This is an open access article distributed under the terms of the Creative Commons
Attribution License (The widely-cited definition of collective action (Wright, Taylor, & Moghaddam, 1990) states that “a group member engages in collective action any time that he or she is acting as a representative of the group and where the action is directed at improving the conditions of the group as a whole” (p. 995). Although this definition is broad and applies to multiple groups, collective action researchers usually implicitly assume that the social protest aims at reducing social inequality in society and is thus associated with liberal movements (cf. Levenson & Miller, 1976). For instance, established theoretical models explaining collective action (e.g., dual pathway models; Becker & Tausch, 2015; Stürmer & Simon, 2004; Tausch et al., 2011; van Zomeren, Postmes, & Spears, 2008; van Zomeren, Spears, Fischer, & Leach, 2004) have been developed using relatively liberal ideas of injustice perceptions and focused mainly on “progressive” collective action (defined as reducing social inequality) in response to collective disadvantage. Consequently, relatively little is known about collective action among Conservatives (for some recent exceptions, see Jost, Becker, Osborne, & Badaan, 2017; Osborne, Jost, Becker, Badaan, & Sibley, 2019). The established models predicting collective action illustrated that “progressive” protest occurs when people perceive that the equality or need rule of fairness is violated. We argue, however, that these perceptions of injustice cannot explain the occurrence of conservative social protests. In the present work we, therefore, focus on injustice appraisals as a key antecedent of collective action (e.g., Runciman, 1966; van Zomeren, Postmes, & Spears, 2008; Walker & Smith, 2002) and test whether the same or different conceptualizations of injustice instigate protest among Liberals and Conservatives. By examining collective action intentions of individuals across the political spectrum, we go beyond prior work and account for ideological preferences for different fairness rules and contribute to the recent literature calling for including political ideology in models of collective action (e.g., Jost et al., 2017; Osborne et al., 2019).
Collective Action Among Conservatives
In the current article, we conceive of political ideology as an interrelated set of moral and political attitudes which “organizes people’s values and beliefs and leads to political behaviour” (Jost, 2006, p. 653). Specifically, we focus on the left/liberal and right/conservative distinction which has proven to be a useful and parsimonious way of classifying political attitudes (e.g., Carney, Jost, Gosling, & Potter, 2008; Jost, Glaser, Kruglanski, & Sulloway, 2003).
The core of conservative (versus liberal) political ideology is centred on two aspects: attitudes toward social stability (versus change) and attitudes toward social hierarchy (versus equality). Both are motivated by a common set of social, cognitive and motivational factors (e.g., Jost et al., 2003; Rossiter, 1968). In line with these assumptions, previous studies focused on the role of conservative ideology for system maintenance, rather than system change (e.g., Cameron & Nickerson, 2009; Feygina, Jost, & Goldsmith, 2010; Jost et al., 2003, 2017; Jost et al., 2012; Osborne et al., 2019; Wakslak, Jost, Tyler, & Chen, 2007). Empirical evidence indicates that, among others, Conservatives are more likely to think the socioeconomic system is fair (e.g., Jost et al., 2003), and are thus less likely to acknowledge the pervasiveness of discrimination. Accordingly, they are less willing to engage in collective action that aims to change the unequal status quo (Jost et al., 2017).
Yet, observations from the social world, such as anti-immigrant movements and ongoing resistance to taxes and government-sponsored programs indicate that Conservatives are not necessarily inactive citizens, but often actively engage in social protest (e.g., Hennes, Nam, Stern, & Jost, 2012; Lo, 1982; van der Brug, Fennema, & Tillie, 2000). Indeed, studies show that even individuals high in system justification are willing to protest if the protest is framed as preserving the status quo and not as promoting social change (Feygina et al., 2010; Osborne et al., 2019). Similarly, economic Conservatives (as approximated by their high scores on social dominance orientation) support affirmative action if it maintains social hierarchies (Chow, Lowery, & Hogan, 2013). Those high in economic system justification, although unlikely to support the hierarchy challenging Occupy Wall Street movement, were more likely to support the economically conservative Tea Party movement (Hennes et al., 2012). These results suggest that such “system-sanctioned changes” may trigger collective action intentions among those who strongly support the system.
More generally, studies so far support the notion that both Liberals and Conservatives are likely to display similar social protest behaviours. What distinguishes them, however, are the different underlying motivations concerning the preference for social stability (versus change) and preference for social hierarchy (versus equality).
In the current paper, we propose that another crucial difference between liberal and conservative collective action stems from a different understanding of what is (un)fair among those on different sides of the political spectrum. While both Liberals and Conservatives are likely to engage in social protest, they are likely motivated by different injustice appraisals. In the following paragraphs, we present a brief overview of the main fairness principles and explain why Liberals and Conservatives consider violations of different principles as unjust. Subsequently, we link the ideological differences in fairness rule preferences to the current models of collective action.
Social Fairness and Ideology
A crucial distinction in the fairness literature has been drawn between procedural justice (e.g., Lind & Tyler, 1988; Thibaut & Walker, 1975) and distributive justice (i.e., outcome fairness, e.g., Deutsch, 1975, 1985; Rawls, 1971). While procedural justice addresses the way justice is administered, distributive justice focuses on the ways, in which available resources and social goods are allocated across society. Distributive justice can be achieved through different allocation principles. The three most commonly distinguished are merit (or equity), equality, and need (e.g., Deutsch, 1985; Miller, 1999). The normative assumptions behind each of these three principles are as follows (for critical discussions, see e.g., Scott, Matland, Michelbach, & Bornstein, 2001): The merit rule states that individuals should be rewarded with and deserve levels of social goods corresponding to their contributions (conceptualized either as ability, effort or performance). The equality rule states that every individual should have and deserve the same level of social goods. Finally, the need rule states that individuals should have levels of social goods corresponding with their needs (Lamont & Favor, 2014).
Past correlational evidence indicates that Conservatives are more likely to allocate goods based on the merit principle, and less likely based on the equality and need principles. In contrast, Liberals are more likely to allocate goods based on the latter two principles (Mitchell, Tetlock, Mellers, & Ordóñez, 1993; Mitchell, Tetlock, Newman, & Lerner, 2003; Scott et al., 2001; Sidanius & Pratto, 2001; Skitka & Tetlock 1992, 1993; Tetlock & Mitchell, 1993). These preferences can be explained by the fact that Conservatives are more likely to endorse values of proportionality emphasizing personal independence, ambition, and economic individualism, which are reflected in the merit rule. Liberals, on the other hand, are more likely to endorse egalitarianism, which stresses the role of interpersonal interactions and mutual dependence and is reflected in a greater societal concern for equality and need (e.g., Mitchell et al., 1993; Rasinski, 1987).
In addition, Conservatives are less likely to endorse the need rule as they are more restrictive in helping, limiting their support to those who they consider not to be personally responsible for their predicament. When, in their opinion, the needy bring misfortune upon themselves, Conservatives punish them for breaching the societal rules of conduct (Skitka & Tetlock, 1992).
The Role of Fairness Perceptions in Collective Action
According to relative deprivation theory, judgments of fairness are central to whether people respond to collective disadvantage (Runciman, 1966; Walker & Smith, 2002). Consequently, perceptions of unfair or illegitimate collective disadvantage promote collective action (e.g., Wright, Taylor, & Moghaddam, 1990; van Zomeren et al., 2004). Although collective action models conceptualize perceptions of group-based injustice as one of three key predictors of social protest (e.g., van Zomeren et al., 2004; van Zomeren, Postmes, & Spears, 2008), they do not differentiate between various notions of fairness. For example, in the studies reviewed in the seminal SIMCA paper, collective action typically seeks to redress disadvantages created by the breach either of the equality (e.g., discrimination against homosexuals, women, or ethnic minorities), or the need rule (e.g., a newly imposed tax, a rise in tuition fees; van Zomeren et al., 2008). To the extent that collective action models assume such ‘liberal’ appraisals of injustice, they are likely to oversee the importance of injustice in instigating conservative collective action, and the prevalence of conservative collective action in general.
In the present research, we seek to expand the existing models of collective action by empirically investigating the role of political ideology in shaping injustice appraisals and subsequent social protest. We predict that, depending on the fairness rule that is violated—equality, need or merit—individuals holding different political attitudes will perceive the outcome as (un)just and, to the extent that injustice appraisals are one of the key triggers for collective action, will be motivated accordingly to engage in protest behaviour.
The Present Research
The aim of the present research is to show how conceptions of fairness among Liberals and Conservatives lead to divergent perceptions of injustice that, in turn, should predict their willingness to engage in collective action. Specifically, we predicted that Conservatives will engage in collective action when the merit principle is violated (H1), whereas Liberals will engage in collective action when the equality and need rules are violated (H2). To test these predictions, we conducted two studies focusing on the allocation of two different social goods: tax money (Study 1) and healthcare (Study 2). We did not expect a difference between the violation of equality and need principles. However, given that they are two different and well-established constructs, in Study 1 we kept them separate to explore possible differences. In both studies, liberal and conservative participants were confronted with a scenario in which one of the fairness rules was violated. We were interested in the extent to which they perceive the violation of this rule as unjust, and whether they are willing to engage in collective action against the proposed changes in the allocation of the social goods. Importantly, in each study, we used the same measure of collective action intentions both for Liberals and Conservatives. In both studies, we adopted a single-dimensional definition of political ideology and analyzed it as a continuous variable (see also e.g., Carney et al., 2008; Fuchs & Klingemann, 1990; Jost, 2006).
Study 1
Method
Participants and Procedure
A power analysis was performed with the G*Power 3.1.9.2 statistical package. The analysis revealed that approximately 88 participants were needed to detect a medium-size effect for interaction (f2 = .15) in a moderated multiple regression, with α = 0.05 and power at .90. A mixed sample of German undergraduate students and online users was recruited via university mail and invitation links posted on the online portal http://www.psychologie-heute.de. After excluding participants who did not complete the survey (n = 33) or provided the wrong answer to the attention check questions (n = 2), the sample consisted of N = 130 participants (59% female; Mage = 26.64; SD = 8.19). Of the final sample, 51 participants were exposed to the violation of merit condition, 42 to the violation of need, and 37 to the violation of equality.
In the invitation to the survey, participants were informed that the purpose of the study was to examine peoples’ beliefs about social issues and social justice. Participants answered the political ideology questions first. Next, they were randomly assigned to one of three conditions presenting a violation of the merit, need, or equality allocation principles. Following the manipulation, they were shown two attention check questions and were asked to answer a series of questions measuring perceived injustice and collective action intentions.
Materials
Political ideology
We used a single-item measure to assess political ideology (1 = right-wing; 7 = left-wing). This measure is commonly used in nation-wide social surveys in Germany to assess political ideology (e.g., GESIS, 2015), and its validity has been demonstrated in previous research (e.g., Fuchs & Klingemann, 1990)i.
Fairness manipulation
Fairness manipulation consisted of two parts. In the first part, participants were presented with an article, published allegedly in an online edition of a major German business newspaper “Wirtschaftswoche”, announcing that the tax reform in Germany is inevitable, and providing a brief description of the proposed changes (that would apply to income, real estate, and welfare tax; see Appendix for the English translation of the manipulation). Next, participants were asked to read an opinion of a Bundestag Committee Member that was voiced in a meeting devoted to the introduction of the proposed guidelines. In the provided transcript of his speech, the politician was voicing his concern that the proposed tax reform threatens social justice in Germany and was calling other politicians to oppose it. Depending on the experimental condition, when speaking of social justice, he was referring either to the merit, equality or the need rule of fairness:
[Merit rule] “In a fair social system, goods are allocated to people based on the contributions they have made to society. People who have contributed more, either by working harder, being more productive or having higher skills, should be given more in return. Similarly, people who have been lazy or unproductive should be given less or nothing.”
[Equality rule] “In a fair social system, goods are distributed equally among all people. Regardless of how large the contribution of each person is, everyone should get equal benefits from the available tax money. No one should be treated worse.“
[Need rule] “In a fair social system, goods are allocated to people based on their social needs. Those who are in need the most should be given priority. Fairness is achieved when people with the greatest needs are given attention first.”
To ensure that participants read the text thoroughly, following the manipulation they were asked to indicate whether the politician was for or against the proposed changes (all participants provided the correct answer) and to summarize the purported consequences of the reform, according to the politician (open-ended question).
Dependent variables
All items were rated on a 1 (definitely disagree) to 7 (definitely agree) scale.
Perceived injustice (four items; α = .90). Respondents were asked: „Thinking about the tax changes mentioned in the article and the opinion of the politician, how strongly do you agree with the following statements?”: “The introduction of the tax reform would be unfair”; “The tax changes would be socially unjust”; “The introduction of the tax reform would be illegitimate”; “The introduction of the tax changes would be justified”. Collective action intentions (six items; α = .91). Subsequently, respondents were asked: “Provided that the tax reform is debated in the Bundestag, how likely is it that you will participate in the following actions to oppose the tax changes?”: “participate in a discussion meeting”; “participate in a demonstration”; “participate in a rally”; “sign an online petition”; “send a letter of protest to the government”; “post flyers”.
In both studies, we also measured threat perceptions, national identity, group efficacy, and radical collective action intentions, for exploratory purposes. Interested readers can obtain detailed results from the first author. At the end of the survey, we additionally asked respondents to what degree each of the three allocation principles aligned most closely with their definition of social fairnessii. Conservative ideology was linked to higher preference for the merit-rule, r(109) = .51, p < .001, and weaker preference for the need rule, r(109) = -.34, p < .001 (the relationship with preference for the equality rule was significant only at the tendency level, r(109) = -.19, p < .052), confirming previous correlational evidence showing the link between political ideology and fairness allocations.
Results and Discussion
As expected, we did not find any differences in a violation of need vs. a violation of equality comparison for Liberals, nor for Conservatives. Thus, in the following analyses, we collapsed these two conditions into one category and, subsequently, compared it with the violation of merit condition (in all analyses the fairness rule was coded as 1 = violation of merit, 2 = violation of equality or need). Separate analyses computed for a violation of merit vs. a violation of equality and for a violation of merit vs. a violation of need comparisons revealed analogous results, thus are not reported below.
Table 1 displays the descriptive statistics and correlations among the study variables. Importantly, across the entire sample, political ideology was unrelated both to injustice appraisals and to collective action intentions.
Table 1
Descriptive Statistics and Zero-Order Correlations Between Measured Variables (Study 1)
| Variable | M | SD | 2 | 3 | 4 |
|---|---|---|---|---|---|
| 1. Political ideology | 5.09 | 1.16 | -.09 | .13 | .09 |
| 2. Injustice | 4.02 | 1.53 | – | .58** | .31** |
| 3. Collective action intentions | 2.81 | 1.56 | – | .25** | |
| 4. Violation of merit (1) vs. violation of need/equality (2) | – | – | – |
Note. Scale range for all continuous variables was 1-7. Lower values indicate a more conservative political ideology.
**p < .001.
Analytical Approach
For perceived injustice and collective action intentions, we ran a regression analysis with political ideology, the manipulated fairness rule (Step 1), and the interaction of both (Step 2) as independent variables using PROCESS macro (Model 1; Hayes, 2013). Results were computed using a heteroskedasticity-consistent standard error estimator (Hayes & Cai, 2007). For each analysis, we requested 10.000 bootstrap samples. We used the Johnson-Neyman’s technique (Hayes & Matthes, 2009), which identifies a range of values of a continuous moderator at which the conditional effect of an independent variable on the dependent variable is significant. Thus, in the text, we provide exact values of political ideology at which the effect of the fairness manipulation on the respective dependent variable was significant. We chose this approach over the pick-a-point approach, which is typically employed in the literature, to avoid sample-specific inferences (cf. Hayes, 2013): as political ideology was somewhat skewed in our sample (i.e., more participants self-defined as liberal than as conservative), values typically denoting low scores (i.e., one standard deviation below the mean) would have aligned very closely with the middle point on the scale (3.93 on a 1-7 scale).
Perceived Injustice
Analysis indicated no significant main effect of political ideology, B = -0.15, t(122) = -1.27, p = .205, 95% CI [-0.37, 0.08], sr2 = .01, a significant effect of the fairness rule, B = 0.99, t(122) = 3.66, p < .001, 95% CI [0.45, 1.52], sr2 = .10, and a significant interaction between political ideology and the fairness rule, B = 1.04, t(122) = 4.88, p < .001, 95% CI [0.62, 1.46], sr2 = .15.
As indicated by Johnson-Neyman’s significance regions (see Figure 1a), the interactive effect was significant for values of political ideology 3.46 and below (i.e., values 1, 2 and 3 on a 1-7 scale, denoting conservative participants; 12% of the sample) and values 4.61 and above (i.e., values 5, 6 and 7 on a 1-7 scale, denoting liberal participants; 71% of the sample). The effect was negative for conservative participants, indicating that Conservatives rated the proposed tax reform as less just when the merit rule was violated than when the equality or need rules were. In contrast, the effect was positive for liberal participants, indicating that Liberals felt more concerned when the equality or need rules were violated compared to the merit rule.
Figures 1
Conditional Effects of Fairness manipulation on a) the Perceived Injustice and b) Collective Action Intentions (Study 1).
Note. Positive effect denotes higher values in the violation of the equality/need rule condition, negative effect denotes higher values in the violation of the merit rule condition.
Collective Action Intentions
We found a corresponding pattern of results for collective action intentions. Analysis revealed no significant main effect of political ideology, B = 0.16, t(116) = 1.31, p = .192, 95% CI [-0.08, 0.41], sr2 = .01, a significant effect of the fairness rule, B = 0.77, t(116) = 2.69, p = .008, 95% CI [0.20, 1.34], sr2 = .06, and a significant interaction between political ideology and the fairness rule, B = 1.01, t(116) = 4.34, p < .001, 95% CI [0.55, 1.47], sr2 = .13.
Again, values of political ideology defining Johnson-Neyman’s significance regions (see Figure 1b) were 3.41 and below (i.e., values 1, 2 and 3 on a 1-7 scale, denoting conservative participants) and 4.86 and above (i.e., values 5, 6 and 7 on a 1-7 scale, denoting liberal participants). The effect was significant and negative for conservative participants, indicating that Conservatives were more willing to engage in collective action to oppose the tax reform when the merit rule was violated than when the equality or need rules were. In contrast, the effect was significant and positive for liberal participants, indicating that Liberals were more strongly motivated to engage in collective action to oppose the tax reform when the equality or need rules were violated that when the merit rule was.
Finally, to test the effect of the fairness manipulation on collective action intentions via perceived injustice, conditional upon political ideology, we performed a moderated mediation analysis with the use of the PROCESS macro (Model 8; Hayes, 2013). Upon requesting 10,000 bootstrap samples, the indirect effect of the manipulation was significant and negative for Conservatives (assessed at the 10th percentile), B = -0.61, 95% CI [- 1.15, -0.21]. For Liberals (assessed at the 90th percentile), the indirect effect of the fairness manipulation was positive: B = 0.98, 95% CI [0.59, 1.50]. Thus, as expected, Conservatives showed a higher motivation to engage in collective action when the merit rule (compared to the need and equality rules) was violated because they experienced more injustice when this rule was violated. In contrast, Liberals showed a higher motivation to engage in collective action when the equality and need rules were violated (compared to the merit rule) because they experienced more injustice when these rules were at stake.
Study 1 provided initial support for our hypotheses. Importantly, liberal and conservative participants alike considered the proposed reform as unjust and were willing to protest the proposed changes taking the same actions, which ranged from signing an online petition to participating in a demonstration or a rally (political ideology was not related to support for any of the specific actions). This supports the notion that individuals across the political spectrum are likely to perceive injustice and engage in the same type of normative actions. However, they differ in their underlying motivations.
To ascertain whether we can replicate these effects in a different (particularly, non-student) sample, and with a different social good (healthcare instead of taxes), we designed Study 2. Moreover, the manipulation in Study 1 referred to an alleged, not a real tax reform. Thus, in Study 2, we aimed to refer to a social good that was subject to a current political debate.
Study 2
The context for Study 2 was the ongoing discussion around the Affordable Care Act (“Obamacare”) in the United States. Although the act itself has been signed by the US president in March 2010, five years in, it would still stir up the political debate at the time when we collected data for the study (March 2015). The health reform bill clearly divided American society: as of March 2015, two in five (41%) surveyed adults had a generally favourable opinion about it, while a corresponding number (43%) held an unfavourable opinion (Kaiser Family Foundation, 2015). The changes proposed by the act were quite complex and, according to opinion polls, many US residents were confused about the content of the act (Washington Post, 2013). Thus, instead of referring explicitly to the act, we referred to healthcare changes more broadly. In Study 2, we focused on the violation of merit and violation of need distinction only. We did not include the equality rule, because we did not find any significant differences between the equality and need conditions in Study 1, and because the need rule seemed to be more important in the healthcare context (see e.g., Scott & Bornstein, 2009).
Method
Participants and Procedure
Power analysis revealed that approximately 88 participants were needed to detect a medium-size effect for interaction (f2 = .15) in a moderated multiple regression, with α = 0.05 and power at .90. A total of 114 US residents (43% female; Mage = 29.68; SD = 10.22), were recruited via Amazon Mechanical Turk (cf. Paolacci & Chandler, 2014) and received $ 0.75 as a compensation for their participation.
The majority of the sample had healthcare coverage (81%; 40% employer-sponsored, 14% private, 13% government-sponsored, 11% insured by a family member, 2% other); 55% had an annual household income below $50,000, 24% between $50,000 and $75,000, 21% above $75,000. On average, participants evaluated their relative socioeconomic status (SES) close to the SES of an average person in American society (M = 6.35; SD = 1.64 on a 1-10 scale).
The procedure was like in Study 1. Participants answered political ideology questions first. Next, they were randomly assigned to one of two conditions: presenting a violation of the merit (n = 60) or a violation of the need (n = 54) rule of fairness. Following the manipulation, they were shown an attention check question (one participant did not provide the correct answer, thus was excluded from further analyses), and items assessing perceived injustice and collective action intentions.
Materials
Political ideology
Political ideology was measured with three items tapping into general political ideology (1 = strongly conservative; 7 = strongly liberal), social conservatism (1 = strongly conservative; 7 = strongly liberal) and economic conservatism (1 = free market; 7 = welfare state), respectively. As the scores on the overall political ideology correlated strongly with items measuring both social, r(113) = .74, p < .001, 95% CI [0.64, 0.81], and economic conservatism, r(113) = .68, p < .001, 95% CI [0.57, 0.77], and the latter two measures were also moderately correlated, r(113) = .58, p < .001, 95% CI [0.44, 0.69], we collapsed all three items to create a composite measure of political ideology (α = .86)iii (cf. Carney et al., 2008).
Fairness manipulation
In both conditions, the first part of the manipulation text read: “Until recently the health care system in the US was mostly based on private or employer-sponsored insurance. Only some people were eligible for public insurance like Medicaid or Medicare.” Following this introductory line, in the violation of merit rule condition participants read: “Imagine that the US Congress put a stop to all private and employer-sponsored insurance by introducing universal, public healthcare. Access to healthcare would no longer be granted based on merit and individual contributions (e.g., no one would be allowed to get special treatment even at additional cost), but care would be equal for all citizens.” In the violation of need rule condition participants read: “Imagine that the US Congress put a stop to all public social insurance programs for the needy (such as Medicare or Medicaid), by retaining only employer-sponsored and private insurance. Access to healthcare would no longer be granted based on needs – to have health coverage, everyone, no matter what age and financial standing, would have to obtain it individually.”
To ensure that participants read the entire manipulation text, following the manipulation they were asked (in an open-ended question) to elaborate on the content of the US Congress amendment proposal they read.
Dependent variables
Perceived injustice (four items; α = .95). Respondents were asked: „After reading about the proposal, how do you feel about it?”: “I think that the introduction of the healthcare proposal would be unfair”; “I think that the healthcare proposal is socially unjust”; “The introduction of the proposal would be illegitimate”; “Introduction of the healthcare proposal would be justified”.iv Collective action intentions (seven items; α = .95). Subsequently, respondents were asked: „How willing would you be to participate in the following actions to prevent the introduction of the proposed changes?”: “sign a complaint against unconstitutionality of the proposal”; “sign an online petition”; “support political candidates opposing the proposal”; “send a letter of protest to the government or the media”; “write a critical comment on an online blog/discussion forum”; “donate money”; “participate in a discussion meeting”.
As in Study 1, at the end of the survey, we additionally asked respondents to what degree each of the allocation principles aligned most closely with their definition of social fairness. Conservative political ideology was linked to higher preference for the merit-rule, r(112) = .43, p < .001, and weaker preference for the need and the equality rules, r(112) = -.50, p < .001 and r(112) = -.55, p < .001), respectively.
Results and Discussion
We employed the same analytical approach as in Study 1. Table 2 shows descriptive statistics and correlations between the measured variables.
Table 2
Descriptive Statistics and Zero-Order Correlations Between Measured Variables (Study 2)
| Variable | M | SD | 2 | 3 | 4 |
|---|---|---|---|---|---|
| 1. Political ideology | 4.50 | 1.48 | -.02 | .06 | -.10 |
| 2. Injustice | 3.99 | 1.94 | – | .82** | .49** |
| 3. Collective action intentions | 3.12 | 1.98 | – | .36** | |
| 4. Violation of merit (1) vs. violation of need (2) | – | – | – |
Note. Scale range for all continuous variables was 1-7. Lower values indicate a more conservative political ideology.
**p < .001.
Next, for each dependent variable, we ran a regression analysis with political ideology, the manipulated fairness rule (coded as 1 = merit, 2 = need; step 1), and the interaction of both (step 2) as independent variables.
Perceived Injustice
Analysis indicated no main effect of political ideology, B = 0.36, t(114) = 0.33, p = .741, 95% CI [-0.18, 0.25], sr2 = .00, a significant main effect of the manipulated fairness rule, B = 1.89, t(114) = 5.93, p < .001, 95% CI [1.26, 2.52], sr2 = .24, and a significant interaction between the experimental condition and political ideology, B = 1.54, t(114) = 9.46, p < .001, 95% CI [1.22, 1.86], sr2 = .34.
As defined by Johnson-Neyman’s significance regions (see Figure 2a), the interaction was significant and negative for values 2.69 and below (denoting conservative participants; 16% of the sample), and significant and positive for values 3.69 and above (denoting liberal participants; 69% of the sample). Conservatives rated the changes in healthcare as more unjust when they violated the merit rule than when they violated the need rule. Liberals, in contrast, rated those changes as more unjust when they violated the need rule than when they violated the merit rule.
Collective Action Intentions
Replicating results of Study 1, we found no main effect of political ideology, B = 0.13, t(114) = 1.08, p = .282, 95% CI [-0.11, 0.36], sr2 = .01, a significant main effect of the manipulated fairness rule, B = 1.44, t(114) = 4.15, p < .001, 95% CI [0.75, 2.13], sr2 = .13, and a significant interaction between the experimental condition and political ideology, B = 1.50, t(114) = 7.85, p < .001, 95% CI [1.12, 1.88], sr2 = .31. As indicated by Johnson-Neyman’s significance regions (see Figure 2b), the effect was significant and negative for values of political ideology 2.90 and below (denoting conservative participants) and significant and positive for values of political ideology 4.00 and above (denoting liberal participants). Thus, Conservatives were more willing to engage in collective action aimed against the proposed changes when the merit rule was violated than when the need rule was. Conversely, Liberals were more willing to engage in collective action when the need rule was violated than when the merit rule was.
Finally, as in Study 1, we tested whether the interactive effect of the fairness manipulation and political ideology on collective action intentions could be explained by perceived injustice. The indirect effect of manipulation (coded as 1 = violation of merit, 2 = violation of need) was significant and negative for Conservatives (assessed at the 10th percentile), B = -1.13, 95% CI [-1.96, -.34], and significant and positive for Liberals (assessed at the 90th percentile), B = 3.62, 95% CI [2.51, 4.72]. Thus, as expected, and in line with Study 1, Conservatives showed a higher motivation to engage in collective action when the merit rule (compared to the need rule) was violated because they were more likely to experience this violation as unjust. In contrast, Liberals showed a higher motivation to engage in collective action when the need rule was violated (compared to the merit rule) because they were more likely to experience the violation of the need rule as unjust.
Figures 2
Conditional Effects of Fairness Manipulation on (a) the Perceived Injustice and (b) Collective Action Intentions (Study 2).
Note. Positive effect denotes higher values in the violation of need condition, negative effect denotes higher values in the violation of merit condition.
General Discussion
Our studies underscore the importance of accounting not only for the strength of injustice appraisals but also for different conceptions of fairness in predicting collective action among individuals across the political spectrum. In two experiments, we show that different notions of what is fair among Conservatives and Liberals lead to varying perceptions of injustice and readiness to protest it. Specifically, Conservatives consider the violation of the merit rule (versus the need and equality rules) as more unfair and are more willing to engage in collective action upon it. In contrast, Liberals consider the violations of the need and equality rules (versus the need rule) as more unfair and are more willing to engage in collective action upon them. We show these effects in samples from two different cultural contexts (Germany and the US), and for two different social goods (tax money and healthcare). Our research is thus an important step towards a fuller understanding of different motivations to engage in collective action among Conservatives and Liberals.
In both studies we used established measures of political ideology (Carney et al., 2008; Fuchs & Klingemann, 1990), and obtained similar results in a sensitivity analysis comparing results for single items assessing social and economic conservatism (Study 1), and overall political ideology versus a composite measure including social and economic conservatism items (Study 2). Additionally, in Study 2 we obtained similar results measuring political ideology using right-wing authoritarianism and social dominance orientation (detailed information about the scales used is reported as endnotes). We believe that these analyses confirm the robustness of our results and are an indication of the study's strength.
Why is this finding meaningful and important? Although the central focus of liberal protest in the prior literature suggests that Conservatives are less interested in protest, our data shows that there is no difference in the overall willingness to protest in terms of readiness to engage in collective action among Liberals and Conservatives. One of the reasons for the apparent absence of Conservatives in protest is likely due to the operationalization of injustice perceptions as violating the equality and the need principle of fairness (e.g., van Zomeren et al., 2008), typically associated with the liberal protests. Conservatives do not regard it as unjust when the equality and need rules are violated – in fact, they welcome forces aiming at maintaining social hierarchies (e.g., Cameron & Nickerson, 2009; Hennes et al., 2012; Jost et al., 2017; Jost & Thompson, 2000; Osborne et al., 2019). In contrast, as soon as the merit principle is at stake, conservative individuals are likely to feel threatened, to perceive injustice and to be willing to engage in action to restore it. The present work thereby highlights the vital role that violations of the merit rule can play in instigating social protest, a relationship that had remained widely unacknowledged.
Implications for Collective Action Models and Social Policy
Our studies show that Liberals and Conservatives were equally interested in collective action against tax reform in Study 1 and against changes in the healthcare system in Study 2. However, observers of protests need to take into account that the same behaviour (a protest against changes in the tax reform or healthcare system) can have completely different underlying motivations depending on different notions of (in)justice. For this reason, we recommend that studies on collective action consider not only measures of the strength of injustice appraisals, but also their content, to ascertain which fairness principles guide one’s perceptions of (in)justice.
Our research also contributes to a wider debate on the role of ideology and personal variables in predicting social protest. Previous studies have shown that willingness to engage in collective action depends on other ideological variables, such as group-based opinions (McGarty, Bliuc, Thomas, & Bongiorno, 2009), personal (van Stekelenburg, Klandermans, & Van Dijk, 2009) and group values (van Zomeren & Spears, 2009), and moral convictions (Milesi & Alberici, 2018; Skitka & Bauman, 2008; van Zomeren, Kutlaca, & Turner-Zwinkels, 2018; van Zomeren, Postmes, Spears, & Bettache, 2011). Engagement in collective action has been proven to also depend on individual-level traits, such as group efficacy beliefs (van Zomeren, Saguy, & Schellhaas, 2013), or regulatory focus (Zaal, Van Laar, Ståhl, Ellemers, & Derks, 2011). This growing body of research suggests that collective action is shaped not only by group-based or context-dependent variables, but also by individual-level and context-independent variables (e.g., Saeri, Iyer, & Louis, 2015).
Regarding social policy, our results show that messages and policies based on moral claims referring to social justice always come with a caveat: definitions of fairness differ depending on who the audience is. What Liberals consider as fair, Conservatives evaluate as unfair and vice versa. This can have significant consequences for how people evaluate social policies and whether they follow them or engage in social protest.
Moreover, our results add to the recent literature by showing that not all social protests lead to a reduction of social inequalities (e.g., Osborne et al., 2019). Whereas protest based on violations of the equality and the need rules is likely to reduce inequality in the society and consider needs of the disadvantaged accordingly, protests based on the merit rule likely lead to a different outcome. Merit is at odds with equality to the extent that it requires that goods or services are distributed in proportion to contribution (e.g., Rawls, 1971), which is usually subjective and unequal (Scott et al., 2001). Knowing that belief in merit justifies inequalities (e.g., McCoy & Major, 2007), it is unlikely that protests aimed at restoring merit will foster equality at the societal level.
Limitations and Future Directions
We acknowledge some limitations of the present work. In both studies, the means for injustice perceptions and collective action items were somehow low (i.e., below the mid-point of the response scale). This is not unusual in studies on collective action, given that traditional forms of collective action (e.g., demonstrations) are rare in real life (Jost et al., 2017). One explanation for the relative infrequency of collective action is that the absence of political opportunities, structural constraints, a lack of ingroup identification or group efficacy beliefs and psychological motivations (such as system justification or just world beliefs) make protest often seem pointless or impossible (e.g., McAdam, 2017). Another explanation lies in the somewhat narrow operationalization of collective action in the current literature which might miss important behaviours people engage in when responding to injustice (for a recent exception, see Stroebe, Postmes, & Roos, 2019). We tried to address this shortcoming by operationalizing collective action quite broadly, including not only actions typically associated with organized social protests or crowd behaviour (e.g., participation in a demonstration or a rally), but also actions that can be enacted by single individuals, such as giving support to political candidates opposing the proposal or money donation. However, we need to acknowledge that being affected by the tax reform and changes in the health care system might not have been very central for participants explaining the lower injustice appraisals and intentions to engage in collective action.
Future studies could examine potential differences in the preferred ways in which Liberals and Conservatives express protest. To our knowledge, this idea has not been investigated yet. Although we did not observe significant differences between Liberals and Conservatives in their willingness to instigate any of the behaviours included in our studies, further exploration of actions taken by Liberals and Conservatives could shed light on collective action more broadly.
Another possible limitation of our studies concerns the cultural contexts in which we measured political ideology. Both our samples were collected in Western countries in which social and economic conservatism are correlated positively for historical reasons (Jost, Federico, & Napier, 2009). This is not the case, for example, in post-Communist countries in which social and economic conservatism are correlated negatively due to recent history (e.g., Kossowska & Hiel, 2003). Future studies could assess to what extent these and other cultural differences affect collective action among Conservatives (vs. Liberals).
Although the inclusion of the merit rule of fairness contributes to the understanding of collective action among Conservatives, by no means can it account for all occurrences of conservative collective action. In cases such as pro-life protests or social movements aimed at defending the traditional family values, other psychological variables, such as moral convictions, can be expected to play a more prominent role (Milesi & Alberici, 2018). Similarly, previous studies looking at liberal collective action indicate, among others, that injustice perceptions are more likely to motivate protest via the emotion-focused route than via the problem-focused route, in which efficacy is a stronger predictor (e.g., van Zomeren et al., 2004). These injustice perceptions have been shown to be an important driver of collective action in general (Runciman, 1966; Walker & Smith, 2002). Examples of conservative collective action, in which the merit rule likely plays a key role, ranges from opposition to progressive taxes and support for fiscally conservative movements more broadly, to protests against the (ostensibly low-skilled) immigrants, refugees and asylum seekers and counter-propositions of merit-based immigration systems giving preference to high skilled immigrants, or support for restrictions to welfare programs limiting eligibility to only those who can prove that they “made an effort” to find a job.
Finally, although our studies speak to the importance of differentiating between various notions of fairness in predicting collective action, we focused only on the three main normative rules of distributive fairness. In future studies, it would be interesting to investigate the importance of different allocation principles against the importance Conservatives typically attribute to the preservation of the status quo, group coherence, and social status. For example, how would Conservatives react if they were faced with a merit-based rule of fairness that threatened their status, such as preferential policies for high-skilled immigrants? (cf. Son Hing et al., 2011). Would they react differently to an equality-based rule of fairness depending on whether it threatens or does not threaten group coherence?
Conclusion
The present work addresses one shortcoming in collective action research by exploring the interactive role of political ideology and injustice appraisals in predicting social protest. We argued that the established models predicting collective action have been developed based on liberal ideas of injustice showing that progressive collective action occurs when people perceive that the equality or need rule of fairness are violated. We demonstrated, however, that these perceptions of injustice cannot explain the occurrence of conservative social protests, because Conservatives were interested in protest when the merit rule of fairness was violated. Thus, this research takes one step forward and contributes to the recent literature highlighting the important role of ideology in collective action research.