Cultural Influence on Judgments of Fairness Running Head: CULTURAL INFLUENCE ON JUDGMENTS OF FAIRNESS Justice for all? Progress in research on cultural variation in the psychology of distributive and procedural justice

We review progress in research attempting to model the influence of culture on judgments of justice. We review research on people’s reactions to resource allocation outcomes, the psychology of distributive justice, as well as on people’s reactions the processes through which authorities make decisions, the psychology of procedural justice. We describe the progress from early work in which culture was equated with country differences to later work which focused on dimensions of values (e.g., individualism-collectivism) that mediate country differences and important contextual factors (e.g., ingroup vs. outgroup) that moderate them. Yet we also describe pitfalls of this research strategy. Finally, we describe the trend toward greater specificity in conceptions of cultural influence-more specific value-dimensions, more specific contextual factors, and the inclusion of specific knowledge structures. Cultural Influence on Judgments of Fairness 3 Justice for all? Progress in research on cultural variation in the psychology of distributive and procedural justice ORGANIZATIONAL JUSTICE Somewhere in the world today, a manager allocates a larger bonus to an older employee than to a younger and more productive co-worker. At another firm, a CEO announces the downsizing of a division without having previously consulted the affected middle-managers. Elsewhere, a foreman interrupts an employee ten times during a brief meeting to discuss the employee’s recent performance problems. Organizational events such as these raise concerns about injustice. Did the manager allocate bonuses in an equitable manner? Did the CEO follow a legitimate procedure for making the downsizing decision? Did the foreman treat the employee fairly? Organizational theorists have long argued that leaders depend on being perceived by organization members as just and legitimate (Barnard, 1938; Weber, 1947). However, allocating organizational rewards in a manner that will be seen as just is not always a straightforward task. Justice and fairness are properties that exist largely in the eye (and the mind) of the beholder. To anticipate the justice judgments that employees and others will make, one must first understand the interpretive processes through which they reach their conclusions. Models of the inference processes that lead to justice judgments have been developed in recent decades by organizational and social psychologists, primarily those working in the United States (for a review, see Sheppard, Lewicki, & Minton, 1992). A key distinction is between fair outcomes addressed by research on distributive justice (Adams, 1965; Deutsch, 1985). Inferences about the fairness of the processes used by authorities to make decisions-both formal procedures and informal personal behaviors--have been captured in research on procedural and interactional justice (Thibaut & Walker, 1975; Lind & Tyler, 1988; Tyler & Bies, 1990). For each kind of justice judgment, models specify the role of abstract criteria of fairness, of contexts that cue these criteria, and of knowledge structures that guide the interpretation of behaviors relevant to the criteria. Cultural Influence on Judgments of Fairness 4 Yet at the same time that justice research in Western organizations has converged on increasingly precise models, studies in non-western settings have revealed cultural differences in the antecedents of justice judgments (e.g., Hundley & Kim, 1997; Leung & Bond, 1984). Although people everywhere seem to go through a computational process of equal complexity in their judgments, the mental software that performs these computations differs across cultures. This paper reviews research on the role of culture in judgments of organizational justice. Before launching into the findings concerning particular justice topics, we first provide an overview of how research on culture and justice has evolved over the past several decades. In so doing, we introduce conceptual distinctions that will structure our literature review. HISTORICAL OVERVIEW The impact of culture on justice judgment does not have a long history as a topic in organizational psychology. For instance, in Barrett and Bass’ (1976) comprehensive review of cross-cultural research in industrial/organizational psychology, the topic of justice did not appear. Nevertheless, many justice-relevant topics were researched under the rubric of other more established topics within industrial/organizational psychology. So although research on these questions has coalesced as a topic only recently, researchers have grappled with these questions for some time. Early Cross-National Comparisons In the 1960s and ‘70s issues relevant to the psychology of justice were addressed within several research areas of organizational psychology. Questions relevant to distributive justice were pursued in research on performance appraisal, such as the question of what criteria are used in assessing the value or deservingness of employees (Barrett & Alexander, 1973). Topics relevant to procedural justice were also addressed in research on leadership, for instance, the question of whether the extent of a managers’ consulting with employees is a key criterion in employees’ preference for their leadership (Haire, Ghiselli, & Porter, 1966). For the most part, early cross-national research was not conducted to test theories of culture or theories of justice. Rather, it assumed that the various criteria salient in Western cultures have the same meaning in non-western cultures and that cultures differ merely in the weight placed on each criterion. Accordingly, the literature on cross-national differences tended to be little more than a compilation of findings of country differences. Figure 1 depicts this conception of cultural influence characterizes the early research reviewed by Barret & Bass Cultural Influence on Judgments of Fairness 5 (1976): Country determines the extent to which a particular criterion is applied in reaching justice judgments. --insert Figure l-Although an exploratory research strategy of mapping country differences in the salience of justice criteria seems a reasonable way to begin research on culture and justice, early crossnational studies on topics related to organizational justice received more criticism than acclaim (Roberts, 1970). Several reasons can be noted. First, country differences are interpretable only if the instrument is demonstrably valid across settings, but this was not always assured. Second, there was little clarity or consensus about which aspects of society or culture were relevant in producing the country differences. The literature consisted of piecemeal findings organized by country, without general principles to organize the patterns and to guide tests of hypotheses. Perhaps as a result of these shortcomings, cross-cultural research did not feedback to influence basic research in the areas (e.g., performance appraisal, leadership style) where early research relevant to justice was conducted. The Ascent of General Principles In the 1980s research on culture and justice underwent a sea change, as more general frameworks emerged to guide research. The identification of more general principles brought a more integrated research paradigm and, consequently, an acceleration of progress and impact. The first development was the coalescence of models of justice judgment. Theories of outcome justice grew from the early work of Adams (1963) on the equity principle to more general models including other principles such as equality and need (Deutsch, 1985). At the same time, theories of process justice developed out of the early work by Thibaut and Walker (1975) on preferences among legal dispute resolution procedures to a more general model of desiderata in organizational decision making processes (Greenberg, 1987). Bies and Moag (1986) and Lind and Tyler (1988) took this a further step in drawing attention to how people interpret their interpersonal interactions with authorities, a set of fairness concerns often referred to as interactional justice. An equally important development was the strategy for cross-cultural research that emerged in the wake of Hofstede’s (1980) well-known analysis of cultural dimensions. By conducting a factor analysis of means from a large sample of countries on a variety of value items, Hofstede distilled dimensions on which countries could be compared. Of greatest Cultural Influence on Judgments of Fairness 6 influence have been the two dimensions that capture the degree of horizontal and vertical differentiation in a society: individualism-collectivism and power distance. The horizontal dimension distinguishes highly individualistic societies (e.g., the United States) from highly collectivistic societies (e.g., China).’ The vertical dimension runs from very egalitarian societies (e.g., Sweden) to very hierarchical societies (e.g., India). Compared with the earlier approach of focusing on country differences, the dimensional approach fostered the integration of findings in terms of general principles of how culture influences social inferences. This approach embodied a culture-general or “etic” approach of analyzing all cultures in terms of the constant set of external constructs rather than an “emic” approach of analyzing each culture in its own terms (Headland, 1990; Pike, 1967). In addition to integrating findings about different countries into a more general picture, the value dimension approach also brought a more refined conception of culture and how it enters the inference process leading to a justice judgment. In justice research employing the individualism-collectivism construct, researchers were able to test competing predictions about the social contexts that moderate cultural differences. For instance, cultural differences are moderated by the relationship between the interactants, i.e., whether they have an ingroup or outgroup relationship (e.g. Leung & Bond, 1984). Figure 2 illustrates this more nuanced conception that grew to dominate cross-cultural social and organizational psychology by the late 1980s: The impact of culture is not seen as a main-effect of country on attention to a criterion but rather as an interactive effect of the cultural dimensions shaping the observer and the social contexts in which the resource allocation event was observed. By drawing attention to the role of social contexts in cueing the application of fairness criteria, this brand of cross-cultural research led to increasing sophistication in basic models of justice judgment. In fact, a major review of cross-cultural research in social psychology (Messick, 1988) singled out justice as an area where cross-cultural research had significantly influenced mainstream theories. --insert Figure 2-A further step in the research program launched by Hofstede’s dimensions has been the effort to measure individual differences in adherence to value dimensions, such as individualismcollectivism. This would allow researchers to assess the extent to which the influence of country on an individual’s behavior is mediated by the individual’s value orientation. This additional step in the research program has proved far more difficult than was originally suspected. Cultural Influence on Judgments of Fairness 7 Various pitfalls have been identified in the reliability and validity of self-report scales measuring abstract values, such as that individuals rate themselves relative to the reference point of the norm in their culture (Peng, Nisbett, & Wong, 1997). Researchers have noted that whereas it is hard to predict specific justice judgments from an individual’s score on very general value dimensions, it is easier to predict justice judgments from more specific belief dimensions (Leung, Bond, & Schwartz, 1995). The questions raised about the role that very general values play in justice judgment echo those that have been debated in other fields. The idea that the best and perhaps only way to account for specific judgments is in terms of highly specific beliefs follows the conclusions of years of research on attitude-behavior consistency (Ajzen & Fishbein, 1983). An even more basic challenge is that individuals do not really possess values and beliefs at this very high level of generality. General structures emerge in aggregate statistics but do not exist in individual minds. In sociology and anthropology, researchers have objected to highly general valueorientations or mentalities imputed to individuals in classical theories; this level of generality may be useful for presenting “composite pictures” that describe differences between societies but not useful in representing the beliefs that guide individual cognition (see Garfinkel, 1967; Rohner, 1984; Shore, 1996). A Turn toward Specificity We have reviewed how the development of more general frameworks brought great progress to research on culture and justice in the 1980s. Dialectically, the progress since then has come primarily from a turn toward more specific: constructs. In the second half of 1990s cultural research has been increasingly losing its resemblance to personality research, in that action is explained in terms of general, ever-present structures, and increasingly gaining a resemblance to social cognition research, in that action is explained in terms of specific, dynamic belief structures (Hong, Morris, Chiu, Benet-Martinez, 1999). More specific cultural constructs are being introduced in several ways. First, conceptions of values are becoming more specific-distinct types of individualism and collectivism are distinguished (e.g., Triandis & Gelfand, 1998). Second, rather than sweeping claims about associations between values and criteria of fairness, researchers increasingly frame their claims in terms of specific contexts that cue different norms or criteria in observers from different cultures (e.g., Leung, 1997). Third, there is increasing recognition that much of the action of Cultural Influence on Judgments of Fairness 8 justice judgment, and cultural differences therein, comes in construing the meaning of behavior so that it can be weighed against a criterion of fairness (Morris, Leung, Ames, & Lickel, 1999). Hence, accounts of cultural differences in justice judgment are being framed in terms of differences in the belief structures used to make sense of the behavior of an authority or conflict opponent. Finally, under the banner of “cultural psychology,” a growing group of researchers advocate value and belief constructs that are specific in the sense that they only have a meaning within a given culture; that is, they are not positions on “etic,” culture-general dimensions. In order to describe these more specific proposals, it is necessary to delve more deeply into the literature on the psychology of justice. Following the chronology of the field, we will do so by first covering research on fair outcomes and then fair process. Within each area, we first describe research on abstract principles or criteria of fairness and then turn to the belief structures relied on in the construal of behavior. In each of these areas, we will first describe research that is based on general value dimensions, which constitutes most of the prominent cross-cultural research. We will then turn to recent studies that are based on more specific cultural constructs, such as sub-dimensions of cultural values, specific beliefs, and the interaction between culture and context. In addition to reviewing these current developments, we will also identify directions that warrant future attention. FAIR OUTCOMES: DISTRIBUTIVE JUSTICE One of the most crucial aspects of management is the allocation of resources to organizational members. There is substantial evidence that the commitment and productivity of employees depend on whether they perceive rewards as having been distributed fairly (see Adams, 1965; Brockner, Tyler, & Cooper-Schneider, 1992; Cowherd & Levine, 1992). Research on the antecedents of distributive justice perceptions involves two components. Some studies focus on the principles or criteria of distributive justice that observers apply to a particular event. For instance, do observers associate justice with equality of allocation or with allocation in proportion to contributions ? Other studies focus on how behavior is construed as relevant to these criteria. For instance, which aspects of a manager’s action are relevant as rewards? Which performances or attributes of employees count toward their deservingness? Applying Criteria Early research on distributive justice suggested that people generally apply the equity principle, that rewards should be directly proportional to contributions, to organizational Cultural Influence on Judgments of Fairness 9 outcomes (Adams, 1963; 1965). Yet subsequent researchers (e.g., Clark & Mills, 1993; Deutsch, 1975) have stressed that in many instances rewards in organizations are distributed according to the equality principle, that everyone’s outcomes should be equal. In addition, Deutsch (1975) has argued that different interaction goals guide the choice of principle: productivity goals are linked with the equity rule, interpersonal harmony is tied to the equality principle, and personal welfare objectives are linked with need-based distributions. General Cultural Value Dimensions Cross-cultural research on distributive justice has been primarily driven by hypotheses about how individualism-collectivism affects people’s choice of criteria to apply. Based on the notion that collectivism requires the preservation of harmony within the ingroup through generosity, Leung and Bond (1982; 1984) predicted and found that Chinese allocative behavior differs from that observed in N. American studies when ingroup members are concerned. In distributions to ingroup members, Chinese participants follow an equality rule when the allocator’s personal input is high but an equity rule when it is low. In distributions to outgroup members, Chinese resembled N. Americans by generally following an equity rule. Nevertheless, subsequent studies have observed violations of this pattern in other collectivist cultures (for review, see Leung, 1997). For instance, Leung and Iwawaki (1988) observed no cultural difference between Japanese, Korean, and American subjects in allocation behavior. Similarly, Marin (1985) found that both Indonesians and Americans preferred equity over equality regardless of the relationship between the allocator and the recipients (strangers, friends, or relatives). And Kim, Park, and Suzuki (1990) reported that Koreans followed an equality rule more closely than Japanese and Americans, who did not differ from each other in their allocation behavior. In an attempt to clarify how collectivism influences reliance on the equity criteria, Hui, Triandis, and Yee (1991) compared the distributive behavior of Chinese and Americans and measured individual differences in individualism-collectivism. An analysis of covariance performed on their data showed that the Chinese tendency to apply the generosity rule was not adequately explained by collectivism. These findings led Hui and colleagues (1991) to conclude that the individualism-collectivism framework may be too global to explain and predict specific allocation behaviors, just as global attitudes are poor predictors of specific behaviors (Azjen & Fishbein, 1983). Testing the implications of this argument, Leung, Bond, and Schwartz (1995) Cultural Influence on Judgments of Fairness 10 found that very general value and attitude constructs are poorer predictors than more specific belief constructs. Advocates of the individualism-collectivism framework have responded by suggesting that constructs at this very general level may be most useful in explaining countrylevel patterns of behavioral syndromes rather than in accounting for how individuals make a particular kind of judgment (Triandis, 1995). More Specific Cultural Constructs To account for cultural differences in particular kinds of judgment, we need mid-range theories that take into account these general cultural dimensions, as well as contextual variables. Recently, Leung (1997) proposed a model to describe how situational factors interact with individualism-collectivism to cue different distributive criteria, and as a result, guide allocation behavior. Based on a goal-directed view of allocation behavior (e.g., Deutsch, 1975; Leung & Park, 1986), this model assumes that interactional goals are the immediate antecedents of the selection of a distributive criterion. Cultures differ in how particular interaction goals are cued by a given social context, such as the relationship between the allocator and the recipient or salient aspects of the context. For instance, Chen (1995) observed allocation following an equity rule among mainland Chinese in an organizational context, which may explained by the salience of the productivity goal in this context, as well as the outgroup relationship between the allocator and the recipient. Murphy-Berman, Berman, Singh, Pacharui, and Kumar (1984) and Berman, Murphy-Berman, and Singh (1985) both found that a need criteria was followed by Indians more than by Americans. Leung (1988) suggested that because resource scarcity is salient in India, the protection of the well-being of the recipient is a more powerful goal than the maintenance of a harmonious relationship. Thus, the collectivist orientation of Indians exhibits itself in the application of a need rule rather than an equality rule. In sum, it is hypothesized that allocators will attempt to follow an allocation criteria based on an interactional goal cued by the situation. This contextual model, which comprises the culture-general construct of individualismcollectivism along with culture-specific factors, provides a reasonably coherent account of the current empirical results. Nevertheless, direct tests of the model are needed to evaluate its usefulness in guiding future cross-cultural work in reward allocation. Construing Behavior In order to apply a fairness criterion such as equity-the rule that rewards be proportional to contributions-an observer must interpret the level of rewards provided, on the one end, and Cultural Influence on Judgments of Fairness 11 the level of contributions, on the other end. The level of rewards conferred by a given managerial act is not self-evident, as it depends in part on the social meanings ascribed to these behaviors. Likewise, the contribution made by particular group members is not self-evident: Which characteristics should be considered as inputs ‘! Which performance dimensions should be weighted most heavily? Even when there is agreement on the allocation criteria, disagreement over what counts as a reward or a contribution could lead to intense conflict. Given the importance of how rewards and contributions are construed, it is surprising that little research has been conducted in this area. To clarify this set of research questions, we will review examples of research on the role of culture-general constructs and culture-specific constructs. General Cultural Value Dimensions Some researchers have used the individualism-collectivism dimension to account for differences in the weight attached to particular employee characteristics when assessing their overall contribution. In collectivist cultures where notions of devotion and sacrifice for the group are salient, long tenure should be valued as an indicator of loyalty and devotion to the group. A recent study by Hundley and Kim (1997) comparing perceptions of input fairness between Americans and Koreans found a pattern consistent with this hypothesis. Respondents considered the fairness of levels of pay for managers with different combinations of job performance, work effort, seniority, education, and family size. Compared to Korean perceptions, American judgments of fairness were more closely tied to performance; Koreans were comparatively concerned with seniority, education, and family size. Before accepting the individualism-collectivism hypothesis, of course, it will be necessary to investigate the extent to which seniority is emphasized in other collectivist settings. Certainly, many studies of Japanese management systems emphasized that that seniority plays a large role in Japanese conceptions of what an employee deserves in compensation (e.g., Ouchi & Jaeger, 1978). Also to fully account for the cultural differences observed, the Korean emphasis on education and family size, for example, it may be necessary to invoke more specific belief structures. Another cultural dimension that is often confounded with collectivism in studies that compare samples from the United States to East-Asian samples is acceptance of hierarchy or power distance (Hofstede, 1980). One way that researchers have conceptualized the effect of this cultural dimension on judgments of deservingness is by determining social comparisons. In Cultural Influence on Judgments of Fairness 12 judging the fairness of an allocation outcome, people often rely on interpersonal comparisons to other employees or groups of employees (e.g., Runciman, 1966). This leads to non-obvious results. For example, the classic findings of Stouffer et al. (1949) revealed that AfricanAmerican soldiers in Southern US were more happy than those in the Northern US despite receiving worse treatment on an absolute scale; the Southern soldiers were more satisfied because the comparison to the condition of African-American civilians was favorable in the South, but unfavorable in the North. The acceptance of hierarchy in a culture may channel the social comparisons that are made. For instance, young Japanese workers rarely report feeling deprived relative to their better-paid elders, probably because in a culture based on age-related hierarchy, young workers do not make social comparisons to older workers. Conversely, the egalitarian ideology in Swedish culture has been implicated in explanations of why pay differentials between blueand white-collar workers have evoked stronger feelings of injustice in Sweden than in other societies (Scase, 1972). Studies of Japanese organizations are also useful in pointing out differences in kinds of performances weighted in assessments of a manager’s rewardingness or an employee’s deservingness. These studies describe workplaces in which employees confer an almost filial loyalty to managers in exchange for paternal regard (Rohlen, 1973). Ethnographic studies present this attention to seniority as part of a system in which the exchange between an employee and organizational authorities involves more than strict task-related performance on the part of the employee and more than monetary compensation on the part of the authority. This analysis can be recast in terms of Fiske’s (1991) framework in that the Japanese system is oriented toward authority ranking, whereas the American system, with a strong emphasis on merit-based compensations, is oriented toward market pricing. Although there is clear evidence that values related to hierarchy play a critical role in employees’ construal of the rewards received and contributions provided, most of the evidence has not explicitly tested whether a culture-general construct, or more specific constructs, are involved. Another influence of cultural values related to hierarchy may be in moderating the bias toward construing one’s own group as more deserving than other groups (Tajfel, 1982). The phenomenon of ingroup favoritism has been systematically related to individual differences in acceptance of hierarchy or “social dominance orientation” (Pratto, Sidanius, Stallworth, & Malle, 1994; Sidanius, 1993). Sidanius, Pratto, and Rabinowitz (1994) proposed that social dominance Cultural Influence on Judgments of Fairness 13 orientation should affect how members of ethnic groups feel about their own group and other ethnic groups. Specifically, for members of high-status ethnic groups, the stronger the social dominance orientation, the more one favors one’s own ethnic group, but the reverse should be true for members of lower-status groups. Sidanius et al. confirmed their hypotheses by showing that for members of the high-status White group, social dominance orientation was positively related to positive perceptions of their own ethnic group, whereas for the lower-status groups, social dominance was negatively related to perceptions of own group. Acceptance of hierarchy may interact with group status to affect how employee characteristics associated with groups are weighted. It might be hypothesized that in a fairly egalitarian culture, people selectively emphasize whichever characteristics portray their group as relatively deserving, showing the classic group enhancing bias (Tajfel, 1982). Yet in a hierarchical culture, people may emphasize attributes that favor the high-status group, such as education level and relevant job experiences (see Jost & Banaji, 1994). However, the extension of Western models of the psychology of groups across cultures must proceed with due caution. Recent cross-cultural studies suggest that perceptions of the ingroups may be less linked to a group’s status in cultures where the social identifications are less a way of gaining self-esteem than a way of gaining connectedness (e.g., Brockner & Chen, 1996). Obviously, the utility of the social dominance construct across cultures is hard to ascertain without empirical substantiation. This would seem to be an important agenda item-not solely for cross-cultural researchers but also for researchers of social dominance-as the cross-cultural differences in social dominance orientation may dwarf the individual differences on which past research has focused. More Specific Cultural Constructs Although very general value dimensions such individualism-collectivism have served as a useful scaffolding for the construction of an integrated cross-cultural psychology, we emphasized several ways in which researchers have found it fruitful to depart from this level of analysis by incorporating other kinds of constructs into accounts of cultural differences. Another response has been to propose more specific value constructs, that maintain a generality across contexts yet have a closer relation to particular sets of judgments. For instance, rather than an uni-dimensional individualism-collectivism construct, researchers have followed the lead of scholars in other disciplines (Lukes, 1973) by distinguishing different strands of individualism and collectivism (Schwartz, 1994). For instance, Triandis and Gelfand (1998) have developed a Cultural Influence on Judgments of Fairness 14 measure of social values and attitudes that distinguishes the economic aspect of collectivism (sharing resources with the group) from the self-expressive aspect of collectivism (affective involvement with the group), and the economic aspect of individualism (belief in achievement through competition) from the self-expressive aspect. of individualism (desire to be unique). Although these constructs have yet to be applied in experiments on resource allocation preferences, the distinctions between types of collectivist cultures may prove useful in developing more fine-grained hypotheses about the role of cultural values in judgments of distributive justice. The potential of less global value constructs in combination with other constructs can be seen in a recent study of managers in different cultures choose to allocate assistance to coworkers. In a multi-country study of managers in the same international organization, Morris, Podolny, and Ariel (1997, 1999, in press) observed evidence that these different strains of individualism and collectivism have distinct consequences. In the United States, where high economic individualism, low collectivism, and short-lived relations make for a market orientation toward workplace relations, managers direct assistance toward others who have directly provided help in the past. Although the data show some aspects of values, relationships, and behaviors in which this individualistic society differs from three collectivist societies, there are as many dimensions on which the three collectivist societies differ amongst each other. In Hong Kong, collectivism takes the form of a familial orientation combining high economic collectivism with endorsement of between-group competition and endorsement of hierarchy; Hong Kong managers were inclined to help powerful others and those tied to them. In Germany, collectivism takes the form of a legal-bureacratic orientation combining low economic individualism, low affective collectivism, and an emphasis on formal rules and categories; German managers were inclined to follow the lines of the formal organization in deciding whom to help. In Spain, collectivism takes the form of an affiliative orientation combining high expressive and economic collectivism with warm and enduring friendship relations. Spanish managers were more likely to help co-workers who were embedded in their friendship network. In addition to influencing which characteristics of others make them deserving of one’s help, these different kinds of cultural collectivism also appear to influence how managers allocate business opportunities. Morris, Podolny, and Ariel (1997) also found differences in managers tendency to direct discretionary business toward their own group or other groups. In sum, Cultural Influence on Judgments of Fairness 15 researchers have found that value dimensions at a slightly more specific level may afford more purchase on how resource allocation varies across cultures. Richer theories of culture and outcome justice may develop if we accept that different models are needed for different types of settings and resources. Highly specific constructs are typically generated by research from an emic perspective. For instance, culturally distinct concerns in interpreting a manager’s action or an employee’s deservingness have been uncovered in this line of work. Ethnographic studies in Chinese societies point to a complex set of rules in considering the value of an employee or colleague as a function of that person’s interpersonal networks. The value of an employee’s connections is understood, no doubt, in many Western settings, but its role is generally left implicit and unarticulated. Not so in Chinese settings, where the concept of guanxi describes the value of an individual’s network of relationships (King, 1991). Studies of how guanxi works to regulate exchanges shed light on why relationship-based patterns of resource exchange are considered fair in Chinese societies while they often evoke charges of cronyism or nepotism in a Western setting. Similar systems of evaluating persons in terms of their relationships, because of the obligation for preferential treatment that relationships carry, have been noted in ethnographic studies of Spanish cultural settings, in which an important resource is a person’s network of persons from whom favors can be asked (conocidos or compadres Femadez-Dols, 1996; Stephens & Greer, 1995). Although arguments about culture-specific constructs have been influential in the applied literature on fairness issues when managing across cultures, they have had little influence on mainstream theories of justice, in part, because qualitative ethnographic evidence is hard to integrate into the psychological literature. In addition to increased focus on more specific cultural dimensions, researchers have also become more attuned to historical change in the kinds of inferences and comparisons that tend to occur in a particular context. As Martin and Harder (1994) argued, political and economic ideologies guide how employees construct the meaning of a total package of rewards offered. For instance, Leung, Smith, Wang, and Sun (1996) examined the social comparison processes that determine justice judgments among local employees of foreign joint ventures in China, in which Chinese and expatriate managers had dramatically different pay despite similar responsibilities. Chinese employees were expected to have justice concerns centering primarily on comparisons within the ingroup. They reported that for the local employees, comparison with Cultural Influence on Judgments of Fairness 16 other local employees was related to job satisfaction. but comparison with expatriate managers was not. Furthermore, the Chinese employees did not perceive the higher salaries earned by the Western managers as unjust. Chinese managers were aware of the fact that these expatriate managers were from developed nations, in which salaries were much higher, and regarded the relatively high pay of expatriates was an acceptable reward for the unique contributions they provided. However, when this survey was repeated after three years of rapid economic and social change in this region, the social comparisons determining justice judgment had shifted (Leung, Wang, & Smith, 1998). In contrast to previous results, comparison with expatriate managers provided additional variance to the prediction of job attitudes. Also, local employees were much more likely to perceive their pay as unfair in light of the high salary of the expatriate managers. Leung et al. (1998) argued that one reason for the heightened sensitivity toward the high pay of expatriate mangers is that many more locals had worked closely with expatriates. After this experience it was more likely that Chinese employees would perceive that expatriates do not greatly differ from locals in their level of expertise. That is, the increased contact leads the locals to engage in direct social comparison with expatriates, leading to a high level of perceived unfairness. This is actually a case in which contextual factors are more important than cultural factors in determining justice judgments. In summary, several developments in recent research on culture and judgments of fair outcome suggest a trend toward modelling culture by taking into account culturally and historically contingent values and beliefs. FAIR PROCESS: PROCEDURAL AND INTERACTIONAL JUSTICE In a program of research in legal settings, Thibaut and Walker (1975, 1978) demonstrated that perceptions of the justice of a decision are not solely a matter of the favorability of the outcome, but also of the process used to arrive at the outcome. Thibaut and Walker (1975) found that people prefer conflict resolution procedures that grant disputants control over the process. In particular, people favor procedures that offer control over the process while yielding control over the final decision to a third party. The most common example embodying these two characteristics is the adversary adjudication legal system used in English speaking countries. In studies of procedural choice, people choose such procedures even when they think other procedures are more likely to bring a favorable outcome. In support of this Lind, Erickson, Friedland, & Dickenberger (1978) showed that even in France and Germany, in which adversary adjudication was not used as a legal procedure, this procedure was still most preferred. These Cultural Influence on Judgments of Fairness 17 studies of legal procedures launched an ever-growing research area on how the process of reaching a joint decision determines the perceived justice of the decision. Current work has shown that perceptions of fair process are related both to formal properties of the decision procedure and to the interpersonal treatment received from authorities. This interpersonal concern is often known as interactional justice (Bies & Moag, 1986; Tyler & Bies, 1990). Cross-cultural research on procedural and interactional justice is quite limited, but the available evidence suggests that on an abstract level, people’s justice perceptions are determined by similar principles across cultures. The original finding by Thibaut and Walker (1975) that procedures granting disputants more process control are preferred has been replicated in many countries other than the U.S., including Britain, France, Germany (Lind, Erickson, Friedland, & Dickenberger, 1978), Hong Kong (Leung, 1987), Japan, and Spain (Leung, Au, Fernandez-Dols, and Iwawaki, 1992). With regard to interactional justice, Leung and his colleagues have shown that similar to the U.S. findings, similar constructs of interactional justice, such as respect and concern received from the decision maker, drive the perception of fairness in Hong Kong and China (Leung, Chiu, & Au 1993; Leung & Li, 1990; Leung, Smith et al. 1996). Lind, Tyler, and Huo (1997) recently reported that the impact of voice, the opportunity to express one’s views, on procedural justice judgments was mediated by relational variables (status recognition, neutrality, and benevolence). More important to our purpose, this pattern is similar across the U.S, Germany, Hong Kong, and Japan. The consequences of perceived procedural and interactional justice are also similar across cultures. For instance, perception of procedural and interactional justice is related to positive perception of decision outcomes and decision-makers in the U.S. (Lind & Tyler, 1988; Tyler & Beis, 1990). Leung and his colleagues have documented similar effects of perceived procedural and interactional justice with Chinese from Hong Kong and China (Leung & Li, 1990, Leung, Chiu, & Au, 1993; Leung, Su, & Morris 1998). Pearce, Bigley, and Branyiczki (1998) replicated the positive effects of procedural justice on organizational commitment and trust in co-workers in a Lithuanian setting. It is important to note that in most recent studies of procedural justice researchers have surveyed participants in different countries about their experiences with authorities rather than comparing responses to a fixed objective event. In these studies, the hypothesized antecedents of procedural and interactional justice perceptions are cast at level that is considerably abstracted Cultural Influence on Judgments of Fairness 18 from the objectively observable features of manager’s behavior. That is, rather than predicting perceived justice from particular managerial statements or actions, researchers have predicted perceived justice from construal of the manager’s words or deeds, such as that the manager conferred “voice” or “dignity.” It is certainly interesting to learn that these abstract components of procedural justice seem to be present across cultures, yet we have not learned whether the same behaviors will be perceived as just or not, because we don’t know whether these behavior would give rise to the same interpretation in terms of abstractions like voice and dignity. Hence, when reviewing the literature, it is again important to include not only abstract reasoning about justice but also the construal of concrete behavior. General Cultural Constructs Applying Criteria Just as researchers of distributive justice have argued that the salience of distributive criteria, such as equity or need, differ as a function of cultural dimensions, the same has argument has been made about salience of procedural criteria, such as harmony or voice. The notion that people in collectivist societies prefer to resolve conflicts through procedures that sustain interpersonal harmony has been advanced by scholars in several disciplines. Anthropologists, such as Nader and Todd (1978) and Gulliver (1979), have argued that in societies in which interpersonal relationships are stable and ongoing, interpersonal harmony is valued and procedures that allow for compromises, such as mediation and negotiation, rather than adjudication, are preferred. Based on these results, Leung (1987) proposed that the supportive evidence for the Thibaut and Walker (1975) theory came from individualist societies, whereas the contradictory anthropological findings were obtained in collectivist societies. To directly test this notion, Leung (1987) compared the procedural preferences of Chinese and Americans. As expected, Chinese preferred mediation and negotiation to a larger extent than did Americans. These procedures are associated with interpersonal harmony because they involve cooperation between the two disputants. Later studies replicated the difference in comparisons of other individualistic and collectivist cultures (Leung, Au, et al., 1992). Leung (1987) also shed some light on the place of culture in a cognitive process model of procedural choice. Contrary to some anthropological accounts, Chinese and Americans differed not so much in their values as their beliefs; Chinese preferred bargaining and Americans preferred adversary adjudication because each believed it would lead to a harmonious outcome, Cultural Influence on Judgments of Fairness 19 which both of them valued. The reason for their stronger preference for these procedures was their greater expectation that these two procedures were conducive to restoration of harmony. This raises the question of why expectations about procedures differed. Morris, Leung, and Sethi (1999) proposed that these expectations may result from differential tendencies to make dispositional attributions for behaviors of conflict opponents, based on several studies indicating that procedural expectations and choices are strongly determined by the characteristics ascribed to one’s opponent, such as uncooperativeness and emotionality (Morris, Larrick, & Su, 1999). They reasoned that Chinese may simply be less likely to ascribe these negative dispositions because of a more situation-sensitive, less dispositional social theory (Morris & Peng, 1994; Menon, Morris, Chiu & Hong, 1999). Morris, Larrick, and Su (1999) replicated the finding of a greater Chinese (than N. American) preference for bargaining and found it was, in part, determined by the fact that Chinese were less likely to ascribe negative dispositions to their. opponent. The direct link between individual scores on an individualism-collectivism scale and procedural choices was weak, suggesting that although the broad value-dimension is useful in characterizing cultures, models of cognition and decision making need to be constructed on more fine-grained constructs, such as the belief structures guiding attributions. Recent work provides compelling process evidence that cultural differences reflect reliance on such structures (Chiu, Morris, Hong, Cheng, & Menon, 2000). Another perspective on culture and procedural preferences is represented in the work of Bierbrauer (1994), who explored the perceived legitimacy of modes of conflict resolution among Germans and among Kurdish and Lebanese immigrant groups in Germany. Bierbrauer (1994) found that compared with Germans, Kurdish and Lebanese immigrants relied on tradition as a criteria: They accepted the religion and tradition as more legitimate guides to resolving conflicts, and state law as less legitimate. In line with the individualism-collectivism framework, which suggests that norms have a bigger impact on collectivists than individualists (Triandis, 1995), Bierbrauer (1992) also found that Kurds and Lebanese reported a higher level of shame and guilt if they violated state laws, traditional norms, or religious norms. For both cultural groups, scores on a collectivism scale were correlated with the level of shame and guilt experienced after violation of norms or rules. These results were interpreted to support the notion that collectivist cultures are associated with a preference for tradition as a criterion. However, it should be noted the use of immigrant groups creates certain problems of Cultural Influence on Judgments of Fairness 20 interpretation: Their predicament as immigrants in Germany, rather than their cultural values, may contribute to their endorsement of collectivism measures concerning the importance of one’s group and for their tendency to accord less legitimacy to legal procedures. A multicountry study of conflict resolution by Morris, Williams et al. (1998) found that an individual’s value on respect for tradition predicts conflict resolution behavior and accounts for cross-national differences, but that this more specific value of respect for tradition cannot be reduced to a general collectivism orientation. Another focus of hypotheses about cultural values and procedural justice has been power distance or acceptance of hierarchy. In a recent review, James (1993) concluded that cultures high in power distance are associated with the tolerance of harsh treatment by authorities. In high power distance cultures, people’s acceptance of unequal social prerogatives promotes the tolerance of unfair treatment. By contrast, in low power distance societies, people’s rejection of inequality makes them less tolerant of unfair treatment received. The evidence for this claim comes largely from differences among countries in Europe in how people respond to organizational authorities. Gudykunst and Ting-Toomey (1988) analyzed the data on anger and justice from 7 European countries reported by Babad & Wallbott (1986) and Wallbott & Scherer (1986), and found that the higher the power distance of a society, the less that perception of unjust treatment triggers a reaction of anger. In a study of conflict resolution, van Oudenhoven, Mechelse, and de Dreu (1998) found that respondents from high power distance countries, such as Spain and Belgium, were less likely to rely on problem solving to resolve a conflict with their superiors. Again, however, findings that seem to indicate a pervasive influence of general cultural values may actually reflect more subtle causal dynamics. Studies contrasting American and Hong Kong samples provide evidence about the moderating conditions and mediating processes. Bond, Wan, Leung, and Giacalone (1985) found that, compared to Americans, Hong Kong Chinese are more willing to accept insulting criticism from a high-status ingroup person. No cultural difference was found, however, when the insult came from a low status individual or from a high-status person who is not from the ingroup. Leung, Su, and Morris (1998) put American and Hong Kong MBA students in the role of an employee whose suggestion was criticized by a manager in a manner that violated the prescriptions of interactional justice: that is, the manager interrupted, failed to listen closely, and was dismissive toward the employee. As Cultural Influence on Judgments of Fairness 21 expected, compared to Americans, Chinese perceived the senior manager’s actions as less unjust and were less likely to reduce their loyalty to this manager. The judgment of injustice again seemed to depend on the attributions made for the manager’s behavior. Hence, it may not be merely that Chinese apply different criteria in judging the fairness of an interaction with a boss but also that they access different beliefs to interpret the manager’s action. In sum, studies suggest that in high power distance societies, legitimate authorities can treat subordinates more harshly before this behavior is perceived as unfair. Again, however, the data do not suggest that individuals’ adherence to hierarchical values directly mediates their fairness judgments; rather it seems that more specific belief structures involved in guiding attributions may be the proximal cause of the cultural difference in judgments. More Specific Cultural Constructs The criteria for judging the fairness of conflict resolution procedures that are specific to particular cultures have not received much explicit research attention. However, directions for research are suggested by the literature on culture-specific conflict resolution procedures. We illustrate this point by three examples, one political, and the other two organizational. The most well known non-western procedure of conflict resolution is probably ahimsa, a procedure employed by Gandhi against the British Colonial regime (Sinha, 1987). Ahimsa involves a very different conceptualization of procedural justice, and the criteria central to Western models, such as process control, seem peripheral to this procedure. Ahimsa involves two basic principles, maha-karuna (great compassion) and maha-praina (great wisdom), and the assumption that the strong emotional attachment and love towards all beings and the perception that all beings are equal would prevent people from harming other beings. Three steps of conflict resolution follow from these two principles. The first step involves persuading the other disputing party through reasoning. If reasoning fails a person next turns to self-suffering, which intends to arouse a feeling of guilt in the opponent and/or put the opponent under moral pressure for inflicting harm on a “helpless” and non-violent individual. If both steps fail, non-violent coercion such as non-cooperation, civil disobedience, boycott, and fast may be adopted. The ahimsa procedure centers on heightening moral self-awareness of the parties involved, which is seen as a potent force to move a dispute toward peaceful resolution. Unfortunately, constructs implied by the ashimsa procedure have not been systematically studied. Cultural Influence on Judgments of Fairness 22 Our second example is drawn from Cho and Park (1998), who have proposed a four-stage model for understanding how Koreans resolve workplace conflict. The first stage is context building, in which information and feelings triggered by the conflict are shared, and effort is made to build strong emotional bonds between the disputants. The second stage is smoothing, in which conflict avoidance and compromises are attempted to protect each other’s face and prevent ill-feelings against each other. If the conflict persists, forcing will be used in the third stage, in which one side attempts to force the other part to yield. Alternately, authorities may be asked to intervene in the conflict, a tendency that reflects the high power distance of the Korean culture. The final stage is tension release, in which joint social activities, such as karaoke singing and drinking, is used as a way to mend the relationship and to establish ingroup harmony. This Korean cycle of conflict management is interesting because it highlights a gap in Western research in conflict resolution. Western conflict models typically do not articulate criteria for the post-resolution stage, whereas in Korea this process is conceptualized as a key element of the conflict resolution process. One research direction prompted by this Korean emit model is that we need to explore how disputants would feel about and interact with each other after they have settled a dispute. Perhaps expectations and feelings about the post-resolution stage may help explain why disputants choose to handle a conflict in a certain way. The third example is concerned with the interpretation of business contracts. It is well documented that in individualist societies, such as the U.S., business contracts are considered to be binding. In collectivist societies, however, business contracts are often regarded as symbols of collaboration, and their contents are taken to signify common understandings, rather than binding prescriptions that are not affected by situational changes (e.g., see Lubman, 1988 for the case of China; Sullivan, Peterson, Kameda, & Shimada, 1981 for Japan). Leung and Stephan (1998) provided an interesting case of how this cultural difference in the meaning of a contractual agreement fueled the Sino-British conflict with regard to the Basic Law, a legal document that prescribes the ruling of Hong Kong after Britain returned Hong Kong to China, and the Joint Declaration by China and Britain about the future of Hong Kong. Britain took a legalistic interpretation, which clashes with the symbolic interpretation adopted by China. In summary, although hypotheses concerning, cultural influence on the criteria applied in procedural justice judgments has focused primarily on general dimensions of values such as collectivism and hierarchy, a number of research programs suggest that the crucial value Cultural Influence on Judgments of Fairness 23 dimensions may be more specific. Also, studies suggest a role of concrete beliefs in mediating the application of criteria and suggest that criteria application is moderated by contextual variables, including the historical context. Although there has been little research directed at the procedural criteria idiosyncratic to particular cultures, examples of emic understandings such as ahisma suggest that there is potential for further investigation. Construing Behavior Before a manager’s behavior can be tallied against criteria of procedural justice, it first must be construed as meaningful. Detecting a managerial sin of omission or commission requires applying knowledge to interpret the behavior. Again, the story begins with research that more or less equates cultural influence with observed country differences. For instance, a popular interpretation of the success of Japanese firms in the late 1970s in industries like automobile manufacturing where American firms were hindered by conflict associated with justice issues was that Japanese employees and managers trust each other and hence make charitable interpretations of each other’s behavior (Ouchi & Jaeger, 1978). However, the claims made about trusting others as a basic aspect of Japanese psychology have been embarrassed by consistent empirical findings that Japanese individuals are less trusting than Americans, when there is no incentive structure that makes the expectation of benevolence a reasonable assumption (see review by Yamagishi & Yamagishi, 1994). As we have seen in our reviews of previous facets of justice, research progressed by linking the psychology of construing of behavior relevant to procedural justice to general cultural dimensions as well as to more specific cultural constructs. We examine the kinds of cultural differences that may exist in the behavior construal relevant to the abstract appraisals referred to as benevolence, neutrality, and status recognition, which are key elements of the relational model of procedural justice (Lind & Tyler, 1988). As we have argued, although these abstract criteria of procedural justice may be culture-general, this does not mean that the specific behaviors that instantiate these constructs are the same. General Cultural Value Dimensions The first key criterion of procedural justice in interactions with an authority is benevolence, which refers to the extent to which the decision-maker appears concerned about the well-being of the recipients. However, what is interpreted as showing concern in one culture may have the opposite effect in another culture and the individualism-collectivism dimension Cultural Influence on Judgments of Fairness 24 may moderate some of these relationships. Smith, Misumi, Tayeb, Peterson, and Bond (1989) found that when a co-worker encounters personal difficulties, Hong Kong and Japanese respondents regarded the discussion of the matter by the supervisor with other team members in the absence of this person as considerate behavior. In direct contrast, respondents from the U. K. and the U.S. regarded talking “behind the back” of this person as inconsiderate. Clearly, the positive gestures that would lead to a manager being perceived as benevolent differ between individualistic and collectivist settings. Likewise, the gestures that undermine the appearance of benevolence also differ. Consider the example of public criticism, especially by an outgroup person. It is well known that in collectivist societies, social face is highly valued (Choi, Kim, & Kim, 1997; Gabrenya & Hwang, 1996). Roongremgsuke and Chansuthus (1998) have provided a dramatic illustration of how an American manager unknowingly generated a great deal of resentment from his Thai subordinate by criticizing him in front of his subordinates. Clearly, it is a great loss of face to be criticized in front of one’s subordinates by one’s senior in Thailand. The Thai subordinate was angry because of the lack of concern for his face by his American boss. To Americans, public affront perhaps carries much less stigma and hence is much less of an issue to deserve thoughtful consideration. Indeed, the willingness to provide candid feedback that might enable a manager to appear benevolent in a North American context might have precisely the opposite effect in Asia.* Another criteria of procedural justice is neutrality, which refers to the extent to which decision-makers are seen to act in an unbiased and fair manner. Construing behavior in terms of neutrality may hinge on beliefs related to hierarchy. In high power distance cultures, a highstatus third party is expected to intervene in a dispute and help settle the dispute in a fair manner. In contrast, in low power distance cultures, disputants prefer to resolve the conflict on their own, and they rather prefer the high-status third party to play the role of a passive referee. The active involvement of the third party may actually be seen as unfair. Consistent with this argument, court litigants look to the judge to provide facts about the case and ultimate justice in high power distance countries like Hong Kong and Japan, whereas American litigants are more inclined to rely on their own efforts to argue for their case (Benjamin, 1975; Leung & Lind, 1986; Tanabe, 1963). Tse, Francis, and Walls (1994) found that compared to Canadian executives, executives from China were more likely to consult their superior in a conflict. Morris and Su (1998) Cultural Influence on Judgments of Fairness 25 demonstrate that when a manager intervenes, as a mediator, in the United States the perception of neutrality is difficult to maintain. They point out that this may be why N. Americans often seek mediators who are strangers to both parties, whereas Chinese seek mediators who have a relationship of being superordinate to both parties. These studies suggest that in high power distance societies, managers are expected to intervene in disputes among subordinates and play an active role in resolving them whereas the same behavior might run afoul of neutrality in a more egalitarian setting. The third major criteria of procedural justice is status recognition, which refers to the extent to which decision-makers show respect toward the recipients. Although it is universal that people value being shown respect, the behaviors that constitute respect are hardly culturegeneral. Smith and Bond (1998) have provided a review of cultural differences in display rules for expression of emotions. In general, people from individualist cultures are more likely to express their negative emotions than people from collectivist cultures. People from collectivist cultures may regard the expression of negative emotions as a sign of disrespect to the recipient, whereas this feeling should be weak in individualist. cultures (e.g., Matsumoto, Takeuchi, Andayani, Kouznetsova, & Krupp, 1998). More Specific Cultural Constructs There is a dearth of research on the relationships between culture-specific constructs and the construal of behavior to arrive at a judgment of its fairness. This line of work is important because a picture painted in broad brush-strokes may miss many of the subtleties of this judgment process. For instance, the individualism-collectivism dimension does not capture many kinds of variation in non-verbal behavior that likely affect whether a manager’s behavior is perceived as respectful. Graham (1985) has documented many dimensions of behavior in negotiations where behavioral norms differ sharply. His research shows that Brazilians negotiators interrupted each other almost 10 times more than did American and Japanese negotiators. Clearly, the occasional interruption in Brazil is less likely to be interpreted as a sign of disrespect than in the U.S. or Japan, and a lack of interruptions by a Brazilian manager might suggest inattention. Similarly dramatic differences can be noted in many other dimensions of nonverbal behavior. What indicates close attention and respect in one culture, such as closing or lowering one’s eyes while listening, may indicate the opposite in other cultures. Cultural Influence on Judgments of Fairness 26 Likewise, there are cultural differences in how evaluations of a manager depend on particular verbal forms. For instance, apologies are a salient feature of Japanese organizational life. Ethnographers have suggested that apologies are a crucial element of procedure in that they satisfy the criteria of restoring a relationship when someone has been harmed (Haley, 1982). Similarly, quantitative studies of verbal accounts have found that Japanese are more likely than Americans to favor types that acknowledge the harm, such as apologies rather than denials (Itsoi, Ohbuchi, & Fukuno, 1996). Sometimes perplexing to observers is that managers apologize for outcomes to organization members even when the outcome would not likely be attributed to the managers’ actions or intent. The answer seems to be that Japanese apologies are less likely to involve regret for intentional actions but merely a recognition of harm and an expression of submission to the normative order (Wagatsuma & Rosett, 1986). In sum, several research programs suggest that there is a large role played by specific cultural constructs in influencing construals of action relevant to procedural justice. Of all the points of possible cultural influence that we have reviewed, this one has been the least thoroughly researched. Thus, this area calls for more serious attention in future work. Prospects for Progress As we have seen, the role of culture in many aspects of procedural and interactional justice has not yet been empirically investigated. Hence, there is less to say concerning pitfalls of previous research. Instead, we have placed more emphasis on sketching possible areas of research and identifying a number of ideas about cultural influence that seem promising. While the general individualism-collectivism and the egalitarian-hierarchical dimensions have a role, so do culture-specific constructs. This conception is illustrated by Figure 3. --insert Figure 3-Again, we can identify three aspects to the trend toward inclusion of more specific constructs. First, current research seeks to avoid the pitfall of conflating useful dimensions for organizing country-level syndromes with useful dimensions for modelling individual cognition. Very global dimensions served the purpose of Hofstede’s study and other country level accounts, but the values that proximally influence justice judgment are better captured by more specific dimensions. Second, researchers seek to avoid the pitfall of ascribing different judgments to different values without considering differences in beliefs. Values, after all, are “trans-situational goals” Cultural Influence on Judgments of Fairness 27 (Schwartz & Bilsky, 1990). Hence, values are primarily useful in accounting for trans-situational styles. Predicting the details of judgments about a single case is better done by incorporating the specific beliefs that mediate between values and inferences. Third, researchers seek to avoid the pitfall of regarding cultural differences in reactions to decision making processes as static, ahistorical phenomena. While not as dependent on social comparison as is outcome justice, process justice is highly dependent on behavioral norms. As habits and expectations change, the expectations against which a manager’s behavior is construed will also change. In times of change, sharp cultural differences in these matters will exist within a given country. For example, Fahr, Earley, & Lin (1997) have shown that the importance of certain kinds of interactional justice dimensions to employees in Taiwan depends on their level of modernism, a value complex that captures their identification with Western values as opposed to traditional Chinese values. In sum, social contexts affect people because of cultural norms governing behavior in these contexts, so we must situate generalizations about effects of social context within an understanding the of the historical and cultural forces sustaining or changing these norms. CONCLUSION We have presented a selective review of the progress, pitfalls, and prospects in research on cultural influence on judgments of outcome and process fairness. We have argued that the trajectory of research in this area moves from compilation of country differences in the 1970s to modelling the influence of general cultural values in the 1980s, to an eclectic 1990s approach of accounting for differences in terms of more specific constructs, such as specific value dimensions, specific cultural beliefs, and specific moderating contexts. In some areas of justice research, this trajectory has fully played out, whereas in others, research is just beginning. Not only have justice models provided a rich set of hypotheses for cross-cultural work, but cross-cultural findings have spurred refinements in basic models of justice judgment. We hope that this review will stimulate more cross-cultural work on justice, as many gaps in our knowledge of culture and justice have been identified. Furthermore, the range of cultures studied so far has been narrow, with many important regions, such as Africa, largely unexplored. Messick (1988) noted that justice research was one of the first areas in which cross-cultural studies and mainstream theory developed in a constructive, dynamic relationship. In the next Cultural Influence on Judgments of Fairness 28 decade, we hope to see a full integration of culture into justice theories and of justice into culture theories. Cultural Influence on Judgments of Fairness 29 Footnote ’ Recently Triandis and his colleagues (e.g., Triandis & Gelfand, 1998), in response to the long-standing complaint of the imprecision of the individualism-collectivism construct, have proposed two sub-types of individual-collectivism: horizontal and vertical. This refinement is obviously a much welcome develop, but because there is virtually no empirical work in the justice literature that is guided this new approach, we will not attempt integrate the current in light of this development. * It should be noted that in an earlier section, we argued that in collectivist societies, harsh feedback from a legitimate authority is more acceptable. Being the target of harsh remarks does not necessarily constitute a loss of face, especially when the feedback occurs in private. Cultural Influence on Judgments of Fairness 30