Procedural Fairness: Existing Findings and Theory
Over the last four decades, research on procedural fairness in social psychology has provided extensive evidence that people look to certain information about how decisions come about when evaluating group authorities and the decisions they implement (Thibaut & Walker, 1975). This literature has uncovered several criteria for what constitutes a widely perceived legitimate or “fair” decision-making process (Leventhal, 1980; Tyler, 1990). The literature has also demonstrated that evaluations of the decision-making process (procedural fairness) are distinct from evaluations of the decision outcomes (e.g., distributive fairness; see Colquitt, 2001) and, hence, require a distinct set of explanations.
Three studies can serve as illustrations of the criteria people used when making judgments of procedural fairness. A seminal illustration comes from Tyler's (1990) book, Why People Obey the Law. Through panel analysis, it is demonstrated how people who had been in recent contact with the legal system were more likely to comply with and accept an unfavorable verdict when the judge adhered to certain procedural criteria such as allowing them to voice their opinions, appeared impartial and unbiased and included all relevant parties in the decision-making process— even after controlling for the outcome of the decision.
In another illustrative study, Cremer and van Knippenberg (2003) had subjects play a public goods game in which subjects were to decide how much of their own endowment they wanted to contribute to a public pot. If the public pot reached a certain size, the pot would be doubled and then divided among all players, creating an incentive to contribute to providing this public good but also an incentive to freeride and let others make the necessary contributions. Each group of subjects played a total of six contribution rounds and was assigned a group leader deciding throughout the game which subjects deserved a share of the public pot. The authors manipulated whether the group leader allowed subjects to voice their opinions (as opposed to denying voice) before deciding how to divide the public pot and whether the group leader had accurate (as opposed to inaccurate) information about subjects' contribution records to base his or her decision on. As expected, a group leader allowing voice in the decision-making process and basing the decision on accurate information increased subjects' subsequent contribution levels, even after controlling for the payoffs received.
In a final illustrative study, Ramirez (2008) investigated through a survey experiment the effects of media coverage concerning procedural fairness of the Supreme Court on public support for the Court and the individual justices. Specifically, the author had subjects read a news article and manipulated whether the Court was portrayed as allowing all parts to voice their opinions, was representative of the American people, and was objective and impartial in their decision-making. As expected, subjects expressed higher levels of support for the Court and its justices when these procedural criteria were met.
These studies illustrate the findings in the literature on procedural fairness. Other studies have demonstrated how procedural fairness matters beyond the legal context, including in educational (Tyler & Caine, 1981), managerial (Lavelle et al., 2009), and political (Bøggild, 2014; Tyler, 1994) settings. Figure 1 reports six widely considered procedural criteria. The list is not exhaustive and the conceptual terminology sometimes varies from study to study within social psychological literature. Nonetheless, there is general agreement that the listed procedural criteria are of key importance when people make moral evaluations of decisions.
To account for why people pay attention to such features of the decision-making process when evaluating decision-makers and their decisions, research in social psychology has offered three different theoretical accounts. First, early work adopted an instrumental model of procedural fairness, holding that people insist on procedural fairness to gain control over and attain the best possible outcome (Thibaut & Walker, 1975). For example, it is argued that people prefer having a voice in decision-making processes because it serves as a means of control over the
Fig. 1 Procedural fairness criteria. Note: The listed criteria reflect the work by Leventhal (1980), Tyler (1990), Colquitt (2001), and Blader and Tyler (2003)
outcome. This model has been criticized for being too narrow in its scope and for disregarding the normative or noninstrumental motivations related to procedural fairness (Tyler, 1990). For example, studies have demonstrated how people care about being granted voice even after a decision is made, which seems beyond the explanatory range of the instrumental model (Lind, Kanfer, & Christopher, 1990).
Second, to accommodate these shortcomings, Tom Tyler and colleagues proposed the relational model of authority (Tyler & Lind, 1992), holding that people's attention to procedural fairness stems from a desire to construct and uphold a social identity. From this perspective, the individual is highly vigilant about how decisions come about because it provides him with a clear indication of the extent to which the group and especially its authority figure(s) regard him as an equal and valuable group member (Lind & Tyler, 1988; Tyler & Lind, 1992). When information on procedural fairness signals that authorities do not appreciate the status and standing of the individual (e.g., by not allowing voice or being partial), it responds with decreased levels of trust in authorities and compliance with decisions. For example, in favor of this relational account, studies have demonstrated how experiencing unfair procedures also lowers self-esteem of the individual (De Cremer, van Knippenberg, van Knippenberg, Mullenders, & Stinglhamber, 2005; Koper, Van Knippenberg, Bouhuijs, Vermunt, & Wilke, 1993). Critics, however, have argued that the relational model does not explain why procedural fairness matters in largescale settings beyond personal interactions. For example, it seems less straightforward why people would look to information on procedural fairness of national courts or governments to determine whether they are valued group members and construct a social identity on this basis (Leung, Tong, & Allan, 2007). Moreover, scholars have questioned whether the relational model constitutes a theory at all. Simply holding that people have a need for constructing a social identity and that decision-making processes somehow provide relevant information in this regard, it is argued, is merely a statement or at best a description without an underlying theoretical logic (Smith et al., 2007). These critics hold that “Tyler has won the empirical battle but there is no victor in the theoretical war” (Hibbing & TheissMorse, 2008, p. 125).
Third, fairness heuristic theory (Van den Bos, Lind, Vermunt, & Wilke, 1997) constitutes the most recent theoretical framework in procedural fairness research. According to this framework, people often lack relevant information in evaluating and making sense of the outcomes they receive from group decisions. For example, people are often faced with situations in which they are uncertain about the trustworthiness of decision-makers and whether their own outcome is fair compared to outcomes of others. Fairness heuristic theory holds that under such uncertainty, people compensate by applying information on procedural fairness as a heuristic in evaluations of decision outcomes (Brockner & Wiesenfeld, 1996; Van den Bos, Wilke, & Lind, 1998). In support of this model, Van den Bos et al. (1997) have demonstrated how information on procedural fairness has a larger impact on evaluations of decisions when subjects only know their own outcomes and not the outcomes of others compared to when the outcomes of all subjects are known. Yet, fairness heuristic theory has been subjected to the same criticism leveled against the relational model of authority. This framework focuses only on the cognitive processes related to how people apply information on procedural fairness and does not address why such information, over other types of information, helps the individual cope with uncertainty (van Prooijen, 2008, p. 36). That is, “what precisely it is about procedural fairness that makes it useful under conditions of uncertainty remains unclear” (Gonzalez & Tyler, 2007, p. 94).
In sum, although existing models have generated important insights on the effects of procedural fairness, they remain contested at the theoretical level. Our view is that each of these theoretical models has important merits and that an integrative account, which is immune to the criticism leveled against the models, can be formulated through further theorizing. Specifically, we argue that criticisms of the previous models can be accommodated by considering where preferences for procedural fairness ultimately stem from. Despite extensive scholarly attention to the concept of procedural fairness, key researchers in this literature have noted that “…little is known about the origins of procedural preferences” (Tyler, 1990, p. 109) and that what is needed is “a deeper account of the reasons people are so concerned with process variables” (Smith et al., 2007, p. 288). Deep, origin-oriented accounts are important in their own right, but they are also important because we become better able to understand how something works when we know why it exists. That is, by knowing why people care about procedures, we inevitably learn more about how and when we care about them.