Surveying the Moral Landscape

The New Explanatory Framework: Psychology’s Fundamental Motivational Distinction

Morality is generally recognized as a system of rules that facilitate and coordinate group living; as such it involves behavioral regulation so as to optimize our existence as social beings. In developing our taxonomy of morality, we turned to the most fundamental psychological distinction in the motivation and self-regulation of behavior—approach versus avoidance, or behavioral activation versus inhibition (Carver, 2006; Carver & Scheier, 1998; J. A. Gray, 1982, 1990; Higgins, 1997, 1998; for reviews, see Carver & Scheier, 2008, and Gable, Reis, & Elliot, 2003). This dual regulatory system is recognized across multiple areas of psychology, and in recent years this approach–avoidance distinction has proved instrumental in understanding diverse phenomena across psychology, including achievement (e.g., Elliot & Church, 1997), attention (e.g., Forster, Friedman, Ozelsel, & Denzler, 2006), power (e.g., Keltner, Gruenfeld, & Anderson, 2003), and interpersonal relationships (Gable & Strachman, 2008).

The behavioral inhibition system, based in avoidance, is an aversive motivational system that is sensitive to punishment and negative outcomes. The behavioral activation system, based in approach, is an appetitive motivational system that is sensitive to rewards and positive outcomes. We recently applied these differences to the moral domain and distinguished between proscriptive and prescriptive morality (Janoff-Bulman, Sheikh, & Hepp, 2009; also see Janoff-Bulman, 2011). More specifically, proscriptive morality (based in avoidance) focuses on what we should not do; it involves restraining a motivation to do something bad, and thus overcoming temptation or desire. Most generally, proscriptive morality protects from harm, and right conduct involves inhibition. Prescriptive morality (based in approach) focuses on what we should do; in contrast to inhibition and restraint, it requires overcoming inertia and establishing a motivation to do something good. Most generally, prescriptive morality provides for well-being, and right conduct involves activation and engaging in helpful behaviors (Janoff-Bulman et al., 2009). Just as a parent strives to protect the child from harm and provide for the child’s basic needs, the motivations to protect and provide are fundamental moral motives that broadly reflect a dual system of moral regulation (i.e., proscriptive vs. prescriptive regulation, respectively).

Research on the development of morality in young children by Kochanska and colleagues (Aksan & Kochanska, 2005; Kochanska, 2002; Kochanska, Coy, & Murray, 2001) supports this motivational distinction in moral regulation. These researchers distinguished between “do’s” and “don’ts,” with the former involving activation and sustaining an activity (e.g., toy cleanups) and the latter involving prohibitions and suppressing behaviors (e.g., not playing with a forbidden attractive toy). Kochanska and colleagues (Aksan & Kochanska, 2005; Kochanska, 2002; Kochanska et al., 2001) found that young children (i.e., 14, 22, 33, and 45 months) complied with prohibitions (don’ts) more readily than with do’s, which were more challenging. Furthermore, fearfulness was positively associated with children’s success at refraining from prohibited behaviors (don’ts), but not with success in the domain of do’s (which is consistent with the greater sensitivity of the proscriptive system to negative outcomes). Based on their findings, Kochanska et al. (2001) concluded that there is “impressive evidence of substantial differences” between do’s and don’ts in early self-regulation (p. 1106).

Our own recent research provides further support for “substantial differences” between proscriptive and prescriptive morality, or avoidance/inhibition and approach/activation in the moral domain. Janoff-Bulman et al. (2009; seven studies are reported) found that proscriptive morality is focused on transgressions and is responsive to threat; it is mandatory and emphasizes blameworthiness. In contrast, prescriptive morality is focused on good deeds and is not responsive to threat; it is more discretionary and emphasizes creditworthiness. Linguistically, proscriptive morality is represented in relatively concrete terms, particularly verbs that specify particular behaviors; in contrast, prescriptive morality tends to be more abstract linguistically, in adjectives reflecting broad categories of behavior (Janoff-Bulman et al., 2009, Study 2). These linguistic differences represent regulatory differences based on approach and avoidance (e.g., see Semin, Higgins, de Montes, Estourget, & Valencia, 2005). The greater threat of error in the proscriptive system requires greater specificity of actions to be inhibited, whereas the more discretionary prescriptive system can provide more general directives and standards for behavior. ¹

Overall, proscriptive moral regulation is condemnatory and strict, whereas prescriptive morality is commendatory and less strict. ² Thus, the negativity bias in psychology (for reviews, see Baumeister, Brataslavsky, Finkenauer, & Vohs, 2001; Rozin & Royzman, 2001; Vaish, Grossmann, & Woodward, 2008), which entails greater motivational potency of negative (vs. positive) outcomes, is evident in the moral domain as well. Proscriptive regulation is harsher and more demanding than prescriptive regulation.

These differences suggest natural affinities to deontological versus consequentialist perspectives on morality. Deontology is predicated on rules of right and wrong and is a strict, mandatory morality—one has a duty to behave in particular ways. Consequentialism posits that the morality of an act is based on its consequences. This is essentially utilitarianism, which argues for the greatest good for the greatest number. The more discretionary nature of prescriptive morality may be inherent in the nature of prescriptions; we know we should help others in need, but it is impossible to help everyone who is in need. We can, however, always refrain from cheating or killing. “Not harming” can be applied universally, but “helping” cannot. The very nature of prescriptions (e.g., “help others”) therefore requires choices and decisions about whom to help (and perhaps how), and standards such as “the greatest good for the greatest number” provide guidance.

Interestingly, the harsher, more mandatory nature of proscriptive (vs. prescriptive) morality can account for the findings in the classic footbridge and trolley problems in moral psychology, which essentially pit proscriptive and prescriptive morality against one another. Five lives can be saved (prescriptive morality) by forfeiting the life of one (proscriptive immorality); or restraint from harming the one (proscriptive morality) will sacrifice the lives of five (prescriptive immorality). Given the stricter, more mandatory proscriptive system, it is not surprising that when forfeiting a life involves direct physical harm (i.e., pushing a stranger off the footbridge), restraint carries the day and five lives are sacrificed. To reverse this pattern, the harm must be “defanged” by becoming distal and impersonal. Only then can prescriptive regulation, with its reliance on consequences, take over; then saving a total of four lives seems like the moral path to follow.

Differences between proscriptive and prescriptive morality may also help us better understand findings by Knobe (2005, 2010), in which negative side effects in moral dilemmas are perceived as more intentional than positive side effects. The more discretionary nature of helping requires evidence of specific intention on the part of the helper to get credit for positive outcomes; in the scenarios this intention is not present, and it is not accorded by research participants. However, the mandatory nature of avoiding harm-doing suggests that people are more obligated to intervene and stop negative, foreseen consequences. The Knobe dilemmas that involve instances of failing to refrain from harm naturally engage an attribution of intentionality, because we are required to intervene to “not harm.”

Model of Moral Motives (MMM): A New Taxonomy

Given the fundamental nature of the activation-approach and inhibition-avoidance systems across psychology and the distinct attributes associated with proscriptive versus prescriptive morality, we believe it is an important dimension to consider in creating a comprehensive model of the moral landscape. The two forms of moral regulation, and in particular their representation in the broad moral motives to protect (via inhibition) and provide (via activation), define the two rows of our model (see Figure 1).

OPEN IN VIEWER

Figure 1.

Model of moral motives.

Each of the three columns of our model represents a distinct focus of moral concern, from self (personal) to other(s) (interpersonal), to the group (collective). This tripartite classification represents different levels of analysis in social psychology. The self-focus involves the impact of individuals’ behavior on themselves. The interpersonal column refers to moral concerns focused on another individual or individuals; and the group category involves a focus on the collective (one’s group) as a whole. Social-psychological phenomena have been categorized as intrapersonal, interpersonal, and collective (Gilbert, Fiske, & Gardner, 1998). And the considerable work on social identity by Brewer and colleagues (Brewer & Gardner, 1996; Sedikides & Brewer, 2001) relies on the same three categories; that is, they distinguish between the individual self, the relational self, and the collective self as three fundamental ways people establish self-representations—in terms of their own unique traits, their dyadic relationships, and their group memberships (also see Greenwald & Breckler, 1985). Morality, too, can operate at each of these levels; moral motives can be directed toward the self, another person or persons, or one’s group. Our model therefore has three columns, each devoted to a particular focus of moral concern. ³

Comparison With Haidt’s Moral Foundations

Currently, the dominant map of the moral domain is the moral foundations theory of Haidt and colleagues, which posits five psychological foundations: Harm/Care, Fairness/Reciprocity, Ingroup/Loyalty, Authority/Respect, and Purity/Sanctity (Haidt, 2007, 2008; Haidt & Graham, 2007; Haidt & Joseph, 2004, 2007). Haidt and Joseph (2004) were explicit in acknowledging that they initially derived these five psychological foundations from their reading of five works ⁶ (also see Haidt & Kesebir, 2010; Haidt & Graham, 2007), and these same five modules constitute Haidt’s moral foundations in his subsequent articles. ⁷ In terms of Shweder and colleagues’ (Shweder, Much, Mahapatra, & Park, 1997), earlier “three ethics” account of morality, Harm/Care and Fairness/Reciprocity roughly correspond to the ethic of autonomy, Ingroup/Loyalty and Authority/Respect correspond to the ethic of community, and Purity/Sanctity corresponds to the ethic of divinity.

Haidt and colleagues refer to Harm/Care and Fairness/Reciprocity as the individualizing foundations because they involve “individual-focused contractual approaches to society” (Graham et al., 2011, p. 369). In contrast, Ingroup/Loyalty, Authority/Respect, and Purity/Sanctity are regarded as the binding foundations, because they “are about binding people together into larger groups and institutions” (Graham et al., 2011, p. 369). The individualizing foundations focus on the rights and welfare of individuals, whereas the group is the locus of moral concern for the binding foundations.

The “interpersonal” column in our model parallels Haidt’s individualizing classification; that is, it focuses on other individuals as the locus of moral concern. From the perspective of our MMM, Harm/Care and Fairness/Reciprocity fall in this interpersonal column, Harm in the “Not Harming” (proscriptive/protect) cell and Care and Fairness/Reciprocity in the Helping/Fairness (prescriptive/provide) cell; here, there is virtually complete correspondence between the two perspectives in terms of content, although we classify that content slightly differently from Haidt based on our addition of approach/avoidance motivation to the MMM framework. We have placed the foundations in cells that we believe reflect the self-regulatory motives underlying the modules.

We believe Haidt and colleagues’ binding foundations, with their focus on the group, have a natural home in the “group” column of our model. Placing the binding foundations in the “self” or “other” columns of our model would mischaracterize them as being focused on something other than the group as the locus of moral concern. Loyalty, Authority, and Purity reflect concerns with order and solidarity, and we therefore locate them in the Social Order/Communal Solidarity (proscriptive) group cell. Our group column includes two distinct types of binding moralities, but several modules in Haidt’s model are reflected in a single group-based morality cell of MMM (see Figure 2).

OPEN IN VIEWER

Figure 2.

Haidt’s moral foundations and the model of moral motives.

Given Haidt’s own label of purity as a binding morality, we include it in the group solidarity cell as well (also see Graham & Haidt, 2010, on the role of purity practices in religion). However, it should be noted that purity involves proscriptive morality more broadly in terms of inhibition or avoidance of “dangerous” stimuli, and can be represented in all three proscriptive cells of our model (i.e., self, other, and group foci of moral concerns). Thus, purity regarding one’s own body (e.g., matters related to sexuality, drinking, or drugs) concerns the self, the desire to avoid contagion associated with another would involve the interpersonal context, and the motivation to maintain group purity through the regulation of deviance (via avoidance and punishment) implicates group-based morality. In many ways, given the distinct origins of the purity foundation of morality, ⁸ from our perspective it can be regarded as a broad-based proscriptive form of morality that coexists with the other proscriptive moral motives.

Microjustice Versus Macrojustice: Focus on the Individual Versus the Group

We argue that there is a group-focused prescriptive (provide) morality and claim that it is based in a Social Justice orientation derived from a sense of shared communal responsibility. From our perspective, Haidt concludes that conservatives rely on both the individualizing and binding foundations, whereas liberals rely only on the individualizing foundations because he has limited his binding foundations to those reflecting a constraint-based (proscriptive) group morality. Interestingly, some of his own research findings, regarding gender in particular, should raise a few red flags about his conclusions. Thus, there is a large literature in psychology addressing gender differences, and one difference that is well acknowledged is the communal versus agentic orientations associated with females versus males, respectively (see, for example, Eagly, 1987; Wood & Eagly, 2010). Although women are group oriented, focused on communal binding, Graham et al. (2011) nevertheless find that men score higher than women on their important binding foundations of Ingroup/Loyalty and Authority/Respect. If these are the communal-based foundations, one would expect women to be stronger or at least equal advocates of these moralities. Surely there could be some binding, group-based morality that reflects a type of communal orientation currently absent from the moral foundations model of morality.

In positing that communal responsibility, and in particular Social Justice, is at the core of this group-based morality, we anticipate an immediate response and critique suggesting that Social Justice is really a form of fairness. In other words, liberals rely on fairness and simply apply this “individualizing” foundation at the group level. While on its face this seems like a valid criticism, we believe that group-based Social Justice is not the same as the individualizing foundation, interpersonal fairness, applied at the group level. We may have a paucity of language and thus necessarily link the constructs through fairness and justice, but these are actually quite different phenomena and moral concerns.

More specifically, interpersonal fairness is based on considerations of another’s deservingness, whereas Social Justice is based on group-level considerations that are focused on the overall distribution of resources in a group. The former is focused on specific individuals or group members, whereas the latter is focused on the group as a whole. These are the differences between microjustice and macrojustice described in a classic article by Brickman, Folger, Goode, and Schul (1981) and recognized as “qualitatively different principles” (p. 174). Interpersonal fairness is individuating and requires the assessment of individuals’ attributes (e.g., need, merit). In contrast, Social Justice is deindividuating and group-focused, and specifies a priori constraints on the pattern or form of a distribution. Interpersonal fairness is based on the particular deservingness of another (or others), whereas Social Justice is based on societal-level considerations focused on minimizing inequality across the distribution. Specific knowledge about individuals’ attributes is superfluous to Social Justice concerns; instead they entail “specifying a desirable property of a distribution rather than an empirical property of a set of individuals” (Brickman et al., 1981, p. 179). Thus, Rawls (2005) in his theory of justice first proposes a veil of ignorance to deindividuate people, and only then does his second principle act as a distributional rule for the group: “Social and economic inequalities are to be arranged so that they are . . . to the greatest benefit of the least advantaged.” (p. 266). Rawls sees this unambiguously as “to everyone’s advantage” (p. 53)—that is, to the advantage of the group as a whole. The group is the primary focus of moral concern.

Rai and Fiske’s (2011) Proportionality motive, which involves fairness calibrations associated with merit, is an example of a microjustice motive, and thus is a (prescriptive) interpersonal motive in our taxonomy and an individualizing motive in Haidt’s moral foundations model. However, Equality is more ambiguous in the Rai and Fiske conceptualization. Equality is defined as the motive for in-kind reciprocity and equal treatment; the former maps onto our Helping/Fairness motive, which resides in the interpersonal domain, whereas “equal treatment” can be group-focused, indicating an interest in greater equality across a distribution. Thus, Rai and Fiske’s Equality motive could potentially operate at the interpersonal and group levels, although they do not make these distinctions; these would reflect microjustice concerns with fairness and reciprocity at the individual level and macrojustice concerns at the group level.

That interpersonal fairness and Social Justice reflect two distinct moral concerns becomes evident as we consider cases in which they clearly clash. An example is affirmative action. Interpersonal fairness and group-based Social Justice are often at odds here, and different outcomes are advocated depending upon whether one uses interpersonal fairness (i.e., the individual’s merit is what matters) or Social Justice (i.e., distribution of resources—jobs, education, etc.—should be more equally distributed across society) as the basis for one’s position (see, for example, Son Hing, Bobocel, & Zanna, 2002). The moral disagreement here is about the focus of moral concern and whether it should be on the individual or the group as a whole. Similarly, faculty members are likely to be familiar with the two different moral concerns when grading exams. Students receive a grade based on their individual performance, but then the test may be curved to reflect a “fairer” distribution across the class (also see Brickman et al., 1981, for other examples). Interestingly, Bar and Zussman (2012) found that the grade distributions of professors identified as Democrats are more egalitarian, with professors identified as Republicans assigning more very low and very high grades.

Individualizing fairness operates between individual members of groups; in contrast, a group-based morality based on Social Justice is not about interpersonal fairness, but rather concerns about the group as a whole. As Jost and Kay (2010) note, “social justice is a property of social systems” (p. 1122); it is a group-level rather than individually oriented moral concern. The focus is the well-being and integrity of the group. ¹⁰ Haidt notes that these are served by adherence to the binding foundations of Loyalty, Authority, and Purity, and a liberal moral orientation, focused on his individualizing foundations of Care and Fairness, threatens “e pluribus unum.” Interestingly, however, there is increasing evidence that the integrity and well-being of a group are served by a binding morality based on Social Justice and egalitarianism. Thus, researchers have found that societal equality is associated with social harmony (see Kesebir, 2011). More specifically, inequality is associated with low social capital, low social trust, lower community participation, and higher levels of violent crime (e.g., Hsieh & Pugh, 1993; Kawachi, Kennedy, Lochner, & Prothrow-Stith, 1997; Uslaner, 2002), and societal equality is associated with greater happiness and health (e.g., Deaton, 2003; Oishi, Kesebir, & Diener, 2011).

Furthermore, recent research based on more than 200 data sets from sources such as the World Bank, the United Nations, the World Health Organization, and the U.S. Census found that among 23 of the world’s wealthiest countries and individual U.S. states, those with the greatest equality (the smallest wealth differentials between rich and poor) are also those that have the highest quality of life for their citizens as a whole and not simply their poorest citizens (Wilkinson & Pickett, 2009). On virtually every index of well-being, deprivation, and quality of life, these researchers found a strong correlation between a country’s economic level of equality and its social outcomes. Inequality has a pernicious effect on a society; it unravels the social ties that bind members of a healthy society. Paradoxically, given Haidt’s perspective, Noah (2012) specifically notes that maintaining the spirit of e pluribus unum is particularly difficult in the context of growing income inequality (also see Stiglitz, 2012). There appears to be more than one moral route to binding a group—one focused on Social Order and the other on Social Justice.

Two Group-Based Moralities

We propose that there are two distinct group-based moralities, both communal in orientation, but reflecting different moral motives. In the case of Social Order and Social Justice moral motives, the integrity of the group is primary, but how to reach this goal differs considerably. The emphasis on order versus justice reflects fundamentally different concerns associated with protecting the group (proscriptive morality) versus providing for it (prescriptive morality), but both Social Order and Social Justice bind people into meaningful groups.

Perhaps it is not surprising that Haidt has focused on a single proscriptive group-based morality, because he strongly emphasizes human selfishness and self-interest in his work. In particular, he notes that the role of moral systems is to constrain selfishness (Haidt, 2007, 2008, 2012). Thus, he and his coauthors write of the individualizing and binding approaches to morality:

Haidt described two common types of moral systems—two ways of suppressing selfishness . . . Some cultures try to suppress selfishness by protecting individuals directly . . . Other cultures try to suppress selfishness by strengthening groups and institutions and by binding individuals into roles and duties in order to constrain their imperfect natures. (Graham et al., 2009, p. 1030)

Furthermore, in his recent book, despite his acknowledgment of group selection, Haidt (2012) explicitly argues for a Glauconian perspective, one based on recognizing the primary role of individual self-interest and the desire to appear good rather than be good; that is, we are selfless and caring when our reputations are at stake. ¹¹ His focus on a single restrictive, proscriptive group morality follows naturally from his emphasis on human selfishness and our “imperfect natures” (see Sowell, 2007, who would regard this as the constrained, vs. more positive and unconstrained, vision of human nature). In our own model we do not deny the selfish side of human nature but instead acknowledge the altruistic side as well; the former is regulated via proscriptive morality, and the latter via prescriptive morality.

Haidt (2008) views “morality as the glue that binds”; crediting Durkheim, he notes that “morality constrains individuals and ties them to each other to create groups” (Haidt, 2007, p. 1000). Durkheim (1893/1984) proposed that morality holds societies together, forming group entities with emergent properties. Although Durkheim emphasized the constraining aspects of moral groups, he nevertheless identified two ways in which a society could be “glued” together: mechanical solidarity, which is characterized by homogeneity, and organic solidarity, which is typified by interdependence. Recent work on entitativity, the degree to which a group is an entity, roughly resembles Durkheim’s dual conception in that there appear to be two qualitatively different ways of achieving this end (e.g., Brewer, Hong, & Li, 2004; Haslam, Rothschild, & Ernst, 2000, 2002; Ip, Chiu, & Wan, 2006; Rutchick, Hamilton, & Sack, 2008; Wilder & Simon, 1998).

Entitativity “is the glue that holds (or is perceived as holding) a group together” (D. L. Hamilton, 2007, p. 1087). Essence theories of groups are based on perceived common attributes and are characterized by ingroup homogeneity and impermeable group boundaries, because all group members share some essential quality. In contrast, agency theories of groups are based on common goals and purpose, and are characterized by ingroup heterogeneity and permeable group boundaries (see Brewer et al., 2004). Common group identity underlies essence theories of group entitativity, whereas group interdependence around shared goals underlies agency theories. Empirical work has found that either approach can account for group entitativity; more specifically, reflecting differences between essence and agency theories of group, respectively, researchers have found that a category construal (emphasizing category membership) or a dynamic construal (emphasizing interacting members of a group) increases perceived entitativity (e.g., Rutchick et al., 2008). There are different ways of binding people into groups, and these correspond to the two group-based moralities in our model.

Social Order is consistent with an essence theory of groups. Group features such as hierarchy and homogeneity, and indices such as loyalty and conformity, all contribute to an increased sense of order and security. For purposes of protection and avoiding harm, a major concern of this proscriptive group-based morality is identifying who is in and who is not in the group. As Graham et al. (2009) note, binding qualities of loyalty and patriotism are combined with an “extreme vigilance for traitors” (p. 103). Group conformity reflects the desire for homogeneity, and the importance of common social identity (and markers of this identity) underlies a restrictive and relatively impermeable intergroup boundary. Engaging in particular behaviors, and more importantly, inhibiting particular behaviors and lifestyles regarded as “deviant” by the group or society (see Blanton & Christie, 2003, on deviance regulation in negative incentive systems), is of crucial importance in Social Order morality.

In contrast, Social Justice is consistent with an agency theory of groups. A major concern of Social Justice-based morality is greater equality in the distribution of social resources and public goods, and thus the emphasis here is on interdependence and common goals rather than identity. The just distribution of public goods and social resources across group members is evidence of societal sharing and communal responsibility; here, the focus is on social welfare and economics, and the emphasis is on the shape of the distribution in providing public goods. Interdependence rather than common identity allows for group heterogeneity and a relatively nonrestrictive, permeable intergroup boundary.

Not surprisingly, then, the two group-based moralities are concerned with distinct social issues. Research on contemporary issues finds that they factor into two clusters (Lewis-Beck, Jacoby, Norpoth, & Weisberg, 2008; also see Janoff-Bulman, Sheikh, & Baldacci, 2008). One cluster, a “Public Goods” factor, includes issues related to economic resources and opportunities such as government welfare and affirmative action, and research has found that this factor is associated with Social Justice, but not Social Order; the other cluster, a “Lifestyles” factor, includes more “personal” issues such as gay marriage and abortion, and is associated with Social Order but not Social Justice (Janoff-Bulman et al., 2008). Group-based moralities involve the regulation of behavior for all members of the group, and thus entail not simply self-regulation, but more broadly, social regulation.

Implications of the Group-Based Moralities: Intergroup Aggression

Perhaps not surprisingly, the two group-based moralities are associated with different emphases on intergroup boundaries. Social Order morality highlights the importance of allegiance and loyalty; these are in the service of protecting group members and facilitating determinations of ingroup (vs. outgroup) membership. The boundary separating the ingroup from the outgroup is likely to be thick and relatively impermeable, relatively restrictive rather than open, thereby minimizing threat and danger. For Social Justice morality, the emphasis is on intragroup differences (i.e., minimizing inequality) rather than intergroup differences. Intergroup boundaries are likely to be far more permeable and less restrictive, and similarly, loyalty and allegiance are not apt to be emphasized.

These distinct orientations have important implications for the treatment of outgroups. Although ingroup loyalty may function to bind groups, research by Cohen, Montoya, and Insko (2006) alerts us to its dangers as well. These researchers examined data from the Standard Cross-Cultural Sample, a database of primarily preindustrial societies chosen from across the globe to be maximally cultural independent (Murdock & White, 1969). Cohen et al. (2006) found that greater loyalty to the ingroup was associated with greater aggression and violence directed toward outgroups, as well as greater valuing and enjoyment of war. Outgroup-directed moralities are implied cells in the MMM, as shown in Figure 3. We would expect that loyalty that manifests as patriotism, particularly blind patriotism and nationalism, would be especially dangerous vis-à-vis outgroup-directed violence. It represents a darker side of Social Order-based morality (Kosterman & Feshbach, 1989; Schatz, Staub, & Lavine, 1999). Similarly, recent work on ingroup glorification versus attachment (Roccas, Klar, & Liviatan, 2006) suggests the dangers of glorification for outgroup members as well. Thus, Leidner, Castano, Zaiser, and Giner-Sorolla (2010) found that ingroup glorification, but not attachment, was associated with resistance to justice and moral disengagement for ingroup-committed atrocities of outgroup members. Although attachment to the ingroup may reflect a fundamental need of individuals, this work suggests that it is not attachment that matters for outgroup-focused atrocities, but rather a form of loyalty and allegiance based on comparison with other groups. Subjective identification with the group (i.e., attachment) is part and parcel of belonging to groups, but glorification entails feelings of ingroup superiority and emphasizes obedience and loyalty to the ingroup (Leidner et al., 2010). Again, these attributes are underscored in morality-based groups focused on Social Order rather than Social Justice.

OPEN IN VIEWER

Figure 3.

Intergroup implications of the model of moral motives.

Of course there may well be vulnerabilities associated with the latter group-based morality as well; thus, an argument from the Right is that liberals are likely to underestimate true dangers and real external threats. This suggests a need to more closely consider the relationship between the two group-based moralities and political ideology. Group-based moralities are focused on the group as a whole; although they implicate an individual’s behaviors, they importantly involve the regulation of behavior for members of the group, and thus entail not simply self-regulation, but more broadly, social regulation, which is the domain of politics.

Political Ideologies: Reconsidering Individualizing and Binding Moralities

Based on the MMM, Haidt’s moral foundations theory supports a false dichotomy that pits individualizing liberals against conservatives who evidence individualizing and binding moralities. In our taxonomy, liberals and conservatives are individualizing and binding in orientation; they engage in interpersonal (individualizing) moral actions as well as group-based morality. Not surprisingly, the two group-based moralities are associated with different political ideologies, with Social Order associated with political conservatism and Social Justice with political liberalism (Janoff-Bulman, 2009; Janoff-Bulman et al., 2008; also see, for example, Bobbio, 1996; Jost, Glaser, Kruglanski, & Sulloway, 2003; Jost, Nosek, & Gosling, 2008; Kerlinger, 1984; Kluegel & Smith, 1986; Skitka, 1999; Skitka, Mullen, Griffin, Hutchinson, & Chamberlin, 2002). Haidt and colleagues recognize a proscriptive group-based morality reflecting conservatives’ communal concerns; however, once a prescriptive group-based morality is acknowledged, liberals’ communal concerns become evident as well. Interestingly, it is libertarians, and not liberals, who lack a group-based morality. Libertarians are low on both Social Order and Social Justice, whereas, communitarians are high on both.

That liberals and conservatives have communal concerns is supported by data we collected using Triandis and Gelfand’s (1998) Individualism–Collectivism Scale. Paralleling the two group-based moralities, Triandis (2001) distinguishes between horizontal and vertical collectivism; individuals high on horizontal collectivism emphasize equality, sharing, and cooperation, whereas those high on vertical collectivism emphasize hierarchy and readily submit to authority. Based on the responses of 1,015 respondents, we found that that liberalism–conservatism was not associated with overall collectivism scores (B = .057, p = .471), but was associated with the horizontal–vertical dimension of the scale. More specifically, greater conservatism (i.e., decreased liberalism) was positively associated with vertical collectivism (B = .333, p < .001) and negatively associated with horizontal collectivism (B = −.254, p < .001); liberalism was thus positively associated with horizontal collectivism and negatively associated with vertical collectivism. The differences between horizontal and vertical collectivism reflect differences (i.e., equality vs. hierarchy) that are associated with Social Justice and Social Order, respectively.

As noted above, Social Order and Social Justice moral motives focus on different factors in the interest of satisfying goals that ultimately serve fundamental motives to protect versus provide for group members. Social Justice morality, focused on communal responsibility and the distribution of group resources, seeks to regulate public resources, which largely represent the economic domain. In contrast, Social Order seeks regulation in the domain of lifestyles and personal behaviors, which essentially serve as social identity markers of group belonging. ¹² When we consider these orientations politically, we see that we are dealing with powerful Left–Right differences in terms of government involvement versus autonomy. We typically think of conservatives as those who want limited government and a strong emphasis on personal liberty, and liberals as those who want a stronger government, with a more limited emphasis on personal liberty. However, these are surely false distinctions, because the domain of regulation is of paramount importance. That is, liberals and conservatives want strong government regulation (i.e., social regulation), but in different domains; and they both want personal liberty and autonomy—no regulation—but again in different domains.

Returning to their distinct social regulatory foci, conservatives want strong government regulation and involvement when it comes to personal behaviors (often involving the body) and lifestyles, but want autonomy when it comes to economics and public goods. In contrast, liberals want autonomy—government hands-off—when it comes to personal behaviors (often involving the body) and lifestyles but want strong government regulation of economics and the distribution of public goods and resources. Interestingly, supporting their protection priorities, conservatives also want strong government involvement when it comes to national security and defense, as well as more local security matters such as the size of police forces.

In the domains in which each side prefers regulation, the protect–provide differences between liberalism and conservatism are often manifested in the nature of the desired regulation, particularly with regard to inhibition (proscriptive) and activation (prescriptive). Thus, in recent culture wars, conservatives have emphasized the importance of inhibition (e.g., restriction of abortion, gay marriage, stem cell research), and liberals have emphasized the importance of activation in the establishment of government programs (e.g., welfare, universal health care, affirmative action program).

Moral Regulation, Politics, and the Uniqueness of Group-Based Moralities

To more fully appreciate the nature of group-based moralities, we return to the full six-cell MMM, with its rows defined by protect/proscriptive regulation and provide/prescriptive regulation and its columns defined by the focus of moral concern (i.e., self, other, group). The two forms of moral regulation are derived from work on self-regulation that distinguishes between a behavioral approach/activation system and a behavioral avoidance/inhibition system. Just as highly adaptive organisms rely on approach and avoidance in responding to their environment, we generally operate on the basis of proscriptive and prescriptive regulation in our daily lives. We try to avoid transgressions, inhibit temptations, engage in good deeds, and lead moral lives. Surely we are not always successful, sometimes failing in self-restraint, other times failing to provide needed help. However, there is nothing antagonistic about the proscriptive and prescriptive systems when dealing with moral regulation of our own behavior (i.e., self-regulation), whether targeting ourselves or others. People who try to avoid harming others are likely to want to help as well, at least some of the time, and vice versa. Similarly, the Self-Restraint/Moderation motive is positively rather than negatively correlated with Industriousness (Janoff-Bulman et al., 2008). Although people may have a dominant perspective or greater sensitivity to one or the other form of self-regulation, proscriptive and prescriptive regulation in the first two columns of the model, referring to morality directed at the self and another, are wholly compatible. A moral life comprises both types of moral regulation.

However, the comfortable relationship between the two forms of moral regulation seems to break down when we move to the group (i.e., third column). Here, we are dealing with social regulation rather than solely self-regulation; although this social regulation implicates individual behavior, it is really about moral rules and standards applied by and for the group. And here proscriptive and prescriptive regulation take distinct forms that seem more antagonistic than compatible. Advocacy of Social Order and Social Justice are negatively correlated (e.g., Janoff-Bulman et al., 2008). This group context is the realm of politics, and as discussed above, Social Order and Social Justice reflect different political orientations. It is intriguing to consider why the two forms of moral regulation (i.e., proscriptive and prescriptive regulation), so compatible at the individual and interpersonal level, become so conflictual at the group level. The two group-based moralities become the battleground for distinct moral orientations focused on protecting versus providing, societal losses versus gains, and stability (inhibition) versus change (activation).

Group-based morality is about societal rules and standards—social regulation rather than self-regulation regarding own behavior toward the self and others. Social regulation implicates one’s own behaviors but is focused on the collective and communal obligations or prohibitions.

For the self and interpersonal contexts, prescriptive–proscriptive balance resides within the individual. That is, the (im)moral actor is the unit of analysis, and here the two regulatory orientations comfortably coexist, called upon as the person and situation demand, constraining our selfishness and enabling our altruistic impulses. However, the relevant unit of analysis for group-based moralities is not the individual, but rather the group or society as a whole. And just as proscriptive and prescriptive morality are relatively balanced in a healthy individual, the two moralities are also relatively balanced at the societal (group) level, with both well-represented across society, as reflected in political ideologies.

When exploring these differences through political liberals and conservatives, we can see the divergence in moral regulation as we move from self-regulation to social regulation. As discussed below, there is increasing evidence that political liberals and conservatives differ in their sensitivities to positive and negative outcomes generally, and to prescriptive and proscriptive morality more specifically; prescriptive regulation is more likely to be dominant for liberals, and proscriptive regulation is more likely to be dominant for conservatives. These differences are minimally apparent when the moral focus is on the self or another, but are extremely apparent when the moral focus is the group.

Janoff-Bulman (2009) suggested that conservatism is broadly based in avoidance motivation, and liberalism in approach motivation. It is not simply that societies become more conservative in times of salient dangers and insecurity, which is the case (e.g., Bonanno & Jost, 2006; Janoff-Bulman & Sheikh, 2006; Landau et al., 2004; McCann, 1997), but that liberals and conservatives themselves differ in their sensitivities to positive and negative outcomes, gains, and losses. These links—between conservatism and avoidance motivation and between liberalism and approach motivation—have been demonstrated in a number of quite distinct studies. For example, participants in a study by Shook and Fazio (2009) played a computer game that involved learning whether novel stimuli produced positive or negative outcomes. Conservative participants learned negative stimuli better than positive; furthermore, they used a more cautious, avoidant strategy, whereas liberals used a more open, approach strategy to novel stimuli. Recent work that manipulated message framing demonstrated that conservatives showed greater sensitivity to losses than gains, whereas liberals were more sensitive to gains than losses (Lavine et al., 1999). Furthermore, Oxley et al. (2008) found that conservatives show higher physiological reactivity to threat (i.e., sudden noises and threatening visual images) than liberals. And in their 20-year longitudinal study of personality, Block and Block (2006) found that 23-year-old conservatives were likely to have been described as fearful and overcontrolled preschoolers two decades earlier, whereas 23-year-old liberals were likely to have been described as resilient and energetic as preschoolers. Similarly, Fraley, Griffin, Belsky, and Roisman (2012) found that fearfulness in childhood was associated with conservatism in late adolescence, whereas high activity levels in childhood were associated with liberalism in late adolescence. In general, conservatives are higher on threat sensitivity (Jost et al., 2003), which is associated with inhibition-based avoidance motivation, whereas liberals are higher on Openness to Experience (McCrae, 1996), which is linked to activation-based approach motivation.

In addition, recent research on purity and disgust has found that the tendency to feel disgust (“disgust sensitivity”) is associated with political conservatism (Inbar, Pizarro, & Bloom, 2009). This relationship has been confirmed in voting behavior and in international samples (Inbar, Pizarro, Iyer, & Haidt, 2011). Disgust is strongly linked to avoidance motivation (no doubt originally to pathogens and rotten foods) and is particularly apparent regarding issues such as gay marriage and abortion, which are related to conservatives’ purity perceptions and regarded as “social dangers” (Inbar et al., 2009).

Together these findings provide converging support for different motivational orientations of liberals and conservatives, with liberals emphasizing approach/activation and conservatives emphasizing avoidance/inhibition. Not surprisingly, in the moral domain these differences translate into differences in emphasis on proscriptive versus prescriptive morality. In recent research that included a measure of proscriptive–prescriptive regulation (i.e., Moralisms Scale, Janoff-Bulman et al., 2009), we found greater conservatism positively correlated with proscriptive morality and negatively correlated with prescriptive morality (and thus liberalism, the opposite end of the scale, was positively correlated with prescriptive morality and negatively correlated with proscriptive morality). In addition, differential sensitivity to proscriptive morality was demonstrated in research that primed moral regulation and explored cognitive categorization (Rock & Janoff-Bulman, 2010). Using categorization of neutral stimuli (e.g., whether “camel” is a member of the category “vehicle”) based on the work of Rosch (1975), when primed with proscriptions, conservatives (but not liberals) used narrower categories; that is, they did not include less prototypical items in the category, thereby raising the bar for inclusion. Liberals were not sensitive to proscriptions but were nonrestrictive regardless of prime.

Recent research on parenting and politics further illustrates these proscriptive–prescriptive differences. In a longitudinal investigation of the antecedents of political ideology, Fraley et al. (2012) found that parents’ authoritarian attitudes (assessed when the child was 1 month old) predicted children’s conservatism 18 years later, whereas parents’ egalitarian attitudes predicted children’s political liberalism. In our own research with parents and their college-aged children, we also found an association between restrictive parenting and greater political conservatism (Janoff-Bulman, Carnes, & Sheikh, 2013). McAdams and colleagues (McAdams et al., 2008) used a very different methodology and explored life narratives of political liberals and conservatives. These researchers found that conservatives emphasized authority figures who were strict enforcers of moral rules, whereas liberals emphasized lessons learned about empathy and openness (also see Lakoff, 2002, on the “strict father” metaphor for conservatism and the “nurturant parent” metaphor for liberalism). Given that the two types of moral regulation reflect broader differences in approach–avoidance regulation, these distinct liberal–conservative sensitivities to proscriptive versus prescriptive regulation are likely to represent more general differences in approach/activation versus avoidance/inhibition orientation.

These differences in approach–avoidance are not very apparent when dealing with moral self-regulation (the first two columns of the model, with the self or another as the focus of one’s moral concern), because the compatibility of the two forms of moral regulation precludes having to make a choice; one can rely on both, and one’s own action (or inaction) will no doubt depend largely on factors such as the particular behavior involved (e.g., gluttony vs. gambling) or the specific other involved (e.g., close friend vs. stranger). Helping, not harming, moderation, and industriousness are likely, in general, to be desired (moral) outcomes in the behavioral repertoires of both liberals and conservatives. However, when we move to the group level, the more antagonistic nature of moral regulation appears to require that we choose one or the other type of group-based morality. And thus conservatives, more sensitive to proscriptive regulation, opt for a restrictive morality that seeks to protect group members, whereas liberals, more sensitive to prescriptive regulation, opt for an enabling morality that seeks to provide for group members.

When we move to the group level of morality, the regulatory balance afforded individuals is now located at the societal level. The balance is represented by the distinct moral orientations represented in Social Order and Social Justice. Given the potent negative reactions to those across the political aisle, it might seem absurd to suggest that the coexistence of the two group-based moralities reflected in our broadest political orientations—liberalism and conservatism—may be important for a healthy society. Of course, the current political paralysis in Washington makes it clear that the differences can produce extreme challenges and difficulties, particularly as the two perspectives get more polarized and the game of politics takes precedence over the moral tasks to be accomplished. It is nevertheless worth considering that in the best of all worlds, each perspective serves as a counterweight to the other—as a corrective to going too far in one direction or the other. Surely a healthy society embraces both tradition and change, stability and innovation, and works to protect and provide for its members.