Introduction

Consumer ethics continues to be an important economic and social issue across the globe, e.g. according to a recent global survey conducted by the BSA (a software alliance representing software makers), the rate of unlicensed software installation in 2015 across the globe was 39% with a commercial value of $52.2 billion; in the USA, this rate was 17% with a commercial value of $9.1 billion (BSA 2016). Consumers also engage in ethically positive activities, e.g. the global fair trade (the world’s largest ethical label) market in 2013 was estimated at £4.4 billion; in the USA, the fair trade market increased by £258 m between 2012 and 2013 (Smithers 2014 in the Guardian). Marketing scholars have examined different personal/psychographic variables as antecedents to consumer ethics (see Vitell 2015 for a review). After reviewing this stream of research on psychological variables and consumer ethics, Dubinsky et al. (2005, p. 1691) noted that “essentially, prior scholars have explored personal variables vis-a-vis ethical decision making or moral philosophy in a seemingly ad hoc fashion without grounding their selected variables in theory”. In response to this concern, the primary contribution of this study is to examine moral foundations theory (Haidt and Graham 2007) as a theoretical framework to explain consumers’ beliefs regarding both ethical and unethical consumption. In particular, this study investigates whether various moral foundations can explain the effects of personal variables on consumer ethics.

Dubinsky et al. (2005) proposed Kohlberg’s (1969) cognitive moral development (CMD) theory as a framework for explaining the effects of personal variables on consumers’ ethical beliefs. CMD theory is a model of moral decision-making that is based upon changes in cognitive development. However, the ‘sentimentalist’ tradition of moral philosophy championed by Hume (1777/1960) has argued that sentiments (emotions) and not reasoning are the primary drivers of moral judgements. Similarly, Haidt (2001) proposed the social intuitionist model of moral judgement’ which posits that emotions and intuitions play a key role in moral decision-making whereas cognition plays a relatively less prominent role (see also Hardy 2006)—in many cases reasoning is involved in post hoc justification of decisions made by intuitions and emotions.

Since CMD theory (Kohlberg 1969) is principally based on cognitive processes, it cannot comprehensively explain the range of responses that transpire in ethical choices. Even more importantly, as noted by Haidt (2012), the Kohlberg (1969) model is focused on only two moral domains or psychological foundations, namely justice (fairness) and caring. Haidt and Graham (2007) propose moral foundations theory as an alternative model of morality that encompasses multiple moral foundations (more than only fairness and caring). Moral foundations theory (Haidt and Graham 2007) is an overarching model that can account for both emotional and cognitive moral decision-making. Moral foundations theory (Haidt and Graham 2007) has been applied in various contexts, e.g. business ethics (Sadler-Smith 2012), legal issues (Buccafusco and Fagundes 2015) and charitable giving (Winterich et al. 2012). In order to examine the validity of moral foundations theory (Haidt and Graham 2007) as a framework for understanding consumer ethics, this study relates moral foundations theory to both ethical and unethical consumer behaviour.

Moral Foundations Theory

Moral foundations theory (Haidt and Graham 2007) is based upon a broadening of the definition of morality to include perspectives from evolutionary psychology and anthropology. Haidt and Graham (2007) proposed the following five psychological foundations (moral domains) that affect moral decision-making: care/harm foundation, fairness/cheating foundation, sanctity/degradation foundation, authority/subversion foundation and loyalty/betrayal foundation. The first two moral foundations—care/harm and fairness/cheating—correspond to Gilligan’s (1982) ‘care’ and Kohlberg’s (1969) ‘justice’ framework of ethics (Haidt 2012). The care/harm foundation is an outcome of the adaptive challenge that humans have of caring for vulnerable offspring. Graham et al. (2013, p. 69) state, “whatever functional systems made it easy and automatic to connect perceptions of suffering with motivations to care, nurture, and protect are what we call the care/harm foundation”. The care/harm foundation is concerned with welfare and generates the emotion of compassion (Haidt 2012).

In regard to the fairness/cheating domain, Graham et al. (2013, p. 69) state that social animals “whose minds are organised in advance of experience to be highly sensitive to evidence of cheating and cooperation, and to react with emotions that compel them to play ‘tit for tat’ (Trivers 1971), had an advantage over those who had to figure out their next move using their general intelligence”. The fairness/cheating foundation is concerned about issues of proportionality/justice (Haidt and Graham 2007) and elicits the emotion of anger (Haidt 2012).

The sanctity/degradation foundation is based on the emotion of disgust, which helps individuals to develop a ‘behavioural immune system’ (Schaller and Park 2011) that provides protection from pathogens and parasites. Graham et al. (2013, p. 71) state, “individuals whose minds were structured in advance of experience to develop a more effective ‘behavioural immune system’ (Schaller and Park 2011) likely had an advantage over individuals who had to make each decision based purely on the sensory properties of potential foods, friends and mates”. Koleva et al. (2012) point out that the sanctity/degradation foundation responds with the emotion of disgust not only to biological contaminants but also to social contaminants such as corruption and other non-physical transgressions that break social conventions (e.g. flag burning).

The authority/subversion foundation is based on respect for dominance hierarchies in society. Adhering to dominance hierarchies provides the evolutionary adaptive benefits of social cooperation and stability (Graham et al. 2013). Depending on where one stands in the social hierarchy, the authority/subversion foundation supports obedience or leadership (Koleva et al. 2012). Two key emotions that are characteristic of the authority/subversion foundation are respect and fear (Haidt 2012).

The loyalty/betrayal foundation is based on intragroup solidarity and intergroup competition. Justifying the evolutionary adaptive benefit of this foundation, Graham et al. (2013, p. 70) note that “individuals whose minds were organised in advance of experience to make it easy for them to form cohesive coalitions were more likely to be part of winning teams in such competitions”. Individuals and actions that build social cohesion are favoured by the loyalty/betrayal foundation (Koleva et al. 2012), and (group) pride is a key emotion related to this foundation (Haidt 2012). In terms of ethical implications, loyalty is a double-edged sword. Loyalty can be a virtue making other virtues, e.g. honesty, compassion, salient (Hildreth et al. 2016); however, it can also be a vice as loyalty may lead to preferential treatment of in-group members (Padgett and Morris 2005).

Dimensions of Consumer Ethics

Vitell and Muncy (2005) identify five dimensions of consumer actions from an ethical perspective that are on a continuum ranging from outright illegal activities to pro-social actions. Actions that are clearly illegal and benefit the consumer are classified as part of the ‘active, illegal’ dimension of consumer ethics, e.g. reporting a lost item as stolen to an insurance company in order to collect the insurance money. Actions that allow consumers to passively benefit from the mistakes of the seller are classified as part of the passive dimension of consumer ethics, e.g. moving into a residence, finding that the cable (pay) TV is still hooked up and using it without paying for it. Actions that may be technically legal but are questionable in nature that benefit the consumer are classified as part of the ‘active, legal’ dimension of consumer ethics, e.g. returning merchandise to a store by claiming it was a gift when it was not. Consumer actions that may be considered unethical by some (but not everyone) although they do not cause direct harm to others are classified as ‘no harm, no foul’ actions, e.g. spending over an hour trying on clothing and not buying any. Finally, consumer actions that are pro-social (altruistic, pro-environmental etc.) are classified as ‘doing good’ actions, e.g. correcting a bill that has been miscalculated in your favour.

Moral foundations theory (Haidt and Graham 2007) provides an alternative framework for examining consumer ethics. As a theoretical model, moral foundations theory (Haidt and Graham 2007) supplements the Hunt–Vitell theory of marketing ethics (Hunt and Vitell 1986). Bray et al. (2011) note that the Hunt–Vitell theory of marketing ethics (Hunt and Vitell 1986) is based on the philosophic traditions of deontology (obligations and rules) and teleology (guided by consequences of actions). However, according to Haidt (2012), both deontology in the tradition of Kant (1785/1993) and teleology in the tradition of Bentham (1789/1996) are based on reasoning and systematic thought, rather than being based on emotions as proposed in the sentimentalist moral theory of Hume (1777/1960). Moral foundations theory (Haidt and Graham 2007) allows for a prominent role for emotions in addition to reasoning in ethical judgements (Haidt 2012). In the following sections, multiple hypotheses regarding the relationships between various moral foundations and (1) unethical consumer actions (‘active, illegal’ dimension, passive dimension and ‘active, legal’ dimension), (2) ‘no harm, no foul’ actions and (3) ‘doing good’ actions are developed.

The Relationships Among the Various Moral Foundations and Beliefs Regarding Unethical Consumer Actions (‘Active, Illegal’, Passive and ‘Active, Legal’ Dimensions)

The care/harm foundation is sensitive to actions that harm others (Haidt 2012). When consumers are actively benefiting from illegal actions, this harms other parties in a transaction (e.g. the seller). Hence, the care/harm foundation should be negatively related to beliefs regarding the ‘active, illegal’ dimension of consumer ethics. The care/harm foundation is closely associated with the emotion of compassion (Haidt 2012), which is related to empathy (Eisenberg 2002). Empathy has been shown to be negatively related to the passive dimension and the ‘active legal’ dimension of consumer ethics (Chowdhury and Fernando 2014). Thus, the care/harm foundation should also be negatively related to beliefs regarding these dimensions. Actively benefiting from illegal actions, passively benefiting from the mistakes of the seller and actively benefiting from legal but questionable actions should also trigger the fairness/cheating foundation as these actions deceive the seller and give disproportionate benefits to the consumer. There should be negative relationships between the fairness/cheating foundation and beliefs regarding these dimensions.

Feinberg and Willer (2013, p. 57) refer to the sanctity/degradation moral foundation as “concerns about preserving purity and sacredness often characterised by a disgust reaction”. Tybur et al. (2009) demonstrated that illegal actions by others, such as stealing bank information online, stealing from a neighbour etc., generate the emotion of disgust. Tybur et al. (2009) also demonstrated that moral disgust can result from evaluating non-reciprocity, which can be considered as a passive act of omission, e.g. when a member of a work group chooses not to contribute anything but shares equally in all the benefits. Furthermore, Tybur et al. (2009) found that legal but questionable actions generate moral disgust, e.g. a business owner making a very high salary but keeping his employees at minimum wage. These findings suggest that the sanctity/degradation foundation is negatively related to beliefs regarding the ‘active, illegal’, passive and ‘active, legal’ dimensions of consumer ethics.

In reference to respect for authority, Graham et al. (2013, p. 70) state that, “the various modules that comprise the authority/subversion foundation are often at work when people interact with and grant legitimacy to modern institutions such as law courts and police departments, and to bosses and leaders of many kinds”. Respect for authority should lead to avoiding overtly illegal actions, as legal authorities strictly forbid these actions. Hence, there should be a negative relationship between the authority/subversion foundation and beliefs regarding the ‘active, illegal’ dimension. Passively benefiting from the mistakes of the seller takes the form of acts of omission, and from a legal authority standpoint acts of omission and acts of commission are treated very differently; acts of omission in many cases are not proscribed by legal authorities (Simester 1995). In regard to the ‘active legal’ dimension, by definition there are no legal mandates from authorities to abstain from such actions. Thus, the authority/subversion foundation should not be related to beliefs regarding the passive dimension or ‘active, legal’ dimension.

Moral traits made salient by loyalty, e.g. honesty, compassion (Hildreth et al. 2016), are related to the fairness/cheating and care/harm foundations. A benefit of using moral foundations theory to understand the relationship between loyalty and consumer ethics is that the effects of moral foundations other than the loyalty/betrayal foundation (e.g. care/harm, fairness/cheating) on consumer ethics can be controlled for, and the specific effects of loyalty can be isolated. As the positive effects of loyalty in relation to ethics seem to be primarily driven by making the care/harm and fairness/cheating foundations salient, after controlling for the effects of the care/harm and fairness/cheating foundations, the loyalty/betrayal foundation should be positively related to unethical actions (‘active, illegal’, passive and ‘active, legal’ dimensions). This supports research that has found negative effects of loyalty on ethicality (Umphress et al. 2010). Hence, the loyalty/betrayal foundation should be positively related to beliefs regarding the ‘active, illegal’, passive and ‘active, legal’ dimensions of consumer ethics.

The following hypotheses are proposed in regard to the relationships of the various moral foundations with beliefs regarding the ‘active, illegal’, passive and ‘active, legal’ dimensions.

H1a) The care/harm foundation, H1b) the fairness/cheating foundation, H1c) the sanctity/degradation foundation and H1d) the authority/subversion foundation are related to negative beliefs regarding the ‘active, illegal’ dimension. H1e) The loyalty/betrayal foundation is related to positive beliefs regarding the ‘active, illegal’ dimension.

H2a) The care/harm foundation, H2b) the fairness/cheating foundation and H2c) the sanctity/degradation foundation are related to negative beliefs regarding the passive dimension. H2d) The authority/subversion is not related to beliefs regarding the passive dimension. H2e) The loyalty/betrayal foundation is related to positive beliefs regarding the passive dimension.

H3a) The care/harm foundation, H3b) the fairness/cheating foundation and H3c) the sanctity/degradation foundation are related to negative beliefs regarding the ‘active, legal’ dimension. H3d) The authority/subversion is not related to beliefs regarding the ‘active, legal’ dimension. H3e) The loyalty/betrayal foundation is related to positive beliefs regarding the ‘active, legal’ dimension.

The Relationships Among the Various Moral Foundations and Beliefs Regarding ‘No Harm, No Foul’ Actions

Consumers do not deem ‘no harm, no foul’ actions as harmful to others (Vitell et al. 1991), and hence, these actions should not trigger the care/harm moral foundation. In the case of ‘no harm, no foul’ actions, the seller is not perceived to be a victim and hence these actions should also not be considered unfair. As consumers do not perceive ‘no harm, no foul’ actions to generate harm or to be unfair, this would also indicate that those consumers who find these actions to be unacceptable are doing so for reasons other than harm avoidance or fairness. Behaviours violating the sanctity/degradation foundation elicit condemnation even when they do not generate harm (Haidt 2012) but violate social norms, e.g. spending over an hour trying on clothing and not buying anything, may not be clearly linked to specific harm or fairness issues but may be considered as a violation of social conventions. Hence, the sanctity/degradation foundation should be negatively related to beliefs regarding ‘no harm, no foul’ actions.

Many types of ‘no harm, no foul’ actions are not specifically restricted by authorities (e.g. returning merchandise after buying it and not liking it is not prohibited by authorities). Thus, for ‘no harm, no foul’ actions, there should be no relationship with the authority/subversion foundation. ‘No harm, no foul’ actions are not assumed to be unfair or harmful by many consumers (Vitell et al. 1991); hence, these actions should also not be expected to provide disproportionate benefits to in-group members. Thus, in-group loyalty should also be unrelated to ‘no harm, no foul’ actions.

The following hypotheses are proposed in regard to the relationships of the various moral foundations with beliefs regarding ‘no harm, no foul’ actions:

H4a) The care/harm foundation and H4b) the fairness/cheating foundation are not related to beliefs regarding ‘no harm, no foul’ actions. H4c) The sanctity/degradation foundation is related to negative beliefs regarding ‘no harm, no foul’ actions. H4d) The authority/subversion foundation and H4e) the loyalty/betrayal foundation are not related to beliefs regarding ‘no harm, no foul’ actions.

The Relationships Among the Various Moral Foundations and Beliefs Regarding ‘Doing Good’ Actions

Consumers engage in pro-social actions to alleviate harm and enhance caring for others. Hence, the care/harm foundation should be positively related to pro-social consumer actions. This is in line with recent research that has demonstrated that the care/harm moral foundation is positively related to climate friendly consumption (Vainio and Makiniemi 2016). ‘Doing good’ actions are pro-social actions that are also based on being fair to others. For example, not buying from companies that do not treat their employees fairly is a prototypical ‘doing good’ action (Vitell and Muncy 2005). This indicates that the fairness/cheating foundation should be positively related to perceptions of pro-social actions.

Feinberg and Willer (2013) demonstrate that pro-environmental messages couched in the sanctity/degradation domain lead to pro-environmental attitudes, particularly among those who have conservative political inclinations (who score high on the sanctity/degradation foundation). Thus, the sanctity/degradation foundation should be positively associated with beliefs regarding pro-social actions. Civil authorities encourage consumers to participate in pro-social actions, e.g. recycling and pro-environmental buying. Hence, the authority/subversion foundation should be positively related to beliefs regarding ‘doing good’ actions. Khatri and Tsang (2003, p. 298) state, “in-group loyalty often leads to out-group derogation, and in-group cooperation is often coupled with fierce out-group competition.” Pro-social consumer actions are altruistic in nature and do not discriminate between in-group members and out-group members (benefiting both); hence, as in-group loyalty increases it is likely that this will lead to less support for ‘doing good’ actions.

The following hypotheses are proposed in regard to the relationships of the various moral foundations with beliefs regarding ‘doing good’ actions:

H5a) The care/harm foundation, H5b) the fairness/cheating foundation, H5c) the sanctity/degradation foundation and H5d) the authority/subversion foundation are related to positive beliefs regarding ‘doing good’ actions. H5e) The loyalty/betrayal foundation is related to negative beliefs regarding ‘doing good’ actions.

The Mediating Role of Moral Foundations in Relation to the Effects of Idealism on Consumer Ethics

Idealism is a personal characteristic that has been consistently shown to be related to negative beliefs regarding unethical consumption (e.g. Arli and Pekerti 2016; Chowdhury 2017; Steenhaut and Van Kenhove 2006; Vitell and Paolillo 2003 etc.) and to be related to positive beliefs regarding pro-social consumption (e.g. Chowdhury 2017). Forsyth (1980) in his Ethics Position Questionnaire (EPQ) proposed idealism as one type of personal moral philosophy (relativism is the other type). Forsyth (1992, p. 462) defines idealism as “the individual’s concern for the welfare of others. Highly idealistic individuals feel that harming others is always avoidable, and they would rather not choose between the lesser of two evils, which will lead to negative consequences for other people”. However, the philosophic grounding of idealism is debatable, as it does not clearly align with established ethical decision-making approaches, e.g. utilitarianism, deontology. Another concern is that Forsyth (1980) originally suggested a typology approach (combining high/low idealism with high/low relativism) for identifying individuals’ ethical ideologies rather than focusing on idealism and relativism separately.

Davis et al. (2001) conducted a detailed investigation of the validity and reliability of the EPQ (Forsyth 1980). Davis et al. (2001) discouraged the use of the typology approach due to the loss of predictive power through categorisation and noted that in subsequent work Forsyth also admits concerns with the typology approach (see Forsyth and Nye 1990). In relation to idealism and relativism, Davis et al. (2001, p. 49) stated, “the effects of relativism on moral judgment were rather meagre. However, the idealism scale exhibited reasonable degree of validity when employed as a predictor of moral judgements. Its predictive strength, though modest, did not change even when personal attributes such as age and gender were included as factors in the regression model. Thus, differences among individuals in their concern for the welfare of others may be a particularly useful ‘personality’ variable for investigating moral judgment”.

In line with Davis et al. (2001), this study does not examine relativism, although the study does examine idealism as a predictor of consumer ethics. However, instead of examining idealism separately as an independent variable in relation to consumer ethics, this study identifies mediating (intervening) variables between idealism and dimensions of consumer ethics. Hayes (2013, p. 7) states, “intervening variables, often called mediators, are conceptualised as the mechanism through which X influences Y” (X is the independent variable, i.e. idealism and Y is the dependent variable, i.e. consumers’ ethical beliefs). The various moral foundations (Haidt and Graham 2007) are proposed as mediators of the effects of idealism on consumer ethics. In the event that it can be empirically established that moral foundations theory (Haidt and Graham 2007) explains the effects of idealism on consumer ethics, it would be theoretically appropriate for future studies in consumer ethics to focus on the role of moral foundations rather than relying upon personal moral philosophies (e.g. idealism). In order to establish such mediation, it is first required to identify conceptual relationships between idealism and various moral foundations.

Idealism, which by definition relates to welfare orientation, should be positively related to the care/harm foundation, as this foundation is sensitive to the welfare of others (Haidt and Graham 2007). Similarly, idealism is also positively related to perceptions of the fairness/cheating foundation, as idealist consumers have concerns for equity and fairness (Sidani et al. 2014). Glenn et al. (2009) found that psychopathy is positively associated with both idealism and the sanctity/degradation foundation. There are two dimensions of psychopathy: a) personality/emotional features of psychopathy and b) antisocial lifestyle/behavioural features of psychopathy (Glenn et al. 2009). Specifically, the personality/emotional features dimension of psychopathy was positively associated with both idealism and the sanctity/degradation foundation; however, the antisocial lifestyle/behavioural features dimension of psychopathy was not associated with both idealism and the sanctity/degradation foundation (Glenn et al. 2009). These findings indicate that idealism and the sanctity/degradation foundation are positively associated. Idealism is also related to collectivism, which is associated with both respect for authority and loyalty to community (Forsyth et al. 2008).

Previous research has demonstrated that idealism is negatively related to beliefs regarding the ‘active, illegal’, passive and ‘active, legal’ dimensions of consumer ethics (e.g. Steenhaut and Van Kenhove 2006). As discussed above idealism is also positively related to the various moral foundations. In earlier sections, the various moral foundations have been related to beliefs regarding the ‘active, illegal’ dimension. Hence, it can be predicted that these moral foundations mediate the effects of idealism on beliefs regarding the ‘active, illegal’ dimension of consumer ethics. In earlier sections, the various moral foundations (except authority/subversion) have also been related to beliefs regarding the passive and ‘active, legal’ dimensions. Thus, it can be predicted that the various moral foundations (except authority/subversion) mediate the effects of idealism on beliefs regarding the passive and ‘active, legal’ dimension of consumer ethics. The following hypotheses are proposed:

The relationship between idealism and beliefs regarding the ‘active, illegal’ dimension H6a) is mediated by the care/harm foundation, H6b) is mediated by the fairness/cheating foundation, H6c) is mediated by the sanctity/degradation foundation, H6d) is mediated by the authority/subversion foundation and H6e) is mediated by the loyalty/betrayal foundation.

The relationship between idealism and beliefs regarding the passive dimension H7a) is mediated by the care/harm foundation, H7b) is mediated by the fairness/cheating foundation, H7c) is mediated by the sanctity/degradation foundation, H7d)is not mediated by the authority/subversion foundation and H7e) is mediated by the loyalty/betrayal foundation.

The relationship between idealism and beliefs regarding the ‘active, legal’ dimension H8a) is mediated by the care/harm foundation, H8b) is mediated by the fairness/cheating foundation, H8c) is mediated by the sanctity/degradation foundation, H8d)is not mediated by the authority/subversion foundation and H8e) is mediated by the loyalty/betrayal foundation.

Idealism has been shown to be negatively related to beliefs regarding ‘no harm, no foul’ actions (Steenhaut and van Kenhove 2006). As noted earlier, idealism is also positively related to the various moral foundations. In earlier sections, the sanctity/degradation moral foundation has been related to beliefs regarding ‘no harm, no foul’ actions. Hence, it can be proposed that the sanctity/degradation moral foundation (but not the other foundations) mediates the effects of idealism on beliefs regarding ‘no harm, no foul’ actions. The following hypotheses are proposed:

The relationship between idealism and beliefs regarding ‘no harm, no foul’ actions H9a)is not mediated by the care/harm foundation, H9b)is not mediated by the fairness/cheating foundation, H9c) is mediated by the sanctity/degradation foundation, H9d)is not mediated by the authority/subversion foundation and H9e)is not mediated by the loyalty/betrayal foundation.

Previous research indicates that idealism is positively related to ‘doing good’ actions (Chowdhury 2017). As discussed above, idealism is also positively related to the various moral foundations. In earlier sections, it has been proposed that all the various moral foundations are related to beliefs regarding ‘doing good’ actions. Hence, it can be proposed that these moral foundations mediate the relationship between idealism and beliefs regarding ‘doing good’ actions. The following hypotheses are proposed:

The relationship between idealism and beliefs regarding ‘doing good’ actions H10a) is mediated by the care/harm foundation, H10b) is mediated by the fairness/cheating foundation, H10c) is mediated by the sanctity/degradation foundation, H10d) is mediated by the authority/subversion foundation and H10e) is mediated by the loyalty/betrayal foundation.

Figure 1 depicts the proposed mediating roles of various moral foundations in regard to the relationships of idealism with different dimensions of consumer ethics.

Fig. 1
figure 1

Proposed mediating roles of moral foundations in regard to the relationships of idealism with dimensions of consumer ethics

Full size image

In order to test the proposed hypotheses, an online survey was conducted with consumers in the USA. The details of the method used in this study are described below.

Method

Sample

A total of 450 US consumers who are part of an online consumer panel managed by Qualtrics, a leading market research firm, participated in this study. Participants were sampled from the panel based on a quota of gender distribution representative of the US population (48.9% male, 51.1% female). As the survey was closed once a predetermined number of respondents were reached, technically a response rate cannot be calculated. The participants were informed that the survey includes questions about moral foundations, personal moral philosophies and consumer ethics (i.e. there was no deception). They were further informed that participation was voluntary and that responses would be anonymous. The demographics of the sample are provided in Table 1.

Table 1 Demographic characteristics of sample
Full size table

Measures

The various moral foundations were measured based on the moral foundations questionnaire (MFQ, Graham et al. 2011). The MFQ has two sections. The first section of the MFQ measures the moral relevance of various considerations when deciding whether something is right or wrong (e.g. the item “whether or not someone suffered emotionally” measures the care/harm foundation). Respondents reported the extent to which each item was relevant to them when making decisions about right or wrong using a six-point scale, (1 = “not at all relevant”, 6 = “extremely relevant”, similar to Graham et al. 2011). There were fifteen items for assessing moral relevance—three for each of the five moral foundations.

The second section of the MFQ measures moral judgments of specific normative statements (e.g. “compassion for those who are suffering is the most critical value” is a statement that measures the care/harm foundation). Agreement with these statements was measured on a six-point scale (1 = “strongly disagree”, 6 = “strongly agree” similar to Graham et al. 2011). There were fifteen moral judgment statements—three for each of the five moral foundations. The score for each of the moral foundations was based on the average responses to the moral relevance items and the moral judgment items pertaining to that foundation (six items for each foundation).

Idealism was measured with the ethics position questionnaire (EPQ, Forsyth 1980), which includes ten items for idealism and ten items for relativism. A five-point scale was used that measures agreement with each item (1 = “strongly disagree”, 5 = “strongly agree” similar to Chowdhury 2017). Since relativism is not a variable of interest in this study, it is not included in any of the analyses. The various dimensions of consumer ethics were measured with a scale based on Vitell and Muncy (2005). The scale assesses the ethicality of various consumer actions. There were six items for the ‘active, illegal’ dimension, four items for the passive dimension, five items for the ‘active, legal’ dimension, five items for ‘no harm, no foul’ actions and five items for ‘doing good’ actions. Participants assessed each item on a five-point scale anchored 1 = “strongly believe that this action is wrong” and 5 = “strongly believe that this action is not wrong”. Hence, smaller numbers indicated that consumers believed the action to be ethically more unacceptable.

Out of all the constructs, one of the moral foundations (fairness/cheating, α = 0.64) had a Cronbach’s α that was lower than the minimally acceptable cut-off level of 0.65 recommended by DeVellis (2012). The six items related to this foundation were further examined, and the item with the lowest inter-item correlations was deleted. This item was: “I think it’s morally wrong that rich children inherit a lot of money while poor children inherit nothing”. On deletion of this item, the reliability for the final five-item fairness/cheating scale (α = 0.73) was acceptable. The reliabilities of all the other constructs were acceptable: care/harm, α = 0.69; sanctity/degradation, α = 0.79; authority/subversion, α = 0.66; loyalty/betrayal, α = 0.69; idealism, α = 0.88; ‘active, illegal’ dimension, α = 0.92; passive dimension, α = 0.89; ‘active, legal’ dimension, α = 0.88; ‘no harm, no foul’ actions, α = 0.74 and ‘doing good’ actions, α = 0.86.

In order to test the validity of the measures, a confirmatory factor analysis was conducted with the items for the various moral foundations, the different dimensions of consumer ethics and idealism. The overall measurement model was significant (Chi-square = 5487.18, df = 1897, p < 0.001); however, the model showed acceptable fit (RMSEA = 0.065, SRMR = 0.091, CMIN/df = 2.893). Steiger (2007) states that a model is acceptable if the RMSEA is <0.07, while Weston and Gore (2006) note that SRMR <0.10 is acceptable (particularly for <500 sample size). The rule of thumb for acceptable fit for CMIN/df is <3 (Kline 2004).

In order to test for common method bias, an additional method factor was added in the measurement model. In this procedure, as per Podsakoff et al. (2003, p. 168), “items are allowed to load on their theoretical constructs, as well as on a latent common methods variance factor, and the significance of the structural parameters is examined with and without the latent common methods variance factor in the model”. An examination of the regression weights in the model without the common method factor and in the model with the common method factor revealed that all the regression weights between items and their relevant constructs remained significant even after adding the common method factor. The common method factor accounted for a very small portion of the total variance (6.76%) indicating that common method bias was not an issue.

Results

The means and standard deviations of the variables are reported in Table 2. The correlations among the variables are reported in Table 3.

Table 2 Means and SD of key variables
Full size table
Table 3 Correlations among key variables
Full size table

Testing for the Relationships of the Various Moral Foundations with Beliefs Regarding Unethical Consumer Actions (‘Active, Illegal’ dimension, Passive dimension and ‘Active, Legal’ dimension)

Three separate regression analyses were conducted with the various moral foundations as the independent variables and beliefs regarding the ‘active, illegal’ dimension (see Regression model 1, Table 4), the passive dimension (see Regression model 2, Table 4) and the ‘active, legal’ dimension (see Regression model 3, Table 4) as the dependent variables.

Table 4 Regression analyses
Full size table

In order to check for multicollinearity, the VIF and tolerance (1/VIF) values for the independent variables in all the regression models were reviewed. Multicollinearity is indicated by a VIF value >10 and tolerance values <0.1 (Meyers et al. 2006). The highest VIF was 2.73 and lowest tolerance value was 0.37. Multicollinearity was not an issue in these regression models (or any other regression model used in this study).

The results of Regression model 1 demonstrate that the fairness/cheating foundation and the sanctity/degradation foundation were significantly related to negative beliefs regarding the ‘active, illegal’ dimension. This supports H1b and H1c. The loyalty/betrayal foundation was significantly related to positive beliefs regarding the ‘active, illegal’ dimension. This supports H1e. However, the care/harm foundation and authority/subversion foundation were not significant predictors. Hence, H1a and H1d were not supported. A Ramsey RESET test was conducted to check for specification error in the linear regression. The results indicated that the RESET test was not significant, indicating that there was no specification error (this was the case for all regression models in this study).

The results of Regression model 2 demonstrate that the fairness/cheating foundation and the sanctity/degradation foundation were significantly related to negative beliefs regarding the passive dimension while the authority/subversion foundation was not a significant predictor. This supports H2b, H2c and H2d. The loyalty/betrayal foundation was marginally significantly (p = 0.057) related to positive beliefs regarding the passive dimension. This provides support for H2e. However, the care/harm foundation was not a significant predictor. Hence, H2a was not supported.

The results of Regression model 3 demonstrate that the sanctity/degradation foundation was significantly related to negative beliefs regarding the ‘active, legal’ dimension. This supports H3c. As proposed, the authority/subversion foundation was not a significant predictor. This supports H3d. However, the care/harm foundation, fairness/cheating and loyalty/betrayal foundation were also not significant predictors. Hence, H3a, H3b and H3e were not supported.

Testing for the Relationships of the Various Moral Foundations with Beliefs Regarding ‘No Harm, No Foul’ Actions

A separate regression analysis was conducted with the various moral foundations as the independent variables and consumers’ beliefs regarding ‘no harm, no foul’ actions as the dependent variable (see Regression model 4 in Table 5).

Table 5 Regression analyses
Full size table

The results demonstrate that the sanctity/degradation foundation was significantly related to negative beliefs regarding ‘no harm, no foul’ actions. This supports H4c. As proposed, the care/harm foundation, authority/subversion foundation and loyalty/betrayal foundation were not significant predictors. This provides support for H4a, H4d and H4e. The fairness/cheating foundation was positively related to ‘no harm, no foul’ actions. This was not predicted, and hence, H4b was not supported.

Testing for the Relationships of the Various Moral Foundations with Beliefs Regarding ‘Doing Good’ Actions

A separate regression analysis was conducted with the various moral foundations as the independent variables and consumers’ beliefs regarding ‘doing good’ actions as the dependent variable (see Regression model 5 in Table 6).

Table 6 Regression analyses
Full size table

The results demonstrate that the fairness/cheating foundation and the authority/subversion foundation were related to positive beliefs regarding ‘doing good’ actions. This supports H5b and H5d. The care/harm foundation was marginally significantly (p = 0.060) related to positive beliefs regarding ‘doing good’ actions. This provides support for H5a. The loyalty/betrayal foundation was related to negative beliefs regarding ‘doing good’ actions. This supports H5e. However, the sanctity/degradation foundation was not a significant predictor. Hence, H5c was not supported.

Testing for the Mediation Effects of Various Moral Foundations in Regard to the Relationship Between Idealism and Beliefs Regarding Unethical Consumer Actions

A multiple mediation model (see Multiple mediation model 1 in Table 7) was utilised that uses bootstrapping (n = 1000) to generate bias-corrected confidence intervals to test for the indirect effects of idealism on beliefs regarding actively benefiting from illegal actions through the various moral foundations. Using Model 4 in the PROCESS macro in SPSS (Hayes 2013), indirect effects through mediation models were estimated by bootstrapping methods. The indirect effect is significant if the bootstrapped confidence interval does not include zero.

Table 7 Testing for the mediation effects of moral foundations in relation to the effects of idealism on unethical consumer actions
Full size table

The results indicated that the fairness/cheating foundation, sanctity/degradation foundation and loyalty/betrayal foundation mediate the effects of idealism on beliefs regarding the ‘active, illegal’ dimension. This supports H6b, H6c and H6e. However, the care/harm foundation and authority/subversion foundation are not mediators. Hence, H6a and H6d were not supported. After including moral foundations as mediators in the model, the direct effect of idealism on beliefs regarding the ‘active, illegal’ dimension was no longer significant.

A separate multiple mediation model (see Multiple mediation model 2 in Table 7) was utilised that uses bootstrapping (n = 1000) to generate bias-corrected confidence intervals to test for the indirect effects of idealism on beliefs regarding passively benefiting from the mistakes of the seller through the various moral foundations. The results indicated that the fairness/cheating foundation and the sanctity/degradation foundation mediate the effects of idealism on beliefs regarding the passive dimension while the authority/subversion foundation is not a mediator. This supports H7b, H7c and H7d. However, the care/harm foundation and loyalty/betrayal foundation also are not mediators. Hence, H7a and H7e were not supported. After including moral foundations as mediators, the direct effect of idealism on beliefs regarding the passive dimension was no longer significant.

A separate multiple mediation model (see Multiple mediation model 3 in Table 7) was utilised that uses bootstrapping (n = 1000) to generate bias-corrected confidence intervals to test for the indirect effects of idealism on beliefs regarding actively benefiting from legal but questionable actions through the various moral foundations. The results indicated that the sanctity/degradation foundation mediates the effects of idealism on beliefs regarding the ‘active, legal’ dimension, while the authority/subversion foundation is not a mediator. This supports H8c and H8d. However, the care/harm foundation, fairness/cheating foundation and loyalty/betrayal foundation also are not mediators. Hence, H8a, H8b and H8e were not supported. After including moral foundations as mediators in the model, the direct effect of idealism on beliefs regarding the ‘active, legal’ dimension was no longer significant.

Testing for the Mediation Effects of Various Moral Foundations in Regard to the Relationship Between Idealism and Beliefs Regarding ‘No Harm, No Foul’ Actions

A separate multiple mediation model (see Multiple mediation model 4 in Table 8) was utilised that uses bootstrapping (n = 1000) to generate bias-corrected confidence intervals to test for the indirect effects of idealism on consumers’ beliefs regarding ‘no harm, no foul’ actions through the various moral foundations.

Table 8 Testing for the mediation effects of moral foundations in relation to the effects of idealism on ‘no harm, no foul’ actions
Full size table

The results indicated that the sanctity/degradation foundation mediates the effects of idealism on beliefs regarding ‘no harm, no foul’ actions. This supports H9c. Furthermore, as proposed, the care/harm foundation, authority/subversion foundation and loyalty/betrayal foundation are not mediators. This supports H9a, H9d and H9e. However, as opposed to predictions, the fairness/cheating foundation mediates the effects of idealism on beliefs regarding ‘no harm, no foul’ actions. Hence, H9b was not supported. After including moral foundations as mediators in the model, the direct effect of idealism on beliefs regarding ‘no harm, no foul’ actions was no longer significant.

Testing for the Mediation Effects of Various Moral Foundations in Regard to the Relationship Between Idealism and Beliefs Regarding ‘Doing Good’ Actions

A separate multiple mediation model (see Multiple mediation model 5 in Table 9) was utilised that uses bootstrapping (n = 1000) to generate bias-corrected confidence intervals to test for the indirect effects of idealism on consumers’ beliefs regarding ‘doing good’ actions through the various moral foundations.

Table 9 Testing for the mediation effects of moral foundations in relation to the effects of idealism on ‘doing good’ actions
Full size table

The results indicated that the fairness/cheating foundation, authority/subversion foundation and loyalty/betrayal foundation mediate the effects of idealism on beliefs regarding ‘doing good’ actions. This supports H10b, H10d and H10e. The care/harm foundation was not a mediator. H10a was not supported. The lower limit of the 95% confidence interval for the sanctity/degradation foundation was −0.007 (very close to zero) indicating a negligible effect. Hence, H10c was also not supported. After including moral foundations as mediators, the lower limit of the 95% confidence interval of the direct effect of idealism on beliefs regarding ‘doing good’ actions was 0.002 (very close to zero) indicating a negligible direct effect.

Table 10 provides a summary of identified relationships among idealism, moral foundations and dimensions of consumer ethics.

Table 10 Identified relationships among idealism, moral foundations and consumer ethics
Full size table

Discussion

The key findings of this study are discussed below:

  1. 1.

    In relation to unethical consumer actions, consumers use perceptions of fairness as the basis of their beliefs regarding consumer ethics in the case of both purely illegal actions and passive acts of omission that are unfair to others, but not for consumer actions that are legal but questionable. The fairness/cheating foundation was also positively related to ‘doing good’ actions. These results support the view that concern for fairness, similar to Kohlberg (1969), is a motivator for ethical consumer actions, but is not a motivator to prevent all types of unethical consumer actions. An interesting finding was that the fairness/cheating foundation was positively related to perceptions of ‘no harm, no foul’ actions. One possible reason for this could be that these actions are not perceived to be unfair to others; hence, a fairness ethic would not necessarily have an issue with engaging in such actions.

  2. 2.

    The care/harm foundation was found not to be related to beliefs regarding unethical consumer actions. However, the care/harm foundation was related to positive beliefs regarding ‘doing good’ actions indicating that Gilligan’s (1982) ethics of care has a role in relation to altruistic consumer actions.

  3. 3.

    The sanctity/degradation foundation is negatively related to all forms of unethical consumer actions as well as ‘no harm, no foul’ actions. This indicates that in relation to unethical consumer actions, the sanctity/degradation foundation has the most wide ranging influence. The sanctity/degradation foundation is based on the moral emotion of disgust, and the key role of the sanctity/degradation foundation in relation to unethical consumer actions highlights the significance of emotions in ethical decision-making (Haidt 2001, Hardy 2006).

  4. 4.

    Hildreth et al. (2016) recently noted that in-group loyalty may be a virtue as it activates related moral traits, e.g. honesty, compassion. However, after controlling for the care/harm foundation (related to compassion) and the fairness/cheating foundation (related to honesty), it was found that the loyalty/betrayal foundation was positively related to beliefs regarding unethical actions and negatively related to beliefs regarding pro-social actions.

  5. 5.

    The authority/subversion foundation is not related to consumers’ ethical beliefs regarding unethical actions but is positively related to consumers’ ethical beliefs regarding ‘doing good’ actions. These results demonstrate that in the context of US consumers, respect for authority does not prevent the support for unethical consumer actions.

  6. 6.

    Collectively, the findings indicate that moral motivations for supporting pro-social actions as a consumer are not necessarily the same as moral motivations for condemning unethical actions. Furthermore, the results of this study demonstrate that moral foundations theory can explain the processes through which personal variables, e.g. idealism, influence consumer ethics. This illustrates that moral foundations theory can be utilised as a theoretical framework for explaining consumers’ ethical beliefs.

Implications

The key implications arising from this research are as follows:

First, as the sanctity/degradation foundation is the most significant in terms of generating negative beliefs towards unethical consumer actions, this implies that in order to motivate consumers to have negative perceptions of unethical consumer actions, communications based on the sanctity/degradation foundation would be the most effective. Public service messages emphasising that unethical consumer actions are against social norms leading to feelings of disgust would be effective in persuading consumers to reject unethical actions. Second, in order to motivate pro-social consumer actions, a number of different paths may be pursued. Consumers may be encouraged to support ‘doing good’ actions by highlighting the caring dimension and the fairness aspect of these actions. Appeals to respecting authority may also work in this context.

Third, the findings from this study (in relation to the loyalty/betrayal foundation) imply that rather than focusing on the benefits of their actions on immediate family and friends, consumers should be encouraged to contemplate the possible consequences of their actions on the greater community. This is in line with the view that ethical behaviour is based on expanding the circle of moral concern from immediate kin to outsiders (Singer 2011).

Fourth, the results demonstrate that simply relying on consumers’ respect for authority may not be adequate to encourage consumers to condemn unethical actions. A case in point is digital piracy, which has not abated in spite of authorities condemning this action (Jackman and Lorde 2014). Based on this research, consumers should be persuaded to avoid unethical actions through sanctity/degradation appeals that elicit moral emotions.

Limitations and Future Research

There are a few limitations of this study that provide insights into possible future research opportunities. The research did not take into consideration cultural differences or political differences. Liberals and conservatives rely upon different moral foundations (Haidt and Graham 2007), and hence, future research can examine how political differences moderate the impact of moral foundations on consumers’ ethical beliefs. Similarly, cultural differences may also affect the findings, e.g. in collectivist cultures respect for authority may have an important role.

The current research is cross section in nature. Hence, the relationships identified are correlational. Theory was developed that proposed directional relationships among the variables, and the findings support many of the proposed relationships. However, experimental research needs to be conducted that examines the effects of manipulating different moral foundations on consumers’ ethical beliefs. Future research should also examine whether moral foundations theory (Haidt and Graham 2007) can explain the effects of other personal variables (e.g. materialism, religiosity) that are related to consumer ethics.

Overall, this research demonstrates that different moral foundations drive consumers’ perceptions of various dimensions of consumer ethics. The findings show that consumption morality is more than about perceptions of fairness and harm. Expanding the domains of consumption morality to incorporate concerns for sanctity, loyalty and respect for authority can provide greater insights into the reasons behind consumers’ rejection of unethical actions (even if such actions are not illegal) as well as the reasons for consumers’ pro-social actions. Moral foundations theory (Haidt and Graham 2007) is a theoretical framework that can help develop a more in-depth understanding of how personal variables influence consumer ethics. Finally, as the results of this study demonstrate that various moral foundations explain the effects of idealism on beliefs regarding different dimensions of consumer ethics, it would be appropriate for future studies to focus on moral foundations rather than idealism as antecedents of consumers’ ethical beliefs.