This is an open access article distributed under the terms of the Creative Commons
Attribution License (Fairness is a universal moral concern (Graham et al., 2013, 2011). People react to systemic injustices by engaging in protests or other forms of collective actions to restore fairness (Wright, Taylor, & Moghaddam, 1990) and they endorse fairness even at the cost of equality (Starmans, Sheskin, Bloom, Christakis, & Brown, 2017). The legitimacy of authorities such as the police and government also depends on the extent to which they are perceived to be fair (Tyler, 2006). Thus, police can create a positive relationship with the public by treating citizens in a respectful and neutral manner, resulting in higher compliance and community engagement (Mastrofski, Spines, & Supina, 1996; Tyler & Fagan, 2008; Tyler & Jackson, 2014). It has been argued that being treated fairly by authorities is important to people because it communicates to citizens that they are respected group members, while being treated unfairly signals the reverse (e.g. Lind & Tyler, 1988). The present research aimed to investigate the extent to which people also want fairness for others and whether this depends upon their group’s status. Put simply, if fair treatment from authorities indicates that a person or group is valued and respected, do people judge high status groups and individuals as more deserving of fairness than those of low status? Our focus is on the role of expectations of authority in intergroup relations. Thus, we consider the extent to which outgroup status determines people’s expectations of how fairly people should be treated by authority, both in terms of whether they should be given a say over decisions that affect them and in terms of the resources that should be allocated to them. In so doing, we set out to understand whether people expect authority fairness universally, for everyone equally.
Relational Models of Procedural Justice
Procedural justice (PJ; also called ‘procedural fairness’) refers to the extent to which people perceive that they have been treated fairly by others and that the procedures leading to reaching any decisions are considered to be fair. Thus, PJ speaks to the process behind the decision-making rather than the decision itself. People are often found to accept decisions that are made in a fair way and in accordance to some commonly agreed rules, even if they may not come with favourable outcomes (Sunshine & Tyler, 2003). PJ has often been contrasted to distributive justice, which refers to the fairness of an outcome rather than a process. The question of how much people really distinguish between fair processes and fair outcomes in their perceptions of authority is controversial, with some assuming that there are two separate factors (Tyler, 1994) and others arguing that there are substantial overlaps between distributive justice and PJ (Pehrson, Devaney, Bryan, & Blaylock, 2017; Reisig, Bratton, & Gertz, 2007). Nonetheless, one aspect of PJ that has been studied extensively, and which can more straightforwardly be distinguished from the outcome of a decision, is whether people are given the opportunity to voice their opinion before a decision is made (van den Bos, 1999). Voice provides an opportunity for a citizen to participate in the decision-making process and thus conveys that one’s opinion matters to the authorities (Folger, 1977). Indeed, studies demonstrate that voice provision, in contrast to voice denial, increases the perceived fairness of the authority (Hildreth, Moore, & Blader, 2014; Lind, Kanfer, & Earley, 1990; Platow et al., 2013). Even non-instrumental voice, whereby participants are told that they are welcome to express their views but that this would have no effect on the decision, increases perceptions of fairness in comparison to voice denial, suggesting that the role of voice is not only an instrumental concern but also a relational one (Lind et al., 1990). Thus, being offered an opportunity to voice one’s opinion has an important symbolic meaning as it conveys respect even when it does not come with any tangible outcomes; just because an authority considers one’s view does not mean that the outcomes will be in any way lenient.
A prominent explanation of this effect, the group-value model (GVM), suggests that being treated fairly is important because it communicates to people something about their relations with authorities, namely their social standing in society (Lind & Tyler, 1988; Tyler, Degoey, & Smith, 1996; Tyler & Lind, 1992). Status and social standing, used interchangeably in this research, are defined as relative positioning in terms of prominence and respect (Anderson, Willer, Kilduff, & Brown, 2012) that is conferred by others in the group (Halevy, Chou, Cohen, & Livingston, 2012). Much as positive distinctiveness is a source of self-esteem (Hogg & Abrams, 1988), being treated fairly by authorities can be used to infer one’s social standing and self-esteem. Indeed, studies find that experiencing PJ is associated with higher self-esteem (Koper, van Knippenberg, Bouhuijs, Vermunt, & Wilke, 1993) and higher feelings of respect (Heuer & Stroessner, 2011; Platow, Brewer, & Eggins, 2008; Smith, Tyler, Huo, Ortiz, & Lind, 1998). Thus, fair treatment from authorities can affirm one’s status in the group while unfair treatment can disaffirm it.
The relational account of why authority fairness matters, exemplified by the GVM, implies that fair treatment should matter more to those who see the authority as somehow representative of a group that they identify with, rather than of an outgroup. That is, fair treatment would only indicate standing and respect within one’s group if it comes from an authority that is supposed to be acting on behalf of that group. Being mistreated by an outgroup authority, on the other hand, does not convey any identity-relevant information about one’s inclusion and standing in that society (Huo, Smith, Tyler, & Lind, 1996; Smith et al., 1998). Accordingly, people want and expect to be given voice in authority decisions to the extent that they identify with the group in question (Platow, Huo, Lim, Tapper, & Tyler, 2015). At the same time, the inclusiveness of group boundaries is rarely fixed and can also be shaped by the actions of authorities. For example, employers who make decisions in a fair way can foster the workplace identity of employees which, in turn, encourage them to engage in workplace behaviour that benefits the workplace as a whole (Blader & Tyler, 2009). From this perspective, PJ has the potential to engage even individuals who did not previously identify as belonging to the group that the authority represents (see also Bradford, Murphy, & Jackson, 2014). Thus, how authorities treat people is deeply interlinked with individuals’ perceptions of their place and social standing in the society.
From Intragroup to Intergroup Processes in Authority Decisions
While the relational models of PJ tend to describe the relationship between authorities and subordinates as intragroup, we suggest that it can also be conceived in intergroup terms. In other words, authority decisions affect different groups in society and so these relationships can be conceived as multipartite, with an authority serving more than one group in society (Wenzel & Jobling, 2006). Authorities often need to arbitrate between multiple social groups, just as in the case of football referees making decisions about two teams in a match or a judge making calls on which side of the dispute is guilty. In contrast to previous research, which typically concerns police-citizen or employer-employee interpersonal interactions (e.g., Blader & Tyler, 2009; Sunshine & Tyler, 2003), the present research recognises that (1) the role of authority frequently extends beyond dealing with individuals as individuals, one at a time, and (2) holding authority sometimes involves making decisions about groups and managing the expectations from different corners of society.
This view of authorities in intergroup contexts differs from one in which authorities represent one group but not another, for example, university students who receive a decision either from an ingroup authority (the university they are affiliated with) or an outgroup authority (unaffiliated university; e.g., Smith et al., 1998). Thus, we argue that authorities assume regulatory positions whereby they represent not only individuals or one group at a time, but they wield their power in multi-group contexts. Authorities such as governments should, at least in principle, represent all sections of societies whether they are men or women, poor or rich, or national or non-national residents. Relationships between groups in society, however, can be construed in intergroup terms even if the groups are represented by the same authority system.
Because authorities are dealing not only with individuals as individuals, but also with subgroups and their (sometimes) competing needs and expectations, authority fairness can also be perceived in group terms. It is well established, for example, that the experience of relative deprivation can be experienced either individually or collectively (Runciman, 1966). In the same way, relational models of fairness can also be extended beyond the intra-group level: where there is a perception of systematic unfairness towards one’s group, rather than merely oneself as an individual, then we would expect the lack of status or standing of the group to be similarly felt as collective. In other words, if one feels individually subjected to unfairness compared to other group members this may convey that one is not respected within the group, but if one’s group is treated unfairly then it more likely conveys that the group is devalued in relation to other groups in society. People may interpret authority unfairness as a devaluation of their group rather than as a devaluation of themselves as individuals.
Fair for All?
Do people think authorities should give fair consideration to others, or is fairness only a consideration when our own treatment is at stake? One line addressing this question is the scope of justice theory, which states that our concerns regarding whether people are treated fairly or not is bounded by a psychological group (Deutsch, 1985; Opotow, 1995). In other words, for those who are perceived to belong to the same moral community, the same rules of fairness apply. To the extent that outgroups are seen to be outside this community, there may be less concern if they are denied fair treatment. Yet, as we have already suggested, authorities often have a role in managing relations and claims among a number of groups. At least some consensus about the legitimacy of this authority is underpinned by the expectation of fairness for all (Tyler, 2006), which would not work if authorities blatantly served one group over another. This then raises the question of when people are concerned for the fair treatment of outgroups and when they are not.
One possibility, given the relationship between fair treatment and implied social status discussed above, is that the relative status of an outgroup affects whether people think they should be treated fairly or not. Groups with high status, such as the economically advantaged, are frequently perceived as more competent than lower status groups, and hence may be seen as more deserving of respect (e.g., Fiske, Cuddy, Glick, & Xu, 2002). On the other hand, it may be that moral and normative commitment to fairness is such that people expect fair treatment of outgroups irrespective of status. Indeed, one study from Huo (1995; as cited in Tyler, 1997) suggests that people may be unlikely to deny treatment as a human being to even outgroups, but this was not studied from the perspective of how authorities should treat those outgroups. Furthermore, if a form of PJ such as voice comes without tangible resources, it is also plausible that people are happy for authorities to give voice to even low status groups since this does not cost the ingroup anything whilst preserving their moral image as fair. Thus, the extent to which people’s support for procedurally fair treatment of outgroups depends on status remains an open empirical question.
Evidence on authority decision acceptance in intergroup settings suggests that people may be less egalitarian when it comes to resource allocation rather than PJ. People have been shown to accept fair decisions even when the outcome is unfavourable for them as individuals (Sunshine & Tyler, 2003), but they appear to be more instrumental in the case of decisions that affect the ingroup. For these kinds of authority decision, outcome favourability, or the extent to which the decision favours the ingroup, appears to matter more than whether it followed a fair process; decisions with favourable outcomes to the group are more likely to be endorsed than unfavourable decisions, regardless of the perceptions of PJ (Leung, Tong, & Lind, 2007). Even decisions that are made in normatively unfair ways can sometimes be accepted as long as the outcomes are what one was hoping for (Skitka & Mullen, 2002). Leung et al. (2007) explain this in terms of different ‘moral mandates’ that apply at individual versus group levels. For matters affecting individuals as individuals, fairness is generally morally praiseworthy compared to selfishness, and therefore there is a strong expectation that people should accept a fair decision even if it has negative consequences for them personally. However, loyalty and commitment to group interest carry a more positive moral valence than pursuit of individual self-interest and thus in the case of group-level decisions people may feel mandated to prioritise the favourability of the decision over fairness.
What might this mean for people’s expectations of authority decisions about resources for outgroups? Relatively high or even equal status outgroups tend to be perceived as more competitive than lower status outgroups (Fiske et al., 2002). Therefore, people may want authorities to restrict resources to equal or higher status outgroups than lower status outgroups in order to make them less competitive. On the other hand, however, lower status groups may be perceived as not deserving resources due to their presumed incompetence (Fiske et al., 2002) or simply because they are disliked (Tajfel & Turner, 1979). Thus, the role of outgroup status in expectations of resource allocation by authorities could also take different forms.
As far as we know, the role of intergroup status in expectations of how authorities should serve the people they represent both in terms of the procedural justice and the favourability of outcomes they offer has not been directly tested before. Exploring the interplay between people's identity and their expectations of fairness will increase our understanding of what may drive perceptions of authority fairness in intergroup contexts.
The Present Research
The present research investigated the role of groups’ relative social standing on people’s expectations of fairness to establish whether lower status outgroups are expected to receive different treatment from authorities than those perceived to be of equal status to the ingroup. By referring to expectations of authority fairness, we focus on two facets, namely PJ (operationalised as the extent to which people are provided voice by authorities; see Folger, 1977) and outcome favourability. Concerning the former, we predicted that because procedural justice may enhance social standing, people may think authorities should give more voice to higher status outgroups than lower status outgroups, and more voice to the ingroup than the outgroup regardless of status. Concerning the latter, given the increased importance of outcomes rather than fairness at the group-level, we predicted that group status may shape expectations of how authorities should distribute outcomes. However, the direction of this effect was not specified as both directions are theoretically plausible: people may expect authorities to give more favourable outcomes to lower status groups (as they are less threatening to the status of the ingroup) or more favourable outcomes to the equal status groups (because they may be perceived as more deserving of the outcomes due to their status). Finally, the present research aimed to replicate the findings of Platow et al. (2015) which demonstrated a positive relationship between social identification and ingroup voice expectations.
The hypotheses were tested in three experimental studies. Across these studies, the status of EU citizens living in the UK (outgroup) was manipulated as either lower or equal to that of a typical British citizen (ingroup) to evaluate the effects of group status on authority fairness expectation. In all three contexts, administratively speaking, the authority serves both British citizens and foreign-born residents. While the literature typically talks about higher and lower status groups, these are not high and low in absolute terms. Since our interest lies in the relative intergroup status, we adopted the lower and equal status terminology in our studies to reflect the relative intergroup status between British and foreign-born residents. Studies 1 and 2 were carried out in the context of British citizens voting to leave the European Union. The UK government’s decision on post-Brexit immigration policy was the focus of those studies. In Study 3, we provided a further test of our hypotheses by changing the authority to a local city council in a context of dispute regarding the organisation of a cultural festival. Going beyond intergroup status, Studies 2 and 3 further tested the role of intergroup threat in shaping expectations of authority fairness. Finally, an internal meta-analysis was conducted to aggregate the effects.
Taken together, these studies contribute to the theoretical conceptualisations of fairness in the context of authority decisions. We address whether groups expect authorities to be fair to all groups, independently of their social status, by providing them with a voice to communicate their opinions and by making favourable decisions. By studying these processes, the present line of investigation has the potential to better inform authorities on the nature of these expectations and how they may predict public dissatisfaction with the authority action towards other social groups. Ethical approval for all three studies was granted from Queen’s University Belfast Faculty of Engineering and Physical Sciences Research Ethics Committee (formerly Psychology Research Ethics Committee).
For Studies 1 and 2, we did not conduct a-priori power analysis. Instead, we calculated whether the design was adequate to detect even small effect sizes using a sensitivity power analysis. This analysis was based on our main hypotheses regarding the impact of lower versus equal status outgroups on the dependent variables of interest and was conducted in G*Power software (Faul, Erdfelder, Lang, & Buchner, 2007). The minimum effect size detectable in the present designs (two groups, 1 df) at 80% power, α = .05 was f = .23 (f2 = .05) for Study 1 (N = 155) and f = .20 (f2 = .04) for Study 2 (N = 202), demonstrating that our sample was of sufficient size to detect small to medium effect sizes. Study 3 was pre-registered (for the study protocol see the Supplementary Materials section).
Study 1
Studies 1 and 2 were carried out in the context of the aftermath of the 23 June 2016 UK referendum vote to leave the European Union (EU). The free movement of people and economic immigration were amongst the most prominently debated issues throughout the campaign with calls for control measures to be implemented (Alfano, Dustmann, & Frattini, 2016). Concerns regarding control over immigration were indeed more prominent among those voting to leave the EU than those voting to remain (Abrams & Travaglino, 2018). Studies were carried out between December 2016 and May 2017 with no plans regarding any policies post-Brexit revealed by the government by that time. The experiments were centred around the potential policies regarding non-UK EU immigrants and their status in the UK following Brexit. In 2015, there were 3.2 million EU citizens living in the UK, thus making up around 5% of the UK population (Office for National Statistics, 2015). We hypothesised that the equal status outgroup would be expected to receive more voice from the UK government than the lower status outgroup. We also assumed the ingroup to expect more voice for themselves than for the outgroup overall. In terms of outcome favourability, we did not predict whether the lower or equal status group would be expected to receive more favourable outcomes, but we anticipated intergroup status to predict outcome favourability. Moreover, since Leave voters tended to have higher levels of anti-immigrant prejudice, we assumed that they would expect government to give less favourable treatment to outgroups.i
Method
Participants
One-hundred-and-seventy-two participants were recruited via social media dedicated to topics surrounding politics and Brexit such as Twitter and Reddit. The study was advertised as seeking views on Brexit and immigration. After excluding participants who responded incorrectly to the attention check (n = 17), the sample consisted of 155 participants aged M = 30.17 (SD = 9.99). They declared their national identity as either British (n = 122), English (n = 16), Scottish (n = 8), Northern Irish (n = 5) or Welsh (n = 3). One participant defined their national identity as other. Only those who voted in the EU referendum on 23 June 2016 were invited to participate as this ensured that participants were of British nationality. The majority of participants voted to remain in the EU (82.6%) with only 17.4% reporting voting to leave. Thus, the sample deviated from the official referendum result, 52% leave versus 48% remain, attracting more ‘remain’ voters.
Design and Procedures
The study, hosted online via Qualtrics, took around 15 minutes to complete. Participants were assigned to read one of the two articles on the status of EU immigrants in the UK, describing them as either (1) comparable to the status of a UK citizen, or (2) of a lower status.
EU citizen status manipulation
Participants read a passage on the policy regarding the EU immigrants in the UK post-Brexit (see supplementary materials for details). In one condition, EU citizens currently residing in the UK were portrayed as having lower salaries and worse living conditions (lower outgroup status), whereas in another condition, having comparable salaries and living conditions (equal outgroup status).
Manipulation check
Participants were presented with a social ladder and asked to indicate on which rung they thought 1) they stand, 2) a typical British person stands, and 3) a typical EU citizen living in the UK stands. The perceived status of typical British person and typical EU citizen living in the UK were compared to evaluate whether they were in line with the outgroup status condition participants were assigned to.
Voice expectation
Participants were asked to what extent ingroup and outgroup members should be consulted by the government to assist with policymaking. Specifically, we emphasised that the present research was concerned with what participants deemed the fairest way of dealing with this issue. This involved responding to six items (three for each side) developed by the authors such as ‘The British government should consult British members of the general public’ (ingroup voice item, α = .96) or ‘The British government should consult EU citizens living in British cities’ (outgroup voice item, α = .98) on a seven-point Likert scale (1 = strongly disagree; 7 = strongly agree).
Outcome expectation
Participants were asked what the fairest decision British government could make regarding the immigration policy post-Brexit. Participants responded on an eight-point scale (1 = EU citizens who have arrived in the UK prior to Brexit should undergo scrutiny to assess whether they should continue to stay or not; 8 = All EU citizens who have arrived in the UK prior to Brexit should be allowed to stay regardless of their prior contributions.). Therefore, higher scores indicated a more outgroup favouring decision. Given the immigration policy context of the study, it was only possible to measure outcome expectation for the outgroup and not for the ingroup.
Social identification strength
Participants indicated their identification with the British people by responding to four items such as ‘I am glad to be British’ or ‘I identify myself with other British people’ on a seven-point scale (1 = strongly disagree to 7 = strongly agree. Items were adapted from Doosje, Ellemers, and Spears' (1995) social identification scale. Higher scores indicated higher identification strength (α = .89).
Results
Manipulation Check
To evaluate the effectiveness of the manipulation, we conducted a mixed ANOVA with outgroup status condition as a between-group factor and relative positioning of a typical British person and an EU citizen living in the UK on the ladder as within group factor. There was a significant Status Condition*Target Group interaction on perceived status F(1, 152) = 35.75, p < .001, η2p = .19. To investigate this interaction, four t-tests were conducted with a Bonferroni-corrected alpha level, p = .013. In line with the manipulation, the difference between status conditions on British citizens’ status was non-significant, t(152) = 2.08, p = .039, whereas participants in the lower outgroup status condition perceived EU citizens as having lower status (M = 3.79, SD = 1.35) than participants in the equal status condition (M = 4.51, SD = 1.08), t(152) = 3.61, p < .001. Similarly, the difference in status between EU citizens and British citizens in the equal status condition was non-significant, t(72) = 2.28, p = .025, whereas in the low outgroup status condition, British citizens were judged as higher in social standing (M = 5.09, SD = 1.25) than EU citizens (M = 3.79, SD = 1.35), t(80) = 8.79, p < .001. To summarise, EU citizens in the lower outgroup status condition were perceived as having lower status in comparison to the ingroup, while the perceived status of the EU citizens in the equal status condition did not differ from the perceived status of the ingroup. Our manipulation of relative status was therefore successful. Furthermore, participants identified with a typical British citizen. Mean identification strength was 5.00 (SD = 1.38) on a seven-point scale.
The Effect of Outgroup Status on Expected Fairness
Voice expectation
A mixed ANOVA was conducted to evaluate the effect of outgroup status condition (between-subjects factor) on the voice expectation for ingroup versus outgroup as a within-groups factor (within-subjects factor). Both main effects and the interaction were non-significant (Fs < 3.71, ps ≥ .06), indicating that people did not expect authorities to give more voice to their own group in comparison to the outgroup. This was inconsistent with our hypotheses which stated that people would expect more voice for an ingroup than outgroup and equal status outgroup more than lower status outgroup.
Outcome expectation
A one-way ANOVA on the effect of outgroup status on outcome expectation for EU citizens living in the UK was not significant, F(1, 153) = .62, p = .434 (see Table 1). That is, in contrast to our predictions, outgroup status did not matter for the expectations of outcome favourability granted to the outgroup by the authorities in the context of decisions made by the British government regarding the immigration policy post-Brexit.
Table 1
Mean Score Differences in Lower Outgroup Status Versus Equal Outgroup Status Conditions in the Key Measured Variables (Studies 1-3)
| Status condition | Study 1
|
Study 2
|
Study 3
|
||||||
|---|---|---|---|---|---|---|---|---|---|
| N | M | SD | N | M | SD | N | M | SD | |
| Voice expectation (ingroup)a | |||||||||
| Lower status | 81 | 4.59 | 1.66 | 104 | 5.03 | 1.59 | 174 | 6.13 | 1.03 |
| Equal status | 74 | 5.13 | 1.62 | 98 | 4.89 | 1.65 | 173 | 6.15 | 1.03 |
| Voice expectation (outgroup)a | |||||||||
| Lower status | 84 | 4.45 | 1.87 | 104 | 4.96 | 1.69 | 174 | 5.81 | 1.28 |
| Equal status | 74 | 4.93 | 1.83 | 98 | 5.13 | 1.61 | 173 | 5.64 | 1.36 |
| Outcome expectationb | |||||||||
| Lower status | 81 | 6.84 | 1.74 | 104 | 5.38 | 2.34 | 174 | 5.32 | 1.81 |
| Equal status | 74 | 6.61 | 1.94 | 98 | 6.18 | 2.06 | 173 | 5.84 | 1.49 |
| Symbolic threata | |||||||||
| Lower status | - | - | - | 104 | 2.37 | 1.41 | 174 | 2.58 | 1.57 |
| Equal status | - | - | - | 98 | 1.59 | 1.04 | 173 | 1.78 | 1.06 |
| Realistic threata | |||||||||
| Lower status | - | - | - | 104 | 2.82 | 1.57 | 174 | 2.85 | 1.68 |
| Equal status | - | - | - | 98 | 1.78 | 1.20 | 173 | 2.00 | 1.18 |
ameasured on 1-7 scale. bmeasured on 1-8 scale.
*p < .05. **p < .01. ***p < .001.
Finally, zero-order correlations are shown in Table 2. Stronger British identification was associated with less favourable outcome expectations for the outgroup r(154) = -.29, p < .001 and higher expectations of ingroup voice, r(154) = .17, p = .030, but not expectations of outgroup voice, r(154) = -.08, p = .301. Thus, higher identification was related to higher voice expectations in line with Platow et al. (2015).
Table 2
Zero-Order Correlations in Study 1
| Variable | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|
| 1. Ingroup voice | – | .75*** | -.01 | .24* | -.06 |
| 2. Outgroup voice | .82*** | – | .20 | -.03 | -.22* |
| 3. Outcome expectation | .17 | .37** | – | -.24* | -.49*** |
| 4. Identification | .08 | -.17 | -.35** | – | .39*** |
| 5. Brexit vote | -.17 | -.33** | -.48*** | .37** | – |
Note. Brexit vote coded 0 = Remain, 1 = Leave. Values for the lower status outgroup condition are above the diagonal and the equal status outgroup condition are below. Full sample zero-order correlations are available in supplementary materials.
*p < .05. **p < .01. ***p < .001.
Study 2
Given the pattern of findings observed in Study 1, it is possible that EU citizens in the equal status condition may have been perceived as a minority that poses a potential threat to British resources (Stephan & Stephan, 2000). Having not manipulated their status as such, EU citizens being a minority in the UK would have been more plausibly characterised as having a lower social status than the British. It is likely that by manipulating EU citizens status to be equal to that of a typical British person may have questioned the stability of outgroup status, thus potentially indicating a competitive threat. For this reason, instead of manipulating the status of EU citizens overall, in Study 2, we chose EU national groups that tend to be of relatively lower and higher status. We also added a measure of threat to ensure that the equal status outgroup, in this case, is not interpreted as more competitively threatening.
Perceiving another group as threatening may have profound consequences for how people may expect authorities to act towards them. For example, people tend to have aversive feelings towards threatening groups, especially those of lower status (Riek, Mania, & Gaertner, 2006). Thus, they may be less sympathetic to authority treating them fairly and may even expect authorities to deny them voice or restrict favourable outcomes. Indeed, Gerber and Jackson (2016) found that people who support group-based dominance (and believe that it is natural that some groups belong on the bottom and some on the top of the hierarchy) are in favour of excessive police physical violence as a way of granting compliance with hierarchy-threatening citizens, for example, non-violent demonstrators. As such, people may expect authorities to deal with groups that are perceived as threatening by restricting their resources or freedom to express their opinions. For this reason, Study 2 employed measures of intergroup threat.
The aim of Study 2 was to further test our hypotheses relating to the impact of outgroup status on expectations of authority fairness (in terms of their voice and outcomes) in a more status-static setting as well as to control for the impact of threat perception in expectations of authority fairness. We predicted that intergroup threat perceptions will be relevant to expectations of authority fairness. Two types of threat are distinguished in the literature (Riek et al., 2006). Other groups can be perceived as threatening because they are in direct competition for some resources, whether they are more tangible (such as jobs) or less tangible (such as political power). However, even if there is no direct rivalry between groups, threat can still be perceived if other groups’ values are seen as drastically different. These two types of threats are referred to as realistic and symbolic threat, respectively. In terms of expectations of fairness, in line with our previous hypotheses, we predicted that outgroup status will influence expectations of authority voice and outcomes. However, perceptions of symbolic and realistic threat may also be an important element in expectations of authority fairness, both in terms of voice expectation and outcome expectation. In both cases, increased perceptions of threat would have a negative impact on expected fairness. In other words, the more threatening the outgroup is perceived to be, the less fairly people would expect authorities to treat them. As ingroup identification strength is considered to be an antecedent of intergroup threat (Riek et al., 2006), we entered this variable as a predictor to control for its effects.
Method
The study design was largely similar to that of Study 1, utilising a different outgroup status manipulation. Furthermore, because the snowballing technique utilised in Study 1 attracted more remain voters than leave voters, Study 2 made use of a recruitment tool to recruit a sample that was more representative of the British population in terms of the referendum vote. The UK participants (N = 202), aged M = 35.43, SD = 11.93, were 63% Remain voters, 24% Leave voters and 13% did not voteii. Declared national identity was either British (n = 140), English (n = 33), Scottish (n = 18), Welsh (n = 6), or Northern Irish (n = 2). Three participants defined their national identity as other. They were recruited from Prolific and paid £1.25 for completing a 15-minute survey. Prolific was selected because it tends to produce high quality and reliable data (Peer, Brandimarte, Samat, & Acquisti, 2017).
Participants read a similar scenario as in Study 1, whereby they were told about the implications of Brexit on immigration policy, this time focusing on specific groups of EU citizens (see supplementary materials for details). Germany and Romania were chosen as the equal and low status outgroup respectively because of similar proportions of those nationals currently living in the UK (286,000 for Germany and 220,000 for Romania). Romania was also picked as low status because of the restrictions imposed on their citizens’ ability to move to the UK and because they only recently became an EU member state.
Participants completed an attention check, and the same set of dependent variables as in Study 1 (outcomes, voice) in relation to either the low or equal status country as well as a measure of British identification (α = .91). Participants also completed symbolic and realistic threat measures in relation to the country they were assigned to. These included responding to items adapted from Velasco González et al. (2008) such as ‘Germans/Romanians are a threat to the British culture’ (symbolic threat) and ‘Because of the presence of Germans/ Romanians, unemployment in the UK increased’ (realistic threat) on a seven-point scale (1 = strongly disagree; 7 = strongly agree). A higher score indicated higher perceptions of threat (α = .95 for each subscale).
Results
Manipulation Check
A mixed ANOVA was conducted with the status condition as a between-group factor and positioning of the ingroup and outgroup on the ladder as a within-group factor on perceived status. There was a significant Status Condition*Target Group interaction on perceived status, F(1, 200) = 94.36, p < .001, η2p = .32. In line with the manipulation, in the low status condition, the Romanian citizens in the UK were perceived as lower status (M = 3.73, SD = 1.51) than the ingroup (M = 5.67, SD = 1.51), t(103) = 12.06. p < .001, d = 1.29. On the other hand, German citizens living in the UK were perceived to be of similar status to the ingroup, t(97) = .47, p = .639. Therefore, the new method of manipulating outgroup status worked as intended. As with Study 1, participants identified with a typical British citizen. Mean identification strength was 5.21 (SD = 1.31) on a seven-point scale.
The Effect of Outgroup Status on Fairness Expectation and Intergroup Threat
A between-groups multivariate one-way ANOVA was conducted to evaluate the effects of outgroup status on expectations of authority voice and outcomes as well as intergroup threat (see Table 1 for descriptive statistics). As in Study 1, outgroup status had no effect on the expectations of voice for the ingroup, F(1, 201) = .35, p = .557, nor for the outgroup, F(1, 201) = .48, p = .490. The lower status group, however, was expected to be given less favourable outcomes from the authority (M = 5.38, SD = 2.34) than the equal status outgroup (M = 6.18, SD = 2.06), F(1, 201) = 6.59, p = .011, η2p = .03. This was in line with our hypotheses.
Furthermore, the perceptions of symbolic threat were higher in the lower status outgroup condition (M = 2.37, SD = 1.41) than in the equal status outgroup condition (M = 1.59, SD = 1.04), F(1, 201) = 19.69, p < .001, η2p = .09. The same pattern occurred for perceptions of realistic threat (lower status outgroup M = 2.82, SD = 1.57; equal status outgroup M = 1.78, SD = 1.20), F(1, 201) = 27.88, p < .001, η2p = .12.
Ingroup Identification Strength
Zero-order correlations are displayed in Table 3. This time, British identification strength did not correlate with any dependent variables relating to the expectations of fairness in contrast to Study 1.
Table 3
Zero-Order Correlations in Study 2
| Variable | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
|---|---|---|---|---|---|---|---|
| 1. Ingroup voice | – | .24* | -.10 | .04 | .26** | .22* | .06 |
| 2. Outgroup voice | .39*** | – | .40*** | -.20* | -.43*** | -.36*** | -.35*** |
| 3. Outcome expectation | -.12 | .34*** | – | -.29** | -.63*** | -.57*** | -.60*** |
| 4. Identification | .06 | -.02 | .01 | – | .28** | .30** | .32** |
| 5. Symbolic threat | -.01 | -.27** | -.32** | .06 | – | .75*** | .55*** |
| 6. Realistic threat | .07 | -.26** | -.36*** | -.07 | .82*** | – | .53*** |
| 7. Brexit vote | .09 | -.27 | -.47*** | -.01 | .39*** | .55*** | – |
Note. Brexit vote coded 0 = Remain, 1 = Leave. Values for the lower status outgroup condition are above the diagonal and the equal status outgroup condition are below. Full sample zero-order correlations are available in the supplementary materials.
*p < .05. **p < .01. ***p < .001.
Exploring the Role of Intergroup Threat
The last set of analyses explored whether perceptions of outgroup threat predict expectations of authority fairness over the effects of the outgroup’s status. Two multiple regression analyses were carried out with the status condition, identification strength, symbolic and realistic threat entered as predictors. The first model explained 12% of the variance, F(4, 201) = 7.85, p < .001. The status condition, identification strength, and realistic threat were non-significant predictors of outgroup voice expectation (see Table 4). Higher symbolic threat, however, predicted lower voice expectation for the outgroup, β = -.29, t = -2.63, p = .009. Thus, symbolic threat was more relevant to the levels of voice expectations than was outgroup status.
Table 4
Regression Models Predicting Authority Fairness Expectation
| Predictor variable | Study 2
|
Study 3
|
||||||
|---|---|---|---|---|---|---|---|---|
| Voice expectation (outgroup)
|
Outcome expectation
|
Voice expectation (outgroup)
|
Outcome expectation
|
|||||
| β | t | β | t | β | t | β | t | |
| Status condition | -.07 | -.97 | < -.01 | < -.01 | -.17 | 3.15*** | .04 | .76 |
| British identification | -.05 | -.73 | -.06 | -1.01 | < .01 | .01 | .07 | 1.43 |
| Symbolic threat | -.29 | -2.62** | -.33 | -3.31** | -.36 | -3.52*** | -.43 | -4.34*** |
| Realistic threat | -.10 | -.90 | -.25 | -2.47* | < -.01 | < -.01 | <-.01 | -.07 |
| Adjusted R2 | 12.0% | 29.7% | 11.3% | 17.3% | ||||
Note. Status condition: 0 = lower status condition, 1 = equal status condition.
The second model had the same predictor variables with outcome expectation as the dependent variable and it explained 30% of the variance, F(4, 201) = 22.19, p < .001. As with the previous model, status condition and identification strength were non-significant predictors of outcome expectation (see Table 4). Symbolic threat, β = -.33, t = -3.31, p = .001, and also realistic threat this time, β = -.25, t = -2.47, p = .014, were significant predictors of outcome expectation. In both cases, higher perception of threat predicted lowered expectation of outcome favourability. Therefore, realistic threat posed by the outgroup mattered only when judging expectations in relation to the outcomes, whereas symbolic threat was important for both voice and outcome expectations.
The threat posed by the outgroup as opposed to the mere status may therefore better explain expectations of authority fairness. British people who perceived the EU citizens as more competitively threatening expected the government to give them less voice and less favourable outcomes in their decision-making processes. This model was stronger for outcome expectations than for voice expectations.
Study 3
Study 3 was carried out to replicate the findings of Study 2 in the context of a local authority decision adjudicating between ingroup and outgroup demands. The same outgroup status manipulation was used as in Study 2 by varying whether the information provided regarded an outgroup of a relatively equal status to the British people (German citizens in the UK) or of a relatively lower status (Romanian citizens in the UK). According to an a-priori power calculation, 346 participants were needed to detect a small to medium effect size (d = .35) in a two-tailed t-test (alpha set at 0.05 and power at 90%). Data from 359 participants was collected and after excluding participants who failed the attention check (n = 12), the final sample consisted of 347 participants. They were Prolific users who were currently residing in, and were nationals of, the UK, aged between 18 and 75 years old (M = 37.61, SD = 12.76). They declared their national identity as either British (n = 240), English (n = 66), Scottish (n = 21), Northern Irish (n = 9) or Welsh (n = 6). Six participants defined their national identity as ‘other’. In terms of gender, there were 65% females and 35% males with one participant defining their gender as other. They were paid £0.90 for a ten-minute survey.
Participants were told that a centre promoting the culture of the outgroup they were assigned to proposed to organise a festival promoting their traditions. The festival was described as open to everyone and taking place on the streets of a city in the UK, which caused the local British residents to be dissatisfied with the potential noise and disruption to their lives. It was suggested that the local council should resolve the matter regarding whether the festival should take place or not. We used the same attention check as in Study 2. As with the previous studies, voice (ingroup and outgroup) and outcome expectations were measured. Participants responded to two single-item statements: “The local city council should consult the Romanian/German people interested to partake in the event” (outgroup voice), “The local city council should consult the local British residents” (ingroup voice) on a seven-point scale (1 = strongly disagree, 7 = strongly agree). For outcome expectation, on an eight-point scale, participants indicated the extent to which the festival should not go ahead in that area (1) or should go ahead in the area (8). Finally, participants completed the same scales measuring symbolic threat, realistic threat and British identity (α = .95, .94, .90 respectively) and indicated the position of the typical Romanian/German citizen living in the UK and the typical British citizen on a social ladder as a manipulation check.
We preregistered the hypotheses before data collection began (see https://osf.io/47tcs/ for preregistration form). Following our findings so far, we predicted that (1) Outgroup status will have a non-significant effect on voice expectation, but (2) a significant effect on outcome expectation with the equal status outgroup judged as expected to receive more favouring outcomes than the lower status outgroup; (3) increased perceptions of realistic and symbolic threat will be related to decreased voice and outcome expectations; (4) higher symbolic threat will be negatively related to voice expectations whereas higher realistic threat will be negatively related to outcome expectations; and (5) the effects of perceived symbolic and realistic threat will override any effects of outgroup status on voice expectation and outcome expectation.
Results
Manipulation Check
To verify that our manipulation worked, a mixed ANOVA was conducted with the status condition as a between-group factor and positioning of the ingroup and outgroup on the ladder as a within-group factor on perceived status. There was a significant Status Condition*Target Group interaction on perceived status, F(1, 345) = 86.59, p < .001, η2 = .06; in the lower status condition, the Romanian citizens in the UK were perceived as lower status (M = 4.27, SD = 1.56) than the ingroup (M = 5.68, SD = 1.33), t(173) = 11.97. p < .001, d = .91. On the other hand, German citizens living in the UK were perceived to be of similar status to the ingroup, t(172) = -.12, p = .906.
The Effect of Outgroup Status on Fairness Expectation
In line with our pre-registration plan, we conducted a series of independent t-tests to test the effect of outgroup status on voice expectations and outcome expectations (see Table 2). In line with our pre-registered predictions, outgroup status did not significantly affect ingroup voice expectations, t(345) = 0.16, p = .871, or outgroup voice expectations, t(343.5) = -1.19, p = .236, in a two-tailed test. For the effect of intergroup status on outcome expectations, we preregistered this hypothesis as one-tailed, predicting to replicate the results of Study 2 that lower status outgroup would be expected to be given less favourable outcomes from the authority than equal status group. The data confirms this hypothesis, t(333.43) = 2.93, p = .002, d = .31 (see Table 2 for descriptive statistics). Therefore, our pre-registered study further supported the previously hypothesised effect of outgroup status on outcome expectation, while providing little evidence that outgroup status determines voice expectations for ingroup or outgroup.
Intergroup Threat in Fairness Expectation
To provide a further test of the hypotheses related to the role of intergroup threat, which we explored in Study 2, we pre-registered hypotheses regarding the role of intergroup threat in fairness expectations. Both symbolic and realistic threat were found to be negatively related to outgroup voice expectations and outcome expectations, but not with ingroup voice expectations (see Table 5). Following this, we entered status condition, British identification, symbolic threat and realistic threat into a multiple regression model.iii Two models, one predicting voice expectation for outgroup and another predicting outcome expectation for outgroup were of good fit, F(4, 342) = 11.93, p < .001, R2 = .11 and F(4, 342) = 19.10, p < .001, R2 = .17, respectively. Symbolic threat but not realistic threat predicted both voice expectation and outcome expectation (see Table 4). Thus, perceiving the outgroup as higher in symbolic threat was related to lower voice and outcome expectation for the outgroup, while the role of the realistic threat diminished in this context. Consequently, we did not find support for the hypothesis that realistic threat would be more important for outcome expectations. This suggests that perceiving other groups as holding values different to that of the ingroup is related to lower expectations that authorities should provide them an opportunity to voice their opinion or make decisions that are favourable to them. Moreover, when controlling for threat, while the role of outgroup status diminished in the outcome expectation model, outgroup status remained a weak predictor of voice expectation despite previous analyses pointing to a non-significant difference between lower and equal status conditions.
Table 5
Zero-Order Correlations in Study 3
| Variable | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|
| 1. Ingroup voice | – | .24** | -.12 | .16* | .01 | -.01 |
| 2. Outgroup voice | .43*** | – | .53*** | -.16* | -.48*** | -.44*** |
| 3. Outcome expectation | .07 | .35*** | – | -.03 | -.50*** | -.43*** |
| 4. Identification | -.04 | .04 | .05 | – | .29** | .26*** |
| 5. Symbolic threat | -.13 | -.17* | -.22** | .08 | – | .87*** |
| 6. Realistic threat | -.07 | -.11 | -.17* | .12 | .82*** | – |
Note. Values for the lower status outgroup condition are above the diagonal and the equal status outgroup condition are below. Full sample zero-order correlations are available in the analysis code for Study 3 on the Open Science Framework project page.
*p < .05. **p < .01. ***p < .001.
Additional Analyses
While we did not preregister the hypothesis regarding the relationship between British identification and ingroup voice, these variables were not significantly related to one another, r(346) = .06, p = .264. This is inconsistent with Study 1 findings and Platow et al. (2015), but consistent with Study 2.
Internal Meta-Analysis
The three studies presented (N = 704) were meta-analysed using fixed effects in which the mean effect size was weighted by sample size. Cohen’s d effects from each study were combined and converted into a single z-score. Two separate analyses were conducted to test our core hypotheses regarding the role of outgroup status (lower status vs equal status) on fairness expactations (voice expectation and outcome expectation respectively). Data was aggregated using Goh, Hall, and Rosenthal’s tool (2016) tool for internal meta-analyses and then entered into Jamovi MAJOR module to create a forrest plot (jamovi project, 2018; see Figure 1). The mean effect of outgroup status on authority voice expectation was non-significant, d = -.02, z = -.31, p = .757, 95% CI [-.17, .12]. Although it is statistically impossible to conclusively demonstrate absence of an effect, equivalence tests can be used to examine whether the observed effect is statistically smaller than a specified equivalence bound. We used Lakens’ (2017) equivalence test primer for meta-analysis to carry out this procedure setting equivalence bounds of d = -.20 (lower bound) and d = .20. (higher bound) for a small effect. The two one-sided tests indicated that the observed effect size (d = -.02) was significantly within the indicated equivalent bounds, Z = 2.36, p = .009. This means that given the effect size and standard error of the outgroup status on voice expectation calculated from the current set of studies, we can reject the presence of even a small effect of d = .20.
Figure 1
Forest plot of standardised mean difference in (a) voice expectation and (b) outcome expectation as a function of outgroup status in a fixed-effects meta-analytic model.
In terms of authority outcome expectation, the meta-analysis revealed a small effect of outgroup status on authority outcome expectations, d = -.23, z = 3.04, p = .002, 95% CI [-.38, -.08]. Participants expected authorities to distribute less favourable outcomes to lower status outgroups than to equal status outgroups.
General Discussion
Three experiments set out to test whether groups expect authorities to be fair to all groups, independently of their social status, by providing them with a voice to communicate their opinions and by making favourable decisions. In doing so, we addressed previously untested assumptions associated with the relational models of PJ, which state that people derive their sense of social self-esteem from how they are treated by authorities (Lind & Tyler, 1988; Tyler et al., 1996; Tyler & Lind, 1992). The present research reverses this question and examines whether the social standing can also inform expectations of authority fairness, how authorities should act.
Voice Expectation
The results provide little empirical support for the hypothesis that expectations of fairness are driven by social standing. In other words, just because a group is of lower social standing, people do not necessarily think authorities should provide them with fewer opportunities to voice their opinions than groups of status equal to that of the ingroup. In general, then, we found no significant differences in how people expect authorities to treat groups as a function of their social standing. The equivalence test of the meta-analytic effect further suggests that, based on the effect sizes obtained in the present three studies, we can reject the presence of even a small effect of outgroup status on voice expectation. Our findings may suggest that people do not necessarily expect authorities to consult some groups more than others based on their relative status in the intergroup contexts we studied, but this does not mean that voice is not desirable or that authorities do not actually give preferential treatment to some groups over other groups.
One could argue that because voice and fair treatment are symbolic, groups cannot tangibly elevate or decrease their social status based on fair treatment in the way that resources can afford to do and thus, groups expect authorities to treat both lower and equal status groups in the similar way. Our findings are in line with Huo and colleagues (1995; as cited in Tyler, 1997) who showed that people are unlikely to deny treatment as a human being to even outgroups. We replicated this finding in the context of how people expect authorities to act. However, it must be stressed that the authors of the relational models of PJ never indicated that the link between PJ and status would work in reverse, so the present findings do not contest any elements of the theory but rather extend them. As such, being treated fairly may elevate one’s status but having higher status in itself does not appear to be the grounds for fairer treatment.
Extending the findings of Platow et al. (2015), we do not find strong evidence that social identity is driving the expectations of fairness, despite the decisions studied in our chosen contexts being relevant to the British identity. Only Study 1 revealed some correlations between social identification strength and expectations of voice for the ingroup with those who identify more strongly as British expecting the government to allow them more voice, which is in line with the original findings. The same relationship in Studies 2 and 3 was non-significant. Thus, the link between social identification and voice expectation requires more empirical evidence to uncover the mechanisms underlying (the absence of) this effect.
Outcome Expectation
People’s expectations of authority distribution of outcomes tell a different story. In Studies 2 and 3, more favourable outcomes were expected to be distributed to the relatively equal status outgroup in comparison to a lower status outgroup (small-to-medium effect). In Study 1, on the other hand, the outcome expectation did not differ between the lower and equal status groups. The internal meta-analysis combined these and showed that outgroup status has a small effect on outcome expectations. This demonstrates that the way authorities are expected to make decisions about other groups varies depending on group status, mirroring our original hypotheses regarding the importance of resources in acting as a leverage to relative status. While there is an assumption that authorities simply follow the established laws and protocols in the way they make their decisions, the way people reason about what it means for authority to make decisions so that outcomes are fair are dependent on who the receiver is in relation to the ingroup. Therefore, in line with Leung et al. (2007), the present research suggests that groups have more nuanced expectations of the authority fairness when it comes to resources than to how others are treated.
Parallel analyses also demonstrated that the lower status groups tended to be characterised as more threatening to the ingroup. While realistic threat and the competition over physical resources appear to play a role, the evidence regarding symbolic threat is more extensive, at least in the studied context. When the values of the outgroup were perceived to be in conflict with those of the ingroup, people were more likely to expect authorities to distribute outcomes in less favouring ways to those groups than to groups that were perceived as less threatening. In cases where other groups’ values are perceived to be critically incompatible with those who the authority represent, people may expect authorities not to represent them to the same degree by not favouring them. While we did not investigate why symbolic threat may be more important, perceptions of who do the authorities serve and who they do not serve may be contingent on the perceptions of the values they share with those who they represent (e.g., see Tyler & Jackson, 2013).
In line with Riek et al. (2006), it was the lower status groups that were found to be more threatening and simultaneously expected to receive less voice and less favourable resources from the authority. Feeling threatened by outgroups creates an expectation that authorities should treat them distinctly to the ingroup, while those who do not feel threatened to the same degree may not have those expectations. Consequently, these divergent expectations may put an extra strain on the ways in which authorities think they should conduct themselves in relation to the citizens.
Expecting Authority Fairness in Intergroup Settings
Findings relating to the negative perceptions of intergroup threat are relevant for our understanding of reactions to authority decisions. Perceptions of other groups clearly shape expectations of fairness, which is problematic for authorities who may want to come across as fair to everyone to maintain their legitimacy (Tyler, 2006). Those who disagree regarding the perceived threat of other groups, biases aside, actually expect authorities to act towards them in rather divergent ways, which may mean that the same decisions will be perceived as less or more fair on the basis of the perceptions of intergroup relations. Therefore, in line with Platow et al. (2015), the results across all studies consistently suggest that expectations of fairness are not universal. Considering how these threat perceptions impact evaluations of fairness is important because it recognises that even though authorities may be motivated to act in fair ways, their actions may not always be received as such.
The present research is among the first to examine authority decisions that are intergroup in nature (for an exception, see Peate, Platow, & Eggins, 2008). In doing so, the role of authority is conceptualised as extending beyond dealing with individuals in intragroup settings. Switching to the intergroup analysis requires recognising that the way social groups perceive authority decision is more complex than the sum of individual members’ perceptions of the decision. In this context especially, holding authority may also involve making decisions about groups and managing the expectations from different corners of the societyiv. The current literature does not address this function of authorities adequately and requires more attention. These questions are particularly relevant for countries that suffer social unrest and post-conflict societies such as Northern Ireland or South Africa where the legitimacy of authorities needs to be re-established parallel to peacebuilding. It is in these contexts that other groups may be perceived as the most ideologically threatening, which may be associated with more divergent expectations of authorities’ actions.
Limitations
Although investigating novel assumptions, the present research has some shortcomings. Firstly, despite manipulation checks pointing to equal and lower status outgroups being perceived as such across all studies, it is possible that in the case of Study 1, the equal status group was more threatening, perhaps on a comparable level to that of the lower status group in Studies 2 and 3. Measuring intergroup threat in Study 1 would have had potentially enabled us to distinguish between status and threat effects, but we did not consider to examine threat perceptions at that stage. Even if it is the case that the equal status outgroup was more threatening, the stability of their social standing was remarkably different. In Study 1, the equal status condition might have been interpreted like they have gained status from a generally-assumed low status position of immigrant groups, whereby the group status conditions in Studies 2 and 3 were relatively fixed and stable. Either way, outgroup status and intergroup threat are interrelated, but the specific effects may depend on the nature of the intergroup context itself.
Relatedly, we did not measure whether lower status outgroups were perceived with less respect than equal status outgroups, which is how relational models of PJ usually define social standing. Relational models typically discuss the extent to which authorities’ procedural fairness increases one’s feelings of respect and pride in their identity. It is a theoretical limitation of the present research that we did not measure or manipulate respect to evaluate whether groups that are more respected would indeed be expected to receive more fair treatment. As the present research operationalises social standing largely in terms of economic status (and it is indeed one way in which groups can be placed in the social hierarchy), our conclusions are limited as they can only speak to the effects of intergroup economic status on expectations of fairness. Future research should consider whether manipulation of intergroup respect would replicate these results.
Finally, the findings relating to voice expectation in the present studies may be reflective of a floor effect. For example, processes underpinning group dehumanisation, whether in relation to race, gender or disability, are characterised by denying human qualities to others (Haslam, 2006). Voice, being a recognition of one’s higher cognition abilities – that one wishes to express their opinion about something – may be one of the aspects of human experience that becomes denied for dehumanised groups. Thus, testing these hypotheses in the context of groups that are derogated would be beneficial to further examine how fairness may not always be expected of authorities. Another explanation for a possible floor effect is that the present set of studies was carried out in the UK, a Western individualistic country where the distance between superiors and subordinates tends to be generally lower (Hofstede, 2001). In those cultural contexts, people are more likely to react against a lack of voice than in other cultures where the power distance is larger (Brockner et al., 2001). Therefore, voice expectation may be more of a norm in Western cultures, independently of the relative status of groups. Whether outgroup status could determine how much voice groups are expected to receive in other cultural context is a question future research could address.
Conclusion
Across three studies, we sought to investigate whether authority fairness is universally pursued for everyone, independently of their social standing. Although in most cases an equal level of authority fairness and voice provision was expected, social groups that were perceived to be threatening to the values and culture may be expected to be granted less favourable outcomes by authorities. Social standing, however, played a role in determining outcomes that authorities should distribute to those groups: lower status groups were expected to be admitted with less favourable outcomes than equal status outgroups. Such research not only informs us of how groups perceive fairness but also potentially enables predictions about patterns of dissatisfaction with authority decisions. A thorough investigation of the routes to disputing authority fairness may lead to better understanding how intergroup-level authority fairness can be enhanced.