Abstract
Populist attributions of blame have important effects on citizens’ attitudes, cognitions, emotions, and behaviors. Extending previous studies that have mostly looked at populist messages blaming political elites, we use an online survey experiment (N = 805) to investigate the effects of blaming different elitist actors in populist and non-populist ways: (1) political elites, (2) corporate elites, (3) scientific elites, and (4) a combination of these elites. We compare mere causal responsibility attribution to populist blame attributions that highlight a central opposition between “pure ordinary people” and “corrupt elites” to tease out the specific effect of populist rhetoric. Results suggest that populist messages blaming all elites have stronger effects than isolated cues. Furthermore, populist blame attribution leads to more negative perceptions of the elites than non-populist blame attribution. The finding that populist frames can – under some conditions – activate negative perceptions toward different elites is alarming in times of declining trust in societal institutions, and calls for a more encompassing understanding of populists’ scapegoats.
1 Introduction
The factual status of empirical evidence and scientific knowledge is increasingly distrusted and doubted (e. g., Van Aelst et al., 2017). Populist rhetoric ties in with this communication era by emphasizing the antagonism between the “honest people” and different elites that are said to deliberately hide the truth from the people. However, to date, most conceptual and empirical literature on populism has applied a rather restricted definition of the elites as the political establishment that is corrupt and/or not representing the ordinary people’s will (e. g., Jagers and Walgrave, 2007; Mudde, 2004; Taggart, 2000; but see Mede and Schäfer, 2020). In this paper, we investigate how messages that blame allegedly corrupt political, corporate, and scientific elites for causing problems experienced by ordinary people – which we understand as populist attributions of blame – influence public perceptions of scapegoated elites. In line with the affinity between populist rhetoric and conspiracy theories (e. g., Bergmann, 2019; Silva, Vegetti, and Littvay, 2017), the effects of populist communication in which all elitist actors are scapegoated together are also relevant to consider. To provide an example, populists may assign blame to scientists and corporations that allegedly collaborate with the established political order to silence the ordinary people and to present them with a “fake” or distorted reality that hides the real sources of power at play in society.
We follow definitions that understand populism as a thin-centered ideology that revolves around the antagonism between the “pure people” and the “corrupt elites” (e. g., Mudde, 2004). Here, we specifically focus on the causal element of the opposition between the people and the elites (see, e. g., Busby, Gubler, and Hawkins, 2019; Hameleers, Bos, and de Vreese, 2017; Vasilopoulou, Halikiopoulou, and Exadaktylos, 2014). Specifically, populist ideas emphasize a “blame frame” in which the ordinary people are seen as an innocent and powerless in-group that is victimized by the culpable elites held responsible for causing problems experienced by the people. As political, corporate, and scientific elites can be regarded as salient scapegoats in the current fragmented and polarized information landscape where trust in different institutions is declining, we consider it important to assess how blame attribution targeting these different elite actors affects people’s perceptions toward them.
Extant literature indicates that populist communication has important effects on the attitudes, cognitions, emotions, and behaviors of citizens (e. g., Aalberg, Esser, Reinemann, Strömbäck, and de Vreese, 2017; Hameleers et al., 2018; Matthes and Schmuck, 2017; Müller et al., 2017; Wirz, 2018). Against the backdrop of the lack of research on blame shifting to non-political elites in populist rhetoric, this study relies on an experiment conducted among a varied sample of Dutch citizens, and uses a between-subjects factorial design with two factors: blame attributions (shifting blame to the political, corporate, scientific, or all elites) and populist rhetoric (present or absent). In the setting of a salient global issue, climate change regulations, we compare the effects of generic blame attributions to populist blame framing (see also Busby et al., 2019), and assess how exposure to these frames activate populist attitudes and negative stereotypical perceptions of the elites.
Together, this study aims to provide insights into how the “chameleonic” nature of populist ideas on the supply-side, and the multifaceted meaning of the “elites” in populist discourse, affect perceptions toward the targets to whom blame is attributed among receivers. By demonstrating whether blame attributions outside of the political realm are effective, this paper aims to arrive at a better understanding of how different communicators can effectively flirt with populist ideas whilst circumventing attacks on the level of governance they are part of.
Populist blame attribution to different elites
Through the lens of populist rhetoric, society is seen as being divided into the ordinary people, on the one hand, and the “corrupt” or self-interested elites, on the other (e. g., Jagers and Walgrave, 2007; Mudde, 2004; Taggart, 2000). Populism can be regarded as a social identity frame that emphasizes an antagonism between the in-group of ordinary people and the out-group of corrupt elites (e. g., Bos et al., 2020). In this social identity frame, blame attributions are central (e. g., Busby et al., 2019; Hameleers et al., 2017): The ordinary people are regarded as a deprived entity that suffers from the failing culpable elites. These elites are held accountable for causing the deprivation of the people’s in-group. This blame frame can be attached to all sorts of issues owned by populist actors, such as immigration (i. e., corrupt elites allegedly allow too many immigrants to enter the nation, which causes the cultural and economic deprivation of native citizens) or the economy (i. e., self-interested elite actors allegedly fill their pockets at the expense of the needs of ordinary citizens).
Populist ideas can cultivate the ordinary people’s opposition to different elitist actors, of which the political elites, corporations, and scientists may be the most salient ones in populist discourse today. The prevalence of these three scapegoated actors in populist discourse can be understood against the background of shifts toward factual relativism and eroding trust in corporate, political, and scientific intuitions in an “epistemic democracy” (e. g., Van Aelst et al., 2017; Waisbord, 2018). In this paper, we stay close to the definition of populism as a thin-cored ideology by considering blame attributions to political elites (see, e. g., Busby et al., 2019; Hameleers et al., 2017). In line with this, populist messages frequently highlight a causal opposition between innocent ordinary citizens and corrupt politicians that do not represent the people’s will.
Next to political elites, corporations and scientists are frequently regarded as part of the “corrupt” elites and blamed for misleading and harming ordinary people and their concerns (Hameleers and Minihold, 2020). Hence, corporations are a salient scapegoat for (left-wing) populists that point to the failed representation of the ordinary people’s concerns by profit-maximizing elites, whereas scientists and political actors are accused of disseminating a worldview that is unresponsive to the needs and truths of the ordinary people (e. g., Mede and Schäfer, 2020). In this paper, we do not explicitly look at media elites as an out-group, since blaming the media may be too strongly related to the ideas of deception, post-truth politics, fake news accusations, and disinformation – which would make the comparison to the other blame attributions less valid.
Following extant research on the effects of populist communication (e. g., Matthes and Schmuck, 2017; Müller et al., 2017), we look at the effects of populist communication on (1) the activation of populist attitudes, and (2) the cultivation of negative out-group perceptions. In addition, we compare the effects of blame attribution that simply holds these elites accountable for causing important problems in society with explicit populist blame framing (also see Busby et al., 2019). More specifically, although populist blame attributions explicitly emphasize a moral divide between “good” ordinary people and “evil” elites, this antagonism and moral interpretation is not central in blame attributions without explicit populist cues, which simply state that some actors are responsible for causing negative situations and/or developments without emphasizing a central divide between “good” people and “evil” elites.
To illustrate this difference more explicitly, we regard attributions of responsibility without populist elements as all attributions of blame that do not refer to a central opposition between ordinary people and corrupt elite actors. For example, a message that argues that the government is responsible for not clearly informing the public on Covid-19 health regulations assigns blame to political elites but does not refer to a moral antagonism between “pure” or “ordinary” citizens and “corrupt” elite actors. If such a message would explicitly argue that failing, self-interested, and corrupt politicians deceive the ordinary people by not clearly informing the public, we would classify this as populist blame attributions. In this paper, we isolate the persuasiveness of populist communication and assess whether the mere attribution of causal responsibility or the Manichean outlook of populism drives the effects of populist communication on out-group perceptions.
We acknowledge that (populist) attributions of blame can come from different sources, including politicians, corporations, and scientists themselves. Even though the source may matter for the persuasiveness of populist communication, we refrained from including explicit source cues in this study, as such references may bias the findings based on previous beliefs related to the source (i. e., a recipient who distrusts the mainstream media may counter-argue populist messages from the mainstream media). In addition, populist ideas may oftentimes come from unspecified or vague sources, especially in online information settings. As we are mainly interested in how populist rhetoric itself (and not the source) influences political beliefs related to scapegoated out-groups, we kept source references similarly vague across conditions by referring to the (populist) messages as items that have appeared online recently.
The effects of populist scapegoating on out-group perceptions
Psychologically, the effects of populist blame attributions can be understood as negative stereotyping (Dixon, 2008) or trait activation (Richey, 2012). This basically means that when people are exposed to populist ideas that emphasize an antagonistic divide between the ordinary people and corrupt elites, similar mental maps or schemata of the people (good associations) and the elite (bad associations) may become more salient in the minds of receivers – and these mental maps are consequentially used when citizens make political evaluations. When populist communication primes a social identity of the ordinary people versus the corrupt elites, people may thus rely on these identity constructions when making political judgements.
The question remains why people would accept populist communication and use blame attributions when forming their own political opinions. The theory of social identity framing and collective action may offer an answer (Bos et al., 2019). Importantly, populist communication expresses a threat to the in-group of the ordinary people. Hence, the elites are accused of depriving the people on a symbolic, cultural, or material level (e. g., Albertazzi and McDonnell, 2008). At the same time, a cause is introduced: The elites are seen as self-interested or corrupt and only care for themselves. Such constructions of deprivation and blame attributions are found to mobilize and activate citizens to engage politically (e. g., Gamson, 1992; Mols, 2012; Polletta and Jasper, 2001). More specifically, when social identity frames stress that (1) the in-group faces a threat, (2) offers credible scapegoats for this threat, and (3) introduces solutions to avert this threat, the in-group is likely to engage with the message and takes action to avert the threat (e. g., Gamson, 1992).
Although people may not directly change their behaviors as a consequence of being exposed to a populist message, their negative perceptions toward the culpable elites and perceived closeness to the ordinary people as an in-group may be activated – even if only lower levels of these schemata are present among receivers (Busby et al., 2019). In this paper, we measure the activation of the perceived divide between the people and the elites in two different ways: negative stereotypes toward the out-group (e. g., Matthes and Schmuck, 2017) and the activation of the perceived divide between the ordinary people and the corrupt elite – also referred to as populist attitudes (e. g., Schulz et al., 2017). Against this backdrop, the following hypothesis is introduced:
|
H1: |
Compared to a neutral control group, exposure to populist messages in which (a) political elites, (b) corporate elites, or (c) scientific elites are blamed activate populist attitudes and negative stereotypes toward each individual elitist actor. |
The effects of cultivating a populist conspiracy between different elites
Populist rhetoric oftentimes cultivates a conspiracy theory – which indicates that different enemies are blamed for secretly collaborating to hide reality from the ordinary people (Bergmann, 2019 Silva et al., 2017). Here, we should emphasize that populism and conspiracy theories are different concepts, although they can be related at times. At its core, populism refers to a central antagonism between pure or ordinary people and corrupt elites, whereas conspiracy theories forward a more encompassing interpretation of reality that points to a “hidden” truth or forces of power that cover up schemes of “evil” forces. More specifically, conspiracy theories can be understood as comprising at least three elements (Barkun, 2003). First, the source and causes of novel or disruptive issues (i. e., Covid-19, climate change) are attributed to evil forces that are kept hidden from the public (Hornsey et al., 2018). Second, events are interpreted as a struggle between good and evil (Barkun, 2003). Consequentially, historical events are seen as planned conspiracies of evil power forces that override the good and innocent people. Third, conspiracy theories refer to a hidden truth. This truth is allegedly hidden to obfuscate power discrepancies: Covid-19 is, for example, alleged to be a biological weapon used by governments to maintain the power discrepancy between the people and the elites.
Populist ideas may, but do not have to, be related to these elements. First, populist ideas and conspiracy theories are both founded on a Manichean perspective on society: Evil outsiders are seen as posing a severe threat on the pure and honest in-group from which the real sources of power are kept hidden (e. g., Barkun, 2003). Second, populist ideas and conspiracy theories both have an antagonistic relationship to established knowledge and expert knowledge. Hence, populists are known to express their distrust toward mainstream media and elitist sources of information – as these sources may spread “fake news” to damage or mislead the people (e. g., Krämer, 2018). Likewise, conspiracy theorists argue that the elitist and mainstream version of reality deceives the people and distracts citizens from the actual sources of power that oppress them (e. g., Fenster, 2008). Third, populist rhetoric often blames media and political elites from hiding reality from the people – which is allegedly done to silence the ordinary people. Based on this, we argue that conspiracists’ narratives can include a populist framework by highlighting a central opposition between the pure and deprived ordinary people versus the “evil” elites.
In this paper, when looking at the connection between populist ideas and more compelling conspiracies, we focus on a combined attribution of blame to political elites, corporations, and scientists. Generally, we expect that populist communication has stronger effects on priming stereotypes and populist attitudes when the elites are framed as part of a conspiracy that threatens the ordinary people by hiding reality from them (see Hameleers and Minihold, 2020). Hence, populist social identity frames cultivate a stronger threat to the ordinary people when a conspiracy between different elites is emphasized – indicating that the people are threatened by hidden forces of power that conspire to hide the reality from the people. When a stronger threat to the ordinary people is framed, and when scapegoats can credibly be associated with this in-group deprivation, a stronger urge to engage with identity frames should be cultivated (e. g., Gamson, 1992). Against this backdrop, the following hypothesis is introduced:
|
H2: |
Populist messages in which all three elites are blamed have stronger effects on activating populist attitudes and negative stereotypes than exposure to isolated blame attributions to the political, corporate, or scientific elites. |
The alignment between populist scapegoating and prior attitudes
Ceteris paribus, we assume that the cultivation of a divide between the people, political elites, scientists, and/or corporations can activate negative perceptions toward these elitist actors. However, this would presume that citizens actually have (some) negative perceptions toward these scapegoats available that lie dormant to be woken up or activated by populist communication (Busby et al., 2019). Although direct effects haven been found in some experimental research (e. g., Hameleers et al., 2018), many studies point to a conditional effect of populist messages on political perceptions (e. g., Bos et al., 2013; Matthes and Schmuck, 2017).
In this paper, we focus on existing levels of (dis)trust toward the three elitist scapegoats to which blame is attributed as the mechanism by which populist communication affects receivers. When people distrust scientists, politicians, or corporations, we expect that they are more likely to accept messages that frame these as culprits. Hence, lower levels of trust in elites should make populist social identity frames more personally relevant – and should strengthen the perceived threat that the elites pose on the ordinary people. Indeed, as demonstrated in prior experimental research, populist communication has the strongest effect on people with higher levels of distrust toward the elites (Bos et al., 2013). More specifically, people who distrust elitist actors may be more likely to agree with populist ideas that hold elites accountable, and may thus also perceive populism’s social identity frame as more personally relevant – triggering their already existing mental schemata that the elites cannot be trusted. Although this effect has only been established for political elites, we extrapolate the mechanism to corporate and scientific elites: The less people trust these different elites, the more populist attributions of blame resonate with people’s pre-existing schemata. Against this backdrop, we put forward the following hypothesis:
|
H3: |
Populist messages that attribute blame to the elites have the strongest effects on populist attitudes and negative out-group perceptions among people with lower levels of trust in the scapegoated elitist out-group. |
2 Method
The hypotheses are tested using an online survey experiment in the Netherlands. To enhance the transferability of the findings to other settings, we used a global politicized topic that is interpreted and framed in equally polarizing ways across the globe: climate change policies. This issue is not owned by populist parties, which avoids effects driven by the issue instead of the framing of the message. We did not refer to specific source cues to make sure that we could isolate populist blame attributions that can be extrapolated to different countries, media sources, and electoral settings. Thus, in the cover story preceding exposure to the message, we simply stated that the news message was “recently published as a news update”. without specifying where it came from. The design of the experiment can be described as a 4 (blame attribution: political elites versus corporate elites versus scientific elites versus all elites) × 2 (populist communication: absent versus present) + control (neutrally framed news message on the same issue) between-subjects factorial design. Although we regard populist ideas and conspiracy theories as separate constructs, we believe that the ideas of a conspiracy theory are central in the condition that blames all elites together: This condition stresses that the allegedly powerful and self-interested elites conspire to hide reality from the people (see Silva et al., 2017 for the connection between populist ideas and conspiracy reasoning). Thus, in this condition, corporations, political elites, and scientists are together referred to as “evil” forces that deceive the people by hiding reality from them – which allegedly helps them to maintain the power imbalance between the people and the ruling elites. This condition therefore integrates a populist antagonism (the people versus the corrupt elites) with a conspiracist narrative.
Sample
We relied on a large international research agency (Kantar) to collect our data. The panel recruited participants from a large database (voluntary opt-in lists; participants were rewarded for participation with a credit system). We achieved 857 valid completes (response rate 86.5 %). The composition of the sample was varied and approaches national representativeness on a number of important factors (left-right self-placement using 10 scale points, region, age, occupation, education, gender). The panel company sent out invitations to match representativeness. We did not enforce hard quota from our side. The mean age of participants was 48.64 years (SD = 13.05). 53.2 % self-identified as female. The lowest educational categories were occupied by 25.2 % of all participants, and 28.8 % fell in the highest categories (26.0 % moderate level of education). 37.2 % identified mostly as left-wing and 48.2 % as right-wing (corresponding to the current situation in the Dutch government). Deviations between the sample composition and census data were relatively small (< 5 %).
Independent variables
The stimulus material is based on an actual news update discussing effective policies to mitigate climate change. The basic premise of the article was kept constant. Climate change is seen as a serious problem, and different actors are accused of not solving it adequately. In the various versions, the different actors were held responsible for obstructing the implementation of effective policies. Furthermore, to manipulate the second factor (populist communication), different versions were created to reflect populist (emphasizing the binary divide between the ordinary people and the corrupt elites) and non-populist rhetoric (attributing causal responsibility to different actors). These conditions were compared to a control group in which no particular actor was blamed. Table 1 provides an overview of the different conditions. The length of the articles was kept as similar as possible (shortest: 291 words; longest: 310 words). The stimulus materials (translated from Dutch to English) can be found in Appendix B. Although it could be argued, on the one hand, that this specific framing of climate change is not a typical issue framed in populist terms, on the other, populist ideas have become mainstream – and can be attached to different issues and positions. We thus investigate the effects of populist blame attribution in a “least likely case” – attached to a global crisis that is typically not part of the (right-wing) populist agenda. We did control for people’s pre-treatment agreement with the problem statement that climate change is real and caused by human interference. This did not change the outcomes, although skeptical beliefs toward climate change are correlated with populist attitudes (r = .35, p < .001). In this setting, we ask to what extent using populist arguments to frame an issue that is unowned by populists can impact political and societal evaluations. Group sizes across treatment conditions were similar, and there were no differences in composition on relevant variables across the groups.
Overview of experimental conditions and mean scores of the dependent variables.
|
Condition |
Description of condition |
M (SD) populist attitudes |
M (SD) stereotypes political |
M (SD) stereotypes corporate |
M (SD) stereotypes scientists |
|
1: Control condition |
Climate change problem and remedial policies discussed, no blame. |
4.30 (.83) |
3.80 (1.17) |
4.09 (1.04) |
5.04 (1.12) |
|
2: Political elite blamed |
Politicians fail to implement climate change policies. |
4.33 (.99) |
3.75 (1.24) |
3.80 (1.04) |
4.69 (1.13) |
|
3: Political elite blamed, populist frame |
Politicians intently obstruct implementing climate change policies at the expense of the ordinary people (antagonism). |
4.29 (.95) |
3.58 (1.37) |
3.75 (1.18) |
4.88 (1.24) |
|
4: Corporations blamed |
Corporations blamed for blocking the implementation of climate change policies. |
4.37 (.86) |
3.59 (1.26) |
3.84 (1.00) |
5.07 (1.14) |
|
5: Corporations blamed, populist frame |
Corporations wilfully obstruct implementing climate change policies at the expense of the ordinary people (antagonism). |
4.35 (.97) |
3.58 (1.27) |
3.81 (1.27) |
5.05 (1.35) |
|
6: Science blamed |
Scientists’ lack of consensus hinders the implementation of climate change policies. |
4.51 (.91) |
3.79 (1.35) |
4.08 (1.14) |
4.69 (1.15) |
|
7: Science blamed, populist frame |
Scientists intently obstruct the implementation of climate change policies at the expense of the ordinary people (antagonism). |
4.27 (.98) |
3.55 (1.39) |
3.75 (1.19) |
4.73 (1.16) |
|
8: All elites blamed |
Elites are all responsible for lack of effective climate change policies. |
4.39 (.90) |
3.86 (1.32) |
4.00 (1.20) |
5.03 (1.03) |
|
9: All elites blamed, populist conspiracy frame |
Elites collaborate to obstruct climate change policies at the expense of the ordinary people – who are deliberately misled (antagonism + conspiracy). |
4.44 (.89) |
3.36 (1.31) |
3.69 (1.24) |
4.71 (1.23) |
Note. Mean scores of the dependent variables are measured on 7-point scales. Higher scores on stereotypical scales indicate more positive evaluations of the respective scapegoat.
Dependent variables
Based on existing scales for populist attitudes (Akkerman, Mudde, and Zaslove, 2014; Schulz et al., 2017), we used a subset of items to tap the perceptual divide between the ordinary people and the culpable elites. We revised existing populist attitude scales to make them more applicable to the different elitist scapegoats involved in this study (see Appendix C for an overview of the eleven items). All individual items were measured on 7-point disagree–agree scales, and formed one unidimensional scale of populist attitudes (M = 4.36, SD = .92, Cronbach’s alpha = .818).
Next to the populist attitude scale, we tapped stereotypical images toward all three scapegoated actors by asking participants to rate these actors on the following traits (tapped with binary oppositions using 7-point scales): dishonest–honest, lazy–hard-working, unreliable–reliable, hostile–friendly. These traits represented unidimensional scales that tapped stereotypical out-group depictions of the political elite (M = 3.65, SD = 1.30, Cronbach’s alpha = .899), corporate elite (M = 3.87, SD = 1.15, Cronbach’s alpha = .854), and scientific elite (M = 4.88, SD = 1.18, Cronbach’s alpha = .904). The higher the scores on these traits, the more positive the elitist actors belonging to this group were rated. Hence, overall, scientists were rated as most positive, whereas the political establishment was ascribed the most negative traits. Although the reliable–unreliable pair semantically relates to the trust measurement, it tapped different beliefs, as the stereotypical evaluations (DV) measured associations with the out-group as a whole, whereas the trust measurement measured the evaluation of the included elite actors’ capability to fulfill their role expectations (correlations are < r = .22 for all the actors). As a robustness check, we additionally ran the analyses without the reliable–unreliable pair, as these statements relate to the measurement of trust as a moderator – this did not change the results.
Moderator: Trust in different elites
Just like we tapped stereotypical perceptions for each group separately, we measured the level of trust that participants had in the political establishment (M = 3.77, SD = 2.46), corporations (M = 4.17, SD = 2.39), and the scientific community (M = 4.35, SD = 1.49). These levels of trust were measured prior to exposure to the experimental conditions (with the item wording: “Can you indicate to what extent you trust the following actors in society on a scale from 1 (not reliable at all) to 7 (very reliable)?” We used the labels “the government”, “corporations”, and “the scientific community” for the three different scapegoats in the final analyses, but also included different wordings as robustness checks (politicians in general, multinationals, corporate managers, and scientists, respectively). Using these alternative variables in our analyses yielded exactly the same results.
Procedure
After accessing the survey via the digital invitation, participants completed the informed consent procedure. Upon agreement, they completed a pre-treatment survey block with questions on political preferences, the central moderator of this study (trust in different elitist actors), and demographics. After this short block of questions, they were forwarded to the experimental block: Participants were randomly allocated to one of the nine conditions (equal group sizes) and read the stimuli on climate change developments. Participants were forced to stay on this page for 30 seconds or more, which ensured that they had sufficient reading time for the full article. They were explicitly instructed to pay attention to the news article and were informed that they could not skip the stimulus material (the article). We also ensured that manipulations were visible in the headers and all paragraphs of the articles to increase the likelihood of exposure to the manipulations. After reading the stimulus material, the post-test survey with items on the dependent variables and manipulation checks followed. Finally, participants were debriefed and thanked for their cooperation (mean completion time = 16.07min, SD = 24.76min).
Manipulation checks
To see whether respondents accurately perceived which actor was blamed in the article, we asked, on a 7-point Likert scale, to what extent the article blamed political elites, scientists, or corporations for problems related to climate change. Results of ANOVAs indicate that the manipulation was successful (details on the manipulation checks are included in Appendix D).
Using post-hoc randomization checks, we ensured that the different conditions did not differ in the composition of the sample regarding the controls we used (left–right self-placement, region, education, gender, or age). We therefore did not include these as additional controls or explanatory variables in the models – which would also bias the estimates of the randomly allocated treatment factors.
Analyses
We tested the hypotheses using OLS-regression models, in which we first recoded the experimental conditions into k – 1 treatment conditions (the control condition was kept as a reference condition). We then compared treatment effects relative to the control condition, in which populist and blame cues were absent. We used this strategy as it allowed us to estimate treatment effects at different levels of the moderator – without recoding trust into (potentially arbitrary) categories. As a robustness check, we ran ANOVAs that compared the mean scores on stereotypical traits across conditions. These resulted in the same main findings (see Appendix E for extensive details on the outcomes).
3 Results
Direct effects of populist blame attributions on populist attitudes and stereotypes
We expected that exposure to populist attributions of blame would activate populist attitudes and negative stereotypical perceptions toward the scapegoated elites (H1). Table 2 (in Appendix A) shows the results for the effects of populist communication on populist attitudes (see Table 1 for mean score comparison, and Appendix E for ANOVAs). The results indicate that exposure to different levels of blame attribution – to the political, corporate, or scientific elites – does not yield higher levels of populist attitudes compared to the absence of blame attribution. Likewise, populist attributions of blame – the combination of blame attribution and populist cues – has no significant effect on populist attitudes. Based on the mean score comparisons in Table 1 and the regression coefficients depicted in Table 2 (Model I), we find no support for H1a to H1c: Exposure to populist blame attribution does not activate populist attitudes.
The regression analyses summarized in Appendix A, Tables 3, 4, and 5 (Model I) report the effects of populist and non-populist blame attributions on positive stereotypical perceptions of the political, corporate, and scientific elites. The findings show that stereotypical portrayals toward the political elites are not activated by exposure to any of the populist or non-populist blame attribution conditions: There are no direct effects of isolated populist cues on negative stereotypes toward the elites (Tables 3, 4, and 5, Model I). However, the combined populist blame attribution frame does correspond to significantly more negative stereotypical perceptions of the political elites compared to the control condition (Table 3, Model I). The findings thus offer only limited support for H1a: Only when the political elites are scapegoated as part of a conspiracy with corporate and scientific elites, negative images toward politicians can be activated by populist communication.
Regarding H1b, the results do indicate that exposure to populist communication that attributes blame to corporations can activate more negative stereotypical perceptions of corporations and multinationals compared to the control condition (Table 4, Model I). Yet, this effect is only marginally significant. Non-populist blame attribution does not correspond to significantly more negative stereotypical perceptions. The findings offer only limited support for H1b: Populist attributions of blame do activate negative stereotypes toward these elites, but these effects are marginally significant, and only clearly pronounced when the corporations are blamed as part of a more encompassing conspiracy that assigns blame to all three elitist actors. We also see that attributing blame to politicians – even more so than blaming corporations – activates negative perceptions of corporate elites.
Turning to H1c, the results show that exposure to both populist and non-populist attributions of blame to scientists activate more negative perceptions of scientists among participants compared to the control condition (Appendix A, Table 5, Model I). Again, there is a spill-over effect: Blaming scientists in a populist way negatively affects perceptions toward corporate elites. Finally, the populist condition that assigns blame to all elites has a significant, negative effect on perceptions of scientists: Participants exposed to this populist blame frame rate scientists as significantly more negative than participants in the control condition. These findings offer support for H1c.
Taken together, our findings show that exposure to populist blame framing versus non-populist framing does not activate populist attitudes, although negative mental schemata of the political, scientific, and corporate elites are fostered when participants are exposed to populist blame frames. Hypothesis 1 finds partial support in the data: There are no direct effects of populist cues on populist attitudes, whereas populist cues can activate message-congruent negative perceptions toward corporations and scientists. Interestingly, negative stereotypes toward the political elites are only triggered by exposure to messages that blame corporate or scientific elites or attribute blame to all three elitist actors together.
Hypothesis 2 predicted that populist messages that attribute blame to all elites would be more effective than messages that only scapegoat the political, corporate, or scientific elites. The regression models summarized in Appendix A, Table 2 do not support this expectation for populist attitudes: Populist attitudes are not affected by exposing people to any of the populist cues. However, in support of H2, stereotypical out-group perceptions are affected most when participants were exposed to populist messages in which all elites were blamed (Tables 3, 4, and 5). More specifically, exposure to populist messages that emphasized a conspiracy between political, corporate, and scientific elites had the strongest effects on negative out-group perceptions on all three levels. Yet, non-populist messages that blamed these three elites did not have a significant effect on out-group representations, which demonstrates that the effectiveness of blame attribution was a combination of scapegoating and populist language, rather than simply pointing the finger at the elites.
Indirect effects of populist blame attributions
We further assessed if, and if so, how, the effects of populist scapegoating on populist attitudes and stereotypical perceptions are conditioned by prior levels of trust in the different elitist actors (H3). The findings depicted in Appendix A, Table 2 (Model III) show that the effects of populist blame attributions on populist attitudes are not different for participants with lower or higher levels of trust in the scapegoated elites. Although higher levels of trust in politicians, corporations, and scientists correspond to lower levels of populist attitudes (Table 2, Model II), there are no significant interaction effects between prior levels of trust in the elites and exposure to populist blame attribution (Table 2, Model III). This means that H3 is not supported.
Turning to stereotypical out-group perceptions, the results indicate that pre-existing levels of trust in the political elites moderate the effects of populist communication (see Table 3, Model III). In other words, the more people trust politicians, the less negative out-group perceptions of politicians are activated by exposure to populist communication. Thus, for participants that distrust the elites, exposure to a populist message that shifts blame to the political establishment versus a neutral message yields stronger negative perceptions than among the more trusting participants. The same interaction effect can be observed for populist messages that attribute blame to politicians, corporations, and scientists in a single message (Appendix A, Table 3, Model II). Although the effect is only marginally significant, lower levels of trust in the political elites also augment the effects of populist blame attribution to politicians on negative out-group perceptions of scientists (Appendix A, Table 5, Model III).
However, it should be noted that there are no significant two-way interaction effects between trust in corporate (Appendix A, Table 4, Model III) or scientific elites (Appendix A, Table 5, Model III) and the effects of exposure to populist blame attribution versus neutrally framed messages. All in all, these findings only provide limited support for H3: Although declining levels of trust in politicians may augment the impact of populist cues on negative stereotypes, this mechanism cannot be extrapolated to trust in corporate or scientific elites.
4 Discussion
Despite increasing scholarly attention to the effectiveness of communicating the ideational core of populism (e. g., Aalberg et al., 2017; Reinemann, Aalberg, Esser, Strömbäck, and de Vreese, 2017), we know little about the effects of populist messages that shift blame to non-political elites or simultaneously scapegoat different elites for causing the ordinary people’s problems – which may lead to conspiracy theories in populist discourse (e. g., Bergmann, 2019). Against this backdrop, this paper reports on an online experiment with a diverse sample of Dutch participants, in which we manipulated blame attribution on four levels: (1) the political, (2) the corporate, (3) the scientific, and (4) all three elites – and varied between “normal” blame attribution (holding elites accountable for national issues) and populist attributions of blame (emphasizing the antagonistic divide between ordinary people and corrupt elites).
The key findings illustrate that exposure to populist and non-populist blame attribution does not activate populist attitudes – even if pre-existing levels of (dis)trust in the scapegoated elites are taken into account. This contrasts earlier work that did find a significant effect of populist cues on populist attitudes (e. g., Bos et al., 2020; Hameleers et al., 2017). However, exposure to populist blame attributions does activate negative stereotypical schemata of political, corporate, and scientific elites. Yet, attributions of blame to political elites on their own do not activate negative images of political elites: Only the combination of blame attribution to all elites in a populist framework activates negative perceptions of political elites. In light of this, the two main findings of this experimental study can be summarized as follows: (1) Blame attributions need populist framing to have a significant impact on stereotypical out-group perceptions, and (2) messages that combine populist blame attributions to all elites have a stronger effect on out-group stereotypes than isolated blame attribution messages that emphasize a single scapegoat for the people’s problems.
We found support for a spill-over effect of attributing blame to political elites – the most “common” enemy in populist discourse. When politicians are blamed, people’s evaluations of corporate elites also become more negative. This spill-over effect can be explained as an association that is cognitively accessible among news users: Political elites may have corporate interests and corporate connections, which also means that causal attributions of blame to politicians may spill over to people’s beliefs about corporations. The same pattern is observed for blame attribution to scientists: Corporations are evaluated more negatively when scientists are blamed – which may again be due to a perceived similarity of different elites that are not acting independently of each other (i. e., scientists may also have corporate interests).
By experimentally distinguishing between non-populist and populist blame attributions, this experiment shows that the combination of assigning blame (a causal interpretation) and emphasizing a Manichean outlook on society (a moral evaluation) drives the effects of populist messages on stereotypical out-group perceptions. Reasoned from an emphasis framing approach (e. g., Entman, 1993), this experimental study thus demonstrates that two frame elements that give meaning to societal issues by highlighting causes and moral interpretations are together responsible for triggering similar frames on the audience-side (Cacciatore, Scheufele, and Iyengar, 2016).
Looking beyond research that investigated the effects of isolated populist cues (Bos et al., 2013; Matthes and Schmuck, 2017; Müller et al., 2017), we show that the centrality of conspiracy reasoning in populist communication (e. g., Bergmann, 2019; Silva et al., 2017) can make populist cues most persuasive. The “thin” ideology of populism on its own did not affect negative stereotypes to the political elites, but it did spill over to negative stereotypical images of scientists and corporations. However, when all elites were together framed as responsible for plotting a scheme against the ordinary people, the strongest negative stereotypes toward the politicians and corporate elites were cultivated. Yet, these messages needed to rely on the integration of blame shifting and populist cues.
This study bears some limitations that can be addressed in future research. First, we assessed the effects of populist messages in a single country and looked at a single issue. The effects may be different in national settings that offer a more or less favorable opportunity structure to credibly shift blame to the elites (e. g., Aalberg et al., 2017; Reinemann et al., 2017). Hence, attributing blame to corporations may be more credible in Southern European nations that have faced more far-reaching consequences of recent crises. Yet, we believe that the choice of our topic – which is equally relevant and salient across all countries – combined with the choice of a neutral news source makes the findings transferable to other settings as well. Second, we focused on two (attitudinal) dependent variables and on one single mechanism on the individual level (trust). As we demonstrate that populist attitudes and out-group stereotypes are not affected in the same way, future research may further explore to what extent different dependent variables can be affected by exposing people to populist blame attributions, for example, by looking at voting intentions, political engagement, emotions, or blame frames on the receiver side (frame alignment between populist messages and frames-in-mind). Another limitation concerns the omission of sources and different communicators in the stimuli. Populist messages can come from many different sources, such as directly from politicians, from (alternative) media platforms or ordinary citizens. Although we did not introduce this additional factor in our experiment, we believe that future research should explore the role of sources in the persuasiveness of populism and the conditional role of source support in particular (i. e., are populist messages most influential when they come from a liked source?).
We should also note that, although populism can be strongly related to accusations of fake news and disinformation (e. g., Egelhofer and Lecheler, 2019), we did not focus on the media as an out-group. As we believe that media hostility and fake-news accusations are relevant to consider, we suggest future research to explicitly incorporate blame attributions to media elites. We should also consider the fact that the three levels of elite actors distinguished here can mean different things for different people and that we also used some terms interchangeably in the survey experiment. For some, corporate elites may mainly consist of big companies and multinationals, whereas others think about managers or media institutions. The same may apply to scientists: Some may mostly consider scientists connected to political agendas, institutions, and salient environmental issues, whereas others perceive them as more related to an independent construction of knowledge. Future research may need to take into account the implication of both different wordings and personal interpretations.
We also suggest future research to delve deeper into the role of populist framing related to issues that are more versus less likely to be owned by populist actors. We now looked at a “most different” scenario where climate change was seen as a real problem – a statement that populist parties in Europe mostly disagree with. Future research may explore to what extent effects of populist framing are contingent upon the resonance of the problem statement with populist positions (i. e., immigration versus climate change).
Despite these limitations, our experimental study offers first insights into the isolated and combined effects of populist blame attribution to multiple elitist scapegoats held accountable for causing the people’s problems. Although we did not find support for a strong impact of populist blame attributions, under some conditions, populist scapegoating can affect the evaluations of elites on the scientific, corporate, and political levels.
References
Aalberg, T., Esser, F., Reinemann, C., Strömbäck, J., & de Vreese, C. H. (Eds.) (2017). Populist political communication in Europe. London, UK: Routledge.Search in Google Scholar
Akkerman, A., Mudde, C., & Zaslove, A. (2014). How populist are the people? Measuring populist attitudes in voters. Comparative Political Studies, 47(9), 1324–1353. doi:10.1177/001041401351260010.1177/0010414013512600Search in Google Scholar
Albertazzi, D., & McDonnell, D. (2008). Twenty-first century populism. New York, NY: Palgrave Macmillan.10.1057/9780230592100Search in Google Scholar
Barkun, M. (2003). A culture of conspiracy: Apocalyptic visions in contemporary America. Berkeley: University of California Press.Search in Google Scholar
Bergmann, E. (2019). Conspiracy & populism: The politics of misinformation. Cham, Switzerland: Palgrave Macmillan.10.1007/978-3-319-90359-0Search in Google Scholar
Bos, L., Schemer, C., Corbu, N., Hameleers, M., Andreadis, I., Schulz, A., …, & Fawzi, N. (2020). The effects of populism as a social identity frame on persuasion and mobilization: Evidence from a 15‐country experiment. European Journal of Political Research. doi:10.1111/1475-6765.1233410.1111/1475-6765.12334Search in Google Scholar
Bos, L., Van Der Brug, W., & de Vreese, C. H. (2013). An experimental test of the impact of style and rhetoric on the perception of right-wing populist and mainstream party leaders. Acta Politica, 48(2), 192–208. doi:10.1057/ap.2012.2710.1057/ap.2012.27Search in Google Scholar
Busby, E. C., Gubler, J. R., & Hawkins, K. (2019). Framing and blame attribution in populist rhetoric. The Journal of Politics, 81(2), 616–630.10.1086/701832Search in Google Scholar
Cacciatore, M. A., Scheufele, D. A., & Iyengar, S. (2016). The end of framing as we know it … and the future of media effects. Mass Communication and Society, 19(1), 7–23. doi:10.1080/15205436.2015.106881110.1080/15205436.2015.1068811Search in Google Scholar
Dixon, T. L. (2008). Crime news and racialized beliefs: Understanding the relationship between local news viewing and perceptions of African Americans and Crime. Journal of Communication, 58, 106–125. doi:10.1111/j.1460-2466.2007.00376.x10.1111/j.1460-2466.2007.00376.xSearch in Google Scholar
Egelhofer, J. L., & Lecheler, S. (2019). Fake news as a two-dimensional phenomenon: A framework and research agenda. Annals of the International Communication Association, 43(2), 97–116. doi:10.1080/23808985.2019.160278210.1080/23808985.2019.1602782Search in Google Scholar
Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51–58. doi:10.1111/j.1460-2466.1993.tb01304.x10.1111/j.1460-2466.1993.tb01304.xSearch in Google Scholar
Fenster, M. (2008). Conspiracy theories: Secrecy and power in American culture. Minneapolis: University of Minnesota Press.Search in Google Scholar
Gamson, W. A. (1992). Talking politics. Cambridge: Cambridge University Press.Search in Google Scholar
Hameleers, M., Bos, L., & de Vreese, C. H. (2017). ‘They did it’: The effects of emotionalized blame attribution in populist communication. Communication Research, 44(6), 870–900. doi:10.1177/009365021664402610.1177/0093650216644026Search in Google Scholar
Hameleers, M., Bos, L., Fawzi, N., Reinemann, C., Andreadis, I., Corbu, N., …, Weiss-Yaniv, N. (2018). Start spreading the news: A comparative experiment on the effects of populist communication on political engagement in sixteen European countries. The International Journal of Press/Politics, 23(4), 517–538. doi:10.1177/194016121878678610.1177/1940161218786786Search in Google Scholar
Hameleers, M., & Minihold, S. (2020). Constructing discourses on (un)truthfulness: Attributions of reality, misinformation, and disinformation by politicians in a comparative social media setting. Communication Research. doi:10.1177/009365022098276210.1177/0093650220982762Search in Google Scholar
Hornsey, M. J., Harris, E. A., & Fielding, K. S. (2018). Relationships among conspiratorial beliefs, conservatism and climate scepticism across nations. Nature Climate Change, 8(7), 614–620.10.1038/s41558-018-0157-2Search in Google Scholar
Jagers, J., & Walgrave, S. (2007). Populism as political communication style: An empirical study of political parties’ discourse in Belgium. European Journal of Political Research, 46(3), 319–345. doi:10.1111/j.1475-6765.2006.00690.x10.1111/j.1475-6765.2006.00690.xSearch in Google Scholar
Krämer, B. (2018). How journalism responds to right-wing populist criticism. In K. Otto & A. Köhler (Eds.), Trust in media and journalism (pp. 137–154). Wiesbaden: Springer VS.10.1007/978-3-658-20765-6_8Search in Google Scholar
Matthes, J., & Schmuck, D. (2017). The effects of anti-immigrant right-wing populist ads on implicit and explicit attitudes: A moderated mediation model. Communication Research, 44(4), 556–581. doi:10.1177/009365021557785910.1177/0093650215577859Search in Google Scholar
Mede, N. G., & Schäfer, M. S. (2020). Science-related populism: Conceptualizing populist demands toward science. Public Understanding of Science, 29(5), 473–491.10.1177/0963662520924259Search in Google Scholar
Mols, F. (2012). What makes a frame persuasive? Lessons from social identity theory. Evidence & Policy: A Journal of Research, Debate and Practice, 8(3), 329–345. doi:10.1332/174426412X65405910.1332/174426412X654059Search in Google Scholar
Mudde, C. (2004). The populist zeitgeist. Government and Opposition, 39(4), 542–564. doi:10.1111/j.1477-7053.2004.00135.x10.1111/j.1477-7053.2004.00136.xSearch in Google Scholar
Müller, P., Schemer, C., Wettstein, M., Schulz, A., Wirz, D. S., Engesser, S., & Wirth, W. (2017). The polarizing impact of news coverage on populist attitudes in the public: Evidence from a panel study in four European democracies. Journal of Communication, 67(6), 968–992. doi:10.1111/jcom.1233710.1111/jcom.12337Search in Google Scholar
Polletta, F., & Jasper, J. M. (2001). Collective identity and social movements. Annual Review of Sociology, 27(1), 283–305. doi:10.1146/annurev.soc.27.1.28310.1146/annurev.soc.27.1.283Search in Google Scholar
Reinemann, C., Aalberg, T., Esser, F., Strömbäck, J., & de Vreese, C. H. (2017). Populist political communication: Toward a model of its causes, forms, and effects. In T. Aalberg, F. Esser, C. Reinemann, J. Strömbäck, & C. H. de Vreese (Eds.), Populist political communication in Europe (pp. 12–28). New York: Routledge.10.4324/9781315623016Search in Google Scholar
Richey, S. (2012). Campaign advertising and the stimulation and activation of the authoritarian personality. Political Communication, 29(1), 24–43. doi:10.1080/10584609.2011.61687410.1080/10584609.2011.616874Search in Google Scholar
Schulz, A., Müller, P., Schemer, C., Wirz, D. S., Wettstein, M., & Wirth, W. (2017). Measuring populist attitudes on three dimensions. International Journal of Public Opinion Research. doi:10.1093/ijpor/edw03710.1037/t74687-000Search in Google Scholar
Silva, B. C., Vegetti, F., & Littvay, L. (2017). The elite is up to something: Exploring the relation between populism and belief in conspiracy theories. Swiss Political Science Review, 23(4), 423–443. doi:10.1111/spsr.1227010.1111/spsr.12270Search in Google Scholar
Taggart, P. (2000). Populism: Concepts in the social sciences. Buckingham, UK: Open University Press.Search in Google Scholar
Van Aelst, P., Strömbäck, J., Aalberg, T., Esser, F., de Vreese, C. H., Matthes, J., …, Stanyer, J. (2017). Political communication in a high-choice media environment: A challenge for democracy? Annals of the International Communication Association, 4(1), 3–27. doi:10.1080/23808985.2017.128855110.1080/23808985.2017.1288551Search in Google Scholar
Vasilopoulou, S., Halikiopoulou, D., & Exadaktylos, T. (2014). Greece in crisis: Austerity, populism and the politics of blame. JCMS: Journal of Common Market Studies, 52, 388–402.10.1111/jcms.12093Search in Google Scholar
Waisbord, S. (2018). The elective affinity between post-truth communication and populist politics. Communication Research and Practice, 4, 17–34. doi:10.1080/22041451.2018.142892810.1080/22041451.2018.1428928Search in Google Scholar
Wirz, D. S. (2018). Persuasion through emotion? An experimental test of the emotion-eliciting nature of populist communication. International Journal of Communication, 12, 1114–1138.Search in Google Scholar
© 2023 bei den Autorinnen und Autoren, publiziert von De Gruyter.
Dieses Werk ist lizensiert unter einer Creative Commons Namensnennung 4.0 International Lizenz.
