics models are generally accepted (Ferrell and Gresham, 1985; Hunt and Vitell, 1986; Trevino, 1986; Ferrel and Skinner, 1988) with Vitell and Hunt (1990) and Carlson and Kacmar's (1997) being the only ones to partially test a model. In Carlson and Kacmar, moral philosophy was the best predictor on 123 students; but, construct consistency was not tested. Theories associated with ethics models are differential association, cognitive consistency, Fishbein's AIB, as well as Piaget's learning theories with key constructs being opportunity, significant others, and individual factors (See Fig. 1).
835
Opportunity (Bartels, 1967) results from “a favorable set of conditions to limit barriers or provide rewards” (Ferrell and Gresham 1985, p. 92) and is theoretically and somewhat empirically linked to ethical decision making (O'Fallon and Butterfield, 2005; Powpaka, 2002). In Powpaka's structural analysis of bribery and Thai managers, opportunity was defined relative to corporate and cultural circumstances that allow bribes to be given. Barriers to commit unethical actions (negative) or rewards (intrinsic and/or extrinsic) for not committing such actions make up the opportunity construct. The increase or decrease of opportunity within business is affected by the formalization of the organization (centralized/decentralized), the retailer's acceptance of authority within the organizational structure, organizational controls (rules), and situation familiarity. To reduce the opportunity (formalization, controls, authority acceptance, and situation familiarity) to commit unethical or illegal actions, the health care industry hired in-house police to help minimize potential fraud in billing because of Columbia/HCA Health Care Corporation's compliance problems with the Federal Government (Kirkpatrick, 1997). Associating with opportunity are formalization of the organization, acceptance of authority, organizational controls, and
Fig. 1. Hypothesized model of ethicalness.
836
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
familiarity. Formalization of the organization is the standardizing or codifying of operations and procedures within a firm. Several studies link the formalization of bureaucracy to improved ethical conduct (Fritzsche and Becker, 1983; Hunt, Chonko, and Wilcox, 1984; Ferrell and Skinner, 1988; Singhapakdi and Vitell, 1990; Akaah, 1993). Acceptance of authority is in Ferrell and Skinner's (1988) centralization construct and associated with compliance of subordinates to a superior/company/society's legitimate right to exercise control (Grimes, 1978). Compliance comes into the ethics equation regardless of whether or not the employee agrees with the superior. The acceptance of such authority should theoretically increase or play a part in the ethical decision making process. Organizational controls include target setting, measuring or monitoring, and feedback (Ouchi, 1977). Ethics controls are operationalized by setting up rules, regulations, and internal procedures and linked to ethical behavior (Ferrell and Skinner, 1988; Thorne-LeClair et al., 1998). Familiarity relates to how routinized a process/action becomes. Familiar situations should, through cognitive consistency theory (Heider, 1958), be regarded as ethical to retailers. Cognitive consistency theory states people attempt to reach a state of decision equilibrium. Familiarity may be akin to ethical sensitivity in that the sensitivity construct measures the ethical/moral value awareness of the situation that changes the weights in the decision process (Sparks and Hunt, 1998). While many support the theory, Fraedrich and Ferrell (1992) found an anomaly that some teleological types say one thing but do something else that subjects define as unethical. Within the model all four factors relate to opportunity positively. H1. Opportunity relate positively to ethicalness. H1a. Formalization of the organization relate positively to opportunity. H1b. Acceptance of Authority relate positively to opportunity. H1c. Organizational controls relate positively to opportunity. H1d. Familiarity of the situation relate positively to opportunity. Significant other is based, in part, on differential association (Sutherland and Crissey, 1970) and role-set theories (Merton, 1957). Differential association states that people are influenced more by those closest to them whether spatially or perceptually. Role-set theory refers to the complement of role relationships which focal persons have by virtue of their social status in an organization (Ferrell and Gresham, 1985, p.91). For example, differential association explains why a person's friends or family have an influence on business-related decisions. Role-set theory explains why, through authority figures such as supervisors or the firm itself, such entities influence the ethical decision process. Zey-Ferrell et al. (1979) and others (Zey-Ferrell and Ferrell, 1982; Hunt et al., 1984; Cunningham, 1991) show a link between business associates and one's superior and their effect on the ethical decision process. In essence if the superior or associates are prone to unethical behavior, one's own decision process is affected in the negative. The
adage, lie down with dogs get up with fleas, corresponds to what may be happening in the business arena. In addition, it appears that friends outside the firm as well as corporate culture may also affect the ethical decision process. H2. Significant others relate positively to ethicalness. H2a. Friends relate positively to significant others. H2b. Business Associates relates positively significant others. H2c. Superiors relate positively significant others. Individual factors represent an individual's philosophical process. The heart of ethical decision making revolves around moral philosophies; specifically deontology and teleology. Current models suggest these two philosophical approaches are used in resolving ethical dilemmas (Ferrell and Gresham, 1985; Hunt and Vitell, 1986; Trevino, 1986; Ferrell et al., 1989) and an empirical link between philosophies and ethical behavior has also been established (Fraedrich, 1988; Ferrell and Fraedrich, 1991; Fraedrich and Ferrell, 1992; 1992a; DeConinck and Lewis, 1997). Deontology relates to philosophies centered on individual rights such as Ross' prima facie duties. Teleology stresses end-states, consequences, and group action above the individual's rights. Religiosity is an individual factor not in research reports relative to ethical decision. Religiosity relates to a person's relationship with a supreme being and is recognized as a force in consumer behavior (McDaniel and Burnett, 1990; 1991; Terpstra et al., 1993; Barnett et al., 1996; Wilkes et al., 1986) and may be a legitimate construct in business ethics (Fort, 1977). Recent research suggests that an increased level of religiosity helps people recognize ethical problems, can positively affect their judgment, and religious people are less willing to engage in unethical behavior (Kennedy and Lawton, 1996; Tse and Au, 1997; Singhapakdi et al., 2000; Razzaque and Hwee, 2002). However, many measure religiosity in differently and Giacalone and Jurkiewicz (2003) report that spirituality does not correlate well with religiosity. Demographic factors such as marital status and age are associated with ethical decision making (Loe, Ferrell and Mansfield, 2000; O'Fallon and Butterfield, 2005). For example, older individuals tend to be perceived as more ethical (Tepstra et al., 1993). H3. Individual factors relate positively to ethicalness. H3a. Moral philosophy type relates positively to individual factors. H3b. Age relates positively to individual factors. H3c. Religiosity relate positively to individual factors. When doing ethics research a large problem is the definition as to what is ethical. The relativist argument from Socratic times states that whatever one defines as right or ethical therefore becomes such. Borrowing from sociology, Kerlinger (1986) answers the relativist's argument with domain theory. Domain theory argues that one needs to view the decision relative to the
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
domain, area, or level of experience that a particular society has attained. In ethics the different domains are the ethical perceptions of the individual, organization, and society. Under cognitive consistency all should be correlated. Krugman and Ferrell (1981) allude to the dimensionality of the ethics construct with their differences between what was individually perceived as ethical and what friends and business associates believed. Singer and Singer (1997) also advocate a multi-dimensional construct as the incorporation of social and ethical concerns within a utilitarian decision structure. No research to date has simultaneously measured these three domains (individual, organization, and society). 3. Method 3.1. Sample, measures and procedure Retail managers/salespeople were chosen because of their boundary-spanning role in their organization. The questionnaire was pre-tested on business personnel and word changes in some of the statements were made to reduce comprehension confusion in our retail sample. An expert panel reviewed the changes and deemed that they did not significantly alter any item meanings. The revised questionnaire was sent to a major retail conglomerate (700 managers) resulting in 186 usable returns (26.6% rate). Because of the sensitivity of the topic and restrictions from the sample corporation, it was impossible to obtain a comparison sample of non-respondents. Three validated vignettes were used (betraying a trust, bribery, and income tax evasion) (Fritzsche and Becker, 1983; Fraedrich, 1988; Fraedrich and Ferrell, 1992). Scales adapted by John (1984) measured the opportunity via formalization of the organization and organizational controls. All items were 7 point Likert-type with several reversed scored. Acceptance of authority was measured by Withey's (1965) scale measuring the extent an individual tends to obey orders and respect authority. Separate statements associated with each vignette asked if the manager had ever come across this type of problem (situation familiarity). Moral philosophy was measured using modified statements from Boyce and Jensen's (1978) Moral Content Test (MCT) across three vignettes. Retail managers were specifically asked which statement or philosophy best explained the rationale for their decision. The forced choice helps respondents distinguish between salient and non-salient factors. Due to respondent time constraints, a multi-item philosophical scale for each vignette was inappropriate, as it would have jeopardized the study. Religiosity was measured both cognitively and behaviorally using McDaniel and Burnett's (1991) three-item scale. Ethicalness was measured by asking respondents about the perceived ethicalness of their decision (individual). Next, a modified version of Ferrell and Skinner's (1988) ethical behavior scale was used for the organizational perspective (Fraedrich, 1993). Societal ethics was measured by taking each vignette and determining the legality of each decision made. For example, evading income taxes by not reporting extra income is in
837
violation of the IRS code. Hence, individuals who state they would not report the extra income (scale item N 4.0) are defined as socially unethical. The sample consisted of white (90%), married (64%), middle-aged (21–39; 81%), males (75%), with 8–15 years (57%) retailing experience. 4. Analysis 4.1. Model fit and convergent and discriminant validity Following Anderson and Gerbing (1988), the measurement quality of the indicators were evaluated and refined before testing the structural component of the model. This two-step procedure is widely used to identify a final set of items with acceptable discriminant and convergent validity, internal consistency, reliability and parsimony. Hu and Bentler (1999) suggest that both incremental and absolute fit measures be used to assess the fit of a model. CFI and IFI were the incremental fit measures. All measures have been widely reported. Generally, CFI and IFI values greater than 0.9 (Hu and Bentler, 1999) are indicators of good fit of the data to the model tested. The standardized loadings, t-values and composite reliability values in Table 1 reveal the following fit (χ2 = 138.42; d.f. = 48; RMSEA = 0.11; GFI = 0.89; NNFI = 0.63; CFI = 0.73; IFI = 0.75). To improve model fit, Anderson and Gerbing (1988) and MacCallum's (1986) method of analyzing standardized residual values and modification indices while keeping in mind the conceptual contribution of each item yields a high modification index in the Theta–Epsilon matrix. The recommended changes are to develop a path (linkage) between formalization (observed) and significant others (latent). Although not hypothesized the formalization of corporate activities provides the corporation with an identity in the mind of the employee. The data also shows the removal of the path between familiarity and opportunity. Upon closer inspection this scale does not clearly represent the item theoretically hypothesized. The data also shows the path removal between moral philosophy, age, and religiosity and individual factors. One reason for the path
Table 1 Factor loadings and t-values of the items in the measurement model Parameter
Construct Factor t-values of Average variance reliability loadings factor loadings extracted
Opportunity 0.75 Formalization Acceptance of authority Organizational controls Significant others 0.98 Friends Business associates Superiors Ethicalness 0.76 Individual Organizational Societal
0.51 0.80 0.52
7.83 5.94
0.53
6.03
0.99 0.82 0.69
16.99 12.86 10.30
0.37 0.31 0.99
4.00 3.64 5.77
0.95
0.58
838
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
change is in the way the overall construct of individual factors was specified. This path change does not point to the conclusion that philosophy, age and religiosity are not pertinent, but that they do not have sufficient explanatory power for statistical applications. This led to a significant improvement in the model fit and one that is quite acceptable (χ2 = 59.09; d.f. = 24; RMSEA = 0.08; GFI = 0.93; NNFI = 0.91; CFI = 0.92; IFI = 0.92). The next step is assessing the convergent and discriminant validity of the model. Table 1 shows all factor loadings as significantly different from zero (consistently large t-values). Fornell and Larcker (1981) assess convergent validity by the average variance extracted per factor. Accordingly, all extracted average variances exceeded the 0.50 level. To assess discriminant validity, the procedure by Bagozzi and Yi (1991) using two models to be estimated for each pair of constructs — one model in which the correlation between the two variables is allowed to be freely estimated and one model in which the correlation is constrained to be equal to zero. If the χ2 value of the unconstrained model is more than 3.84 but less than the constrained model (where 3.84 is the critical value for one degree of freedom) there is evidence of discriminant validity. In all cases the χ2 for the unconstrained model was significantly less than the constrained model suggesting that the factors exhibit discriminant validity. Finally, Bagozzi and Yi's (1991)
procedure testing the reliability of the scales via the calculation of composite reliability scores ranged from 0.75 to 0.98 and are above the cutoff of 0.6. The results suggest sufficient validity and reliability of the measures as to allow the testing of the hypotheses. 4.2. Structural model and hypotheses testing The structural component of the revised model in Fig. 2 was tested using LISREL 8.30 (Jöreskog and Sörbom, 1993). The fit of the data to the revised model (see Fig. 2) is quite good (χ2 = 61.30; d.f. = 24; RMSEA = 0.08; GFI = 0.93; NNFI = 0.90; CFI = 0.92; IFI = 0.92), but the path from Opportunity to Ethicalness (t-value = 1.19) is non-significant. Hence H1 is not supported. However, Hypothesis 1A is significant and the path between Formalization and Opportunity positive (t = 7.83). Hypothesis 1B, the path between Acceptance of Authority to Opportunity is also positive and significant (t = 5.94) as well as the path between Organization Controls to Opportunity (H1C; t = 6.03). Hypothesis 2 is significant in that the path coefficient from Significant Others to Ethicalness is −0.28 (t-value = −2.08) as well as the sub-hypotheses (H2A: t = 16.99; H2B: t-value = 12.86; H2C: t-value = 10.30). Because the LISREL analysis precludes the path between individual factors and ethicalness, Hypothesis 3 is rejected.
Fig. 2. Revised model of ethicalness.
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
5. Conclusions, limitations, and future research From the revised model significant others such as business associates and supervisors appear to have more importance than opportunity. A surprising result is that friends contributed most instead of a person's superiors or business associates, especially given that two of the three vignettes were business related. One possible explanation for this finding may be that respondents make a distinction between business associates and friends within the business or company. Role-set theory and business associates of superiors being more significant is somewhat overshadowed by an employee's friends. Partial support for this conclusion comes from the criminal literature that found a relationship between peers, perpetrators, and behavior (Agnew, 1993; Warr, 1993; Graham and Bowling, 1996; Hawkins, 1996). Another explanation for the result is because these decisions have an ethical component, employees may trust individuals with whom they have a prolonged relationship over casual associates. In addition, conformity theory within social psychology confirms that such a finding likely exists. The results of age, length of business tenure, and contribution between friend and business associates may also be useful in result explanation. Formalization is most important in explaining ethicalness within the opportunity construct. This implies that standardization of ethical codes and subsequent behaviors be stressed within organizations more than looking at an employee's acceptance of authority figures. For example, if a retailer has limited resources, more money should be spent in developing comprehensive codes with penalties and rewards for ethical behavior than if employees readily accept authority figures or legitimate power. In hiring, retailers might look for employees who understand the why behind rules and procedures instead of militarytype personalities that obey without understanding. Another significant finding is the exclusion of individual factors. Even when the individual factors construct is relaxed and moral philosophy, religiosity, and age are directly run they yield non-significant path associations. One explanation for moral philosophy not being included is the manner it was measured. The dichotomous evaluation of deontology/teleology across three situations was dissimilar to other scale-types. Additionally, the moral philosophy scale did not show the actual continuum. Rather it represents the moral justification state that an employee uses in determining their decision. Although many studies show values and moral philosophies as important, their overall contribution to ethicalness is not. From a retailing perspective, companies should place more resources into training their trainers and less on honesty or moral philosophy tests. Because when new hires join the firm, they absorb the values of the company and the people around them, thus making such pen and paper tests useless for prediction. Although important in decision making, religiosity is not a significant contributor to any of the three ethics domains even when checked independently of other individual factors. The religiosity research shows conflicting results and the research substantiates it. An alternative explanation is even though a person may have a strong moral value system that is being
839
supported by various religious activities it just doesn't become significant when ethical/unethical business decisions are made. Finally, although age is tied to antisocial behavior age does not explain a significant amount of variance. The sample was fairly homogenous; hence the findings may be biased. However, O'Fallon and Butterfield's (2005) review on age and ethics is also inconclusive. From the findings, retailers need to reassess the adage that as one gets older one becomes more ethical. This is not necessarily the case. The research limitations should be seen as markers for future investigations. For example, in order to analyze more constructs a larger sample is needed and, depending upon the model tested, either a more homogenous/heterogeneous sample chosen. Moral philosophy needs to be better operationalized as well as somehow decreasing the questionnaire length yet increasing the amount of information obtained. Increasing the number of vignettes from three to many, as well as including more theoretical constructs for testing needs to be explored. For factors that influence unethical behavior the literature suggests that significant others (Robideaux, 1989), philosophy type (Fraedrich, 1993), and perhaps corporate culture (Herndon, 1991; 1992) have an impact. While some retailers have codes of ethics, many are quite general in language and scope. Research on whether the antecedents of customers' perceptions of retail ethics include such factors as product price or retailer advertising is needed. Work on helping retailers prevent unethical behavior; creating and communicating ethical guidelines; selecting hiring and training ethical salespeople; and measuring the impact of both ethical and unethical behavior on performance is also needed. Scale development from all three perspectives is a research opportunity, as is theory development for explaining retailer ethics. Future research within the empirical testing of ethical decision making models needs to investigate the anomalies discovered as well as testing the ethicalness construct. This is important relative to the issue of situational ethics (Kellaris, Dahlstrom and Boyle, 1996; Banerjee et al., 1998). Another issue is the differences between how individuals make decisions concerning ethical issues and dilemmas. An ethical issue is one that has some synergy between the three ethics domains as it relates to right/wrong choices. An ethical dilemma is where the matrix of ethical/unethical is not synergistic between the three domains; hence, the individual believes that there are only wrong/unethical solutions to the situation. Kohlberg's vignettes are excellent examples of dilemmas where decisions involve breaking laws, death, and stealing to survive. Far too often in business, the situations are not so severe. Retailers need to become more proactive in their codes and their effectiveness relative to sales training, amount and type of supervision, and desired type of salespeople. The multinational aspect of the models being used and tested is very important. Much of the theoretical work is done with a western perspective. Many of the scales used were developed in western cultures. The question is whether researchers are missing unique factors in other cultures thus restricting generalizability. For example, would the values construct change if the retailers tested were from predominantly Muslim, Buddhist,
840
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841
Hindu, or Jewish samples? Are academicians actually capturing the pertinent philosophies associated with all business people? And finally, would the ethicalness construct be the same in other cultures? A major contribution to the field is that business ethics is not just about formal rules and procedures, or how long or familiar someone is with the organization. The acceptance of authority, religion, and age are not very important relative to the ethicalness of individuals within business either. The findings suggest that friends, business associates, superiors, and business associates that are friends, contribute most to ethicalness. The findings suggest that the adage, lie down with dogs get up with fleas, is one step closer to being empirically proven within business ethics. This research does not purport to be error free. Many choices between parsimony and data richness were made. Not all constructs identified in all of the ethics models were incorporated. Yet the contribution to the literature is in testing three major and ten associated constructs together and demonstrating their relative explanatory strengths to a construct that explains the multi-faceted reality of ethics. Our research is a continuation of the many researchers involved in teasing out, bit by bit, the secrets of how ethically charged decisions are made, what causes business people to make unethical choices, and how we can help the individual, organization, and society make better decisions. References Agnew R. Why do they do it? An examination of the intervening mechanisms between social control variables and delinquency. J Res Crime Delinq 1993;30(3):245–66. Akaah IP. An organizational culture and ethical research behavior. J Acad Mark Sci 1993;21(1):59–63. Anderson JC, Gerbing DW. Structural equation modeling practice: a review and recommended two-step approach. Psychol Bull 1988;103:411–23. Bagozzi RP, Yi Y. Assessing construct validity in organizational research. Adm Sci Q 1991;36:459–89. Banarjee D, Cronan TP, Jones TW. A modeling IT ethics—a study in situational ethics. MIS Q 1998;22(1):31–60. Barnett T, Bass K, Brown G. Religiosity, ethical ideology, and intentions to report a peers' wrong doing. Journal of Business Ethics 1996;15(11): 1161–75 [November]. Bartels R. A model for ethics in marketing. J Mark 1967;31:20–6 [January]. Boyce WD, Jensen LC. Moral Reasoning: a Psychological–Philosophical Integration. Lincoln, Nebraska: University of Nebraska Press; 1978. Brass DJ, Butterfield KD, Skaggs BC. A relationship and unethical behavior—a social network perspective. Acad Manage Rev 1998;23(1):14–31. Carlson DS, Kacmar MK. A perception of ethics across situations: a view through three different lenses. J Bus Ethics 1997;16(2):147–60 [February]. Cunningham, M. Walking the thin line: a role conflict model of ethical decisionmaking behavior in the marketing research process. 1991; Ph.D. dissertation, Texas A & M University. DeConinck JB, Lewis WF. The influence of deontological and teleological considerations and ethical climate on sales. J Bus Ethics 1997;16(5): 497–507 [April]. Ferrell OC, Gresham L. A contingency framework for understanding ethical decision-making in marketing. J Mark 1985:87–96 [Summer]. Ferrell OC, Skinner SJ. Ethical behavior and bureaucratic structure in marketing research organizations. J Mark Res 1988;25:103–9 [February]. Ferrell OC, Gresham L, Fraedrich JP. A synthesis of ethical decision models for marketing. J Macromark 1989;9(2):55–64 [Fall].
Ferrell OC, Fraedrich JP. Moral philosophies and marketing managers:an empirical investigation. J Mark Manage 1991;1(1):18–27 [Fall]. Fornell C, Larcker DF. Evaluating structural equation models with unobservable variables and measurement error. J Mark Res 1981;18:39–50. Fort TL. Religion and business ethics: the lessons from political morality. J Bus Ethics 1977;16: 263–73. Fraedrich, J.P. Philosophy type interaction in the ethical decision making process of retailers. 1988; Ph.D. Dissertation, Texas A & M University. Fraedrich JP. The ethical behavior of retail managers. J Bus Ethics 1993(12): 1–12. Fraedrich JP, Ferrell OC. The impact of perceived risk and moral philosophy type on ethical decision making in business organizations. J Bus Res 1992;24(4):283–96 [June]. Fraedrich JP, Ferrell OC. Cognitive consistency of marketing managers in ethical situations. J Acad Mark Sci 1992a;20(3):245–52. Fritzsche DJ, Becker H. Ethical behavior of marketing managers. J Bus Ethics 1983;2:291–9. Giacalone RA, Jurkiewicz CL. Right from wrong: the influence of spirituality on perceptions of unethical business activities. J Bus Ethics 2003;46(1): 85–97. Graham J, Bowling B. Young people and crime, home office research study. #145. UK: Research and Statistics Department; 1996. UK Home Office. Grimes AJ. Authority, power, influence and social control: a theoretical synthesis. Acad Manage Rev 1978;3:724–35 [October]. Gundlach GT, Murphy PE. Ethical and legal foundations of relational marketing exchanges. J Mark 1993;57:35–46 [October]. Hawkins JD. Delinquency and crime: current theories. Cambridge Criminology Series. Cambridge University Press; 1996. Heider F. The psychology of interpersonal relations. New York: Wiley; 1958. Herndon NC. Commitment, Satisfaction and Turnover. The Impact of Individual Moral Value Structures when exposed to a Conflicting or Matching Organizational Sales Culture. Doctoral Dissertation, Texas A & M University; 1991. Herndon NC. Perceived corporate ethicalness and salesforce turnover: a new rationale for more ethical corporate behavior. Dev Mark Sci 1992:330–5 XV. Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Model 1999;6(1): 1–55. Hunt SD, Vitell S. A general theory of marketing ethics. J Macromark 1986;6: 5–16 [Spring]. Hunt SD, Chonko LB, Wilcox JB. Ethical problems of marketing researchers. J Mark Res 1984;21:309–24 [August]. John G. An empirical investigation of some antecedents of opportunism in a marketing channel. J Mark Res 1984;21:278–89 [August]. Jones TM, Ryan LV. The link between ethical judgment and action in organizations—a moral approbation approach. Organization Science 1997;8(6):663–80. Joreskog KG, Sorbom D. LISREL 8.30: A guide to the program and applications. Chicago: Scientific Software International; 1993. Kellaris JJ, Dahlstrom RF, Boyle BA. Contextual bias in ethical judgment of marketing practices. Psychol Mark 1996;13(7):677–94 [October]. Kennedy EJ, Lawton L. The effects of social and moral integration on ethical standards: a comparison of American and Ukrainian Business students. J Bus Ethics 1996;15(8):901–11. Kerlinger FN. Foundations of behavioral research. New York: Holt, Rinehart and Winston; 1986. Kirkpatrick DD. Investors Circle Columbia/HCA Healthcare Corp. Wall Street J 1997:B12 [October 22]. Krugman DM, Ferrell OC. The organizational ethics of advertising: corporate and agency views. J Advert 1981;10(1):21–48. Loe TW, Ferrell L, Mansfield P. A review of empirical studies assessing ethical decision making in business. J Bus Ethics 2000;25(3):185–204. MacCallum RC. Specification searches in covariance structure modeling. Psychol Bull 1986;100:107–20. Malhotra NK, Miller GL. An integrated model for ethical decisions in marketing research. J Bus Ethics 1998;17(3):263–80. McDaniel SW, Burnett JJ. Consumer religiosity and retail store evaluative criteria. J Acad Mark Sci 1990;18(2):101–12 [Spring].
J. Fraedrich, R. Iyer / Journal of Business Research 61 (2008) 834–841 McDaniel SW, Burnett JJ. Targeting the evangelical market segment. J Advert Res 1991;31:26–33 [August–September]. Merton RK. The role-set. Br J Sociol 1957(8):106–20. O'Fallon MJ, Butterfield KD. A review of the ethical empirical decision-making literature: 1996–2003. J Bus Ethics 2005;50:375–413. Ouchi WG. The relationship between organizational structure and organization control. Adm Sci Q 1977(22):95–113. Powpaka S. Factors affecting managers' decision to bribe: an empirical investigation. J Bus Ethics 2002;40(3):227–46. Razzaque MA, Hwee TP. Ethics and purchasing dilemma: a Singaporean view. J Bus Ethics 2002;35(4):307–26. Reidenbach RE, Robin DP. An application and extension of a multidimensional ethics. J Acad Mark Sci 1991;9(2):83–93 [Spring]. Robideaux DR. Significant others and ethical behavior of retail managers. Doctoral Dissertation 1989; Louisiana Tech University. Singer MS, Singer AE. Observer judgments about moral agents' ethical decisions: the role of scope of justice and moral intensity. J Bus Ethics 1997;16(5):473–84 [April]. Singhapakdi A, Vitell SJ. Marketing ethics: factors influencing perceptions of ethical problems and alternatives. J Macromark 1990:4–18 [Spring]. Singhapakdi A, Marta JK, Rallapalli KC, Rao CP. Toward an understanding of religiousness and marketing ethics: an empirical study. J Bus Ethics 2000;27(4):305–19. Sparks JR, Hunt SD. Marketing researcher ethical sensitivity: conceptualization, measurement, and exploratory investigation. J Mark 1998;62(2):92–109 [April]. Sutherland E, Crissey DR. Principles of criminology. 8th edition. Chicago: J.B. Lippincott; 1970.
841
Tepstra DE, Rozell EJ, Robinson RK. The influence of personality and demographic variables on ethical decisions related to insider trading. J Psychol 1993(4):375–89. Thorne-LeClair D, Ferrell OC, Fraedrich J. Integrity management. Tampa: University of Tampa Press; 1998. Trevino LK. Ethical decision making in organizations: a person–situation interactionist model. Acad Manage Rev 1986(3):601–17. Tse ACB, Au AKM. Are New Zealand business students more unethical than non-business students? J Bus Ethics 1997;16(4):445–50. Upchurch R. A conceptual foundation for ethical decision-making: a stakeholder perspective in the lodging industry. J Bus Ethics 1998;17(12): 1349–61. Vitell SJ, Hunt SD. The general theory of marketing ethics: a partial test of the model. Res Mark 1990;10:237–65. Warr M. Parents, peers, and delinquency. Soc Forces 1993;1(72):247–64 [September]. Wilkes RE, Burnett JJ, Howell RD. On the meaning and measurement of religiosity in consumer research. J Acad Mark Sci 1986;14(Spring):47–56. Withey S. The US and USSR: a report of the public’s perspective on the United States-Russian relations in late 1961. In: Borrow D, editor. Components of Defense Policy. Chicago: Rand McNally; 1965. pp. 164–174. Zey-Ferrell MK, Weaver M, Ferrell OC. Predicting unethical behavior among marketing practitioners. Hum Relat 1979;7(32):557–69. Zey-Ferrell MK, Ferrell OC. Role-set configuration and opportunity as predictors of unethical behavior in organizations. Hum Relat 1982;7(35): 587–604.