This is an open access article distributed under the terms of the Creative Commons
Attribution License (Throughout the 20th century, radical political ideologies served as the propeller of tragic events that have left a deep mark in human history (Baumeister, 1997; Midlarsky, 2011). Movements of the radical right, such as Fascism and Nazism, were ideologically driven to commit unimaginable crimes against humanity which resulted in the death of millions. On the left, communist regimes in the former Soviet Union and its satellite states, as well as in numerous Asian countries were responsible for comparable atrocities. Even though there are substantial differences between these ideologies, political psychologists have noted that extreme left- and right-wing ideologies may share a similar underlying psychological pattern (Greenberg & Jonas, 2003). In order to assess whether rigid devotion to extreme political ideology on both the left and the right corresponds to a similar psychological understanding of the world, it is necessary to compare radical individuals and test if their comprehension of the functioning of society is identical.
Although extremist regimes of the past are in many ways hard to compare to political actors that are considered relatively extreme in the current political landscape (e.g., populism), in present times the extreme left and right seem comparable on various dimensions. For instance, electoral support for political parties located at both the far-left and far-right of the ideological spectrum is often explained as a vote of protest against mainstream parties, which are usually government incumbents. Political cynicism and distrust in institutions, which typically increase in times of economic crises, are frequently associated with the better electoral performance of extremist parties (Miller, 1974; Rydgren, 2005). The core proposition of the present paper is that the political extremes have a more pessimistic outlook on society than moderates, as reflected in a range of important socio-political variables. For instance, it has been noted that both political extremes are—more so than moderates—prone to attribute events in society to secret and malevolent activities by power holders. More specifically, research suggests that political extremism predicts a general susceptibility to conspiracy theory belief (van Prooijen, Krouwel, & Pollet, 2015; see also Inglehart, 1987).
Economic conditions also play an important role in the acceptance of radical political ideology. The relationship between electoral outcomes and economic performance has been intensively debated within social science disciplines. Most studies on the topic examine and test the basic assumptions of the theory – whether macroeconomic conditions are related to electoral preferences (Downs, 1957; Fiorina, 1981; Kinder & Kiewiet, 1981; Lewis-Beck, 1988). These scholars argue that economic factors have a significant influence on political preferences, considering that voters hold the government at least partially responsible for the functioning of the country’s economy and their own well-being and that economic growth or decline influences the level of support for established political authorities.
Clearly, the relationship between conspiracy belief, political distrust and negative economic evaluations on the one hand and political ideology on the other is twofold – with the deterioration of socioeconomic conditions the degree of political extremism and distrust within a society increases (Miller, 1974; Rydgren, 2005), while political extremists are normally distrustful and critical towards the political system and its performance (Greenberg & Jonas, 2003).
By using data from a Swedish online survey, this study offers empirical analyses of the predictions that people who adhere to extreme political ideologies, of both the left and the right, are similar in the sense that they are more pessimistic about the functioning of society as reflected in increased conspiracy theory belief, negative economic evaluations and lower interpersonal trust and trust in the political system. Previous research has revealed that conspiracy belief could be seen as stemming from alienation and distrust (Abalakina-Paap, Stephan, Craig, & Gregory, 1999), with greater economic pessimism also being associated with conspiracy perceptions (Radnitz, 2012). The present study aims to test whether there is an association between conspiracy belief, negative economic perceptions and distrust in the case of Sweden. We use economic evaluations, conspiracy beliefs, and various indicators of trust together in order to test if there is a common broad mindset of radical individuals that is reflected on multiple interrelated dimensions. Acknowledging that extremism is relative, and radical political ideology in contemporary Swedish society can hardly be compared to early 20th century extreme political beliefs, we are nevertheless interested in testing whether it is possible predict certain responses of people who are relatively extreme as compared to people who are relatively moderate in their political position. To our knowledge this also constitutes the first empirical study of public opinion on conspiracy beliefs in Sweden.
Conspiracy Beliefs and Extreme Political Ideology
Political ideology constitutes a common belief and value system expressed in the way individuals view and react to their social environment (Jost et al., 2009; Knight, 2006). Erikson and Tedin (2003, p. 64), define ideology as a “set of beliefs about the proper order of society and how it can be achieved”. Specific ideologies tend to unify and disseminate widely (though not necessarily universally) shared beliefs, policy preferences and values of specific groups within society (Knight, 2006). Ideologies aim at describing and interpreting the functioning of society by relying on assertions or assumptions about “human nature, historical events, present realities, and future possibilities” (Jost et al., 2009, p. 309). Next to this, ideologies offer a vision for how societies should function and offer specific means for accomplishing this vision.
Earlier research (e.g. Adorno et al., 1950) operationalizes ideology as a unidimensional concept ranging from liberal to conservative (particularly in the US) or more generally from left to right. However, a large volume of work suggests that such a simplified conceptualization could easily overlook underlying aspects of ideological heterogeneity (Jost et al., 2009; Treier & Hillygus, 2009). The findings of Feldman and Johnston (2014) show a differentiation between the cultural and economic dimension of political ideology, which is crucial for understanding ideology in modern societies. In this study we test whether these different dimensions of ideology converge in terms of producing similar effects regarding conspiracy belief, economic evaluations and trust. With regard to ideological extremism, there has been much debate over its exact definition. Mudde’s (1996) differentiation between „radical“ and „extreme“ entails a distinction in terms of attitudes towards democracy – while radical political actors acknowledge the legitimacy of democracy and its institutions, political extremist aim at profoundly changing the system of governance by rooting out the democratic regime.
A common feature of political extremists is that their perceptions of the world are characterized with rigid self-conviction. Specifically, political extremism is related to a black-and-white style of thinking in which social actors and processes fall under distinctions such as good or evil, positive or negative etc. (Greenberg & Jonas, 2003). Hardin (2002) notes that extremists are “epistemologically crippled”, in the sense that the main sources of information about political issues they trust, come from their own milieu, whereas outside sources that offer more nuanced perspective tend to be ignored. Put differently, extremists only receive or trust information about socio-political issues that was provided by other extremists. This crippled epistemology is reflected in the rigid closed-mindedness of extremists, expressed by the ideologically-driven belief that the only solutions to societal problems are simple and lie in the implementation of their preferred policies (Fernbach, Rogers, Fox, & Sloman, 2013; see also Midlarsky, 2011). Such a conviction functions as an instrument to cope with feelings of uncertainty and fear by constructing a more comprehensible and predictable understanding of the world (Neuberg & Newsom, 1993).
Previous research suggests that extremism is grounded in feelings of uncertainty and fear. Somewhat paradoxically, uncertainty in one domain of life results in an increase of certainty, reflected in extreme convictions, in other (frequently political) domains (McGregor, 2006; McGregor & Marigold, 2003). Thus, political extremism is related to a structured way of thinking aimed at making sense of society and the way it functions. Conspiracy beliefs relate to similar mindset as they give simple and structured answers for difficult questions (Kossowska & Bukowski, 2015; see also Marchlewska, Cichocka, & Kossowska, 2017). Indeed, it has been noted that conspiracy theories address a range of basic human needs, including epistemic needs (Douglas, Sutton, & Cichocka, in press). These considerations suggest that political extremism and conspiracy theories are driven by comparable sense-making processes, and are hence heavily intertwined. In this study we test whether there is a fundamental relationship between belief in conspiracy theories and political extremism.
We define conspiracy beliefs as suspicions that a group of individuals are involved in secret agreement, attempting to achieve hidden goals which are generally regarded as unlawful or malevolent (Swami, Chamorro-Premuzic, & Furnham, 2010; Zonis & Joseph, 1994). In most cases, such conspiracies are considered as a plot of either powerful groups (such as politicians, CEOs, bankers) or minorities (such as Muslims or Jews). Prominent examples of contemporary conspiracy beliefs pertain to issues such as climate change (Jolley & Douglas, 2014), the financial crisis and the common belief about the corrupt and nocuous nature of decision-makers, among others. Although there is a wide range of possible conspiracy theories, individuals who believe in one conspiracy theory are more likely to believe in other conspiracy theories (Lewandowsky, Oberauer, & Gignac, 2013; Swami et al., 2011, 2013). In fact, previous research has revealed that even conspiracy theories that are mutually exclusive (such as believing that Princess Diana was murdered versus believing that she staged her own death) are positively correlated (Wood et al., 2012). This implies that, even if various conspiracy theories differ significantly in content, they are driven by similar underlying psychological processes.
Numerous scholars have described how these processes are characterized by an aspiration for understanding threatening societal events by explaining them in a clear and unambiguous manner. Earlier contributions on the topic (e.g. Hofstadter, 1966) have already implied that conspiracy beliefs are intended to provide causal explanations for complex and distressing social events, an observation that was reflected upon theoretically by numerous authors (e.g., Bale, 2007; Clarke, 2002; Douglas et al., in press). Empirical findings confirm these ideas. For instance, experimental conditions that induce a lack of control or uncertainty increase belief in conspiracies as compared to other conditions (Newheiser, Farias, & Tausch, 2011; Sullivan, Landau, & Rothschild, 2010; van Prooijen & Acker, 2015; van Prooijen & Jostmann, 2013; Whitson & Galinsky, 2008). This implies that conspiracy theories provide epistemic clarity by explaining events through an array of explicit presumptions about the functioning of the world.
In sum, extreme political beliefs and conspiracy beliefs are a subject to similar underlying motivations. Consistently, Inglehart (1987) observed an association between extreme political positions and trust in the legal system, which he interpreted as diagnostic for conspiracy thinking. More recent studies using individual level data from the United States (US) and the Netherlands find that extreme political ideology, at either side of the political spectrum, is positively associated with an increased tendency to believe in conspiracy theories (van Prooijen, Krouwel, & Pollet, 2015). The current contribution aims to test this prediction in the Swedish context. Following this logic we arrive at our first hypothesis:
H1: Political ideology has a quadratic relation with conspiracy beliefs, such that the political extremes are more likely to endorse such beliefs than political moderates.
Economic Perceptions and Extreme Political Ideology
The relationship between electoral outcomes and economic performance has been intensively debated within the social science disciplines in general, and political science in particular. Numerous authors seeking to explain why incumbents fail to get re-elected examine the link between factors which have a profound impact on voters’ life quality such as the management of the economy on the one hand, and the electoral performance of incumbent leaders and their administrations on the other.
The theory of economic voting has been a subject of interest for numerous studies which deal with incumbent approval. Most of them examine and test the basic assumptions of the theory that macroeconomic conditions predict electoral preferences (e.g. Downs, 1957; Duch & Stevenson, 2008; Fiorina, 1981; Kinder & Kiewiet, 1979; Lewis-Beck, 1988). In a recent review, Lewis-Beck and Stegmaier (2007) estimate that there are more than 400 published articles on economic voting and that the vast majority of these studies confirm the basic expectation of the theory. These scholars argue that economic factors have a significant influence on political preferences, considering that voters hold the government at least partially responsible for the functioning of the country’s economy and their own well-being and that economic growth or decline influences the level of support for the established political authorities. An abundant number of studies examine the effect of objective economic conditions or subjective economic assessments on electoral preferences and government approval in different regions of the world (Harper, 2000; Lewis-Beck & Stegmaier, 2007).
The seminal work of Downs (1957) is among the first to relate voting behavior to economic outcomes. His main assumption is that voters are rational and will choose a party to vote for after estimating which one would provide them the greatest economic benefits if it wins the election. Most current studies, however, rather depart from the basic reward-punishment of economic voting theory: “when economic conditions are bad, citizens vote against the ruling party” (Lewis-Beck, 1991, p. 2). While many scholars use objective economic indicators to assess whether the bad shape of the economy results in a decline of electoral support for incumbents (e.g. Carlsen, 2000; Clarke et al., 2000; Harper, 2000; Nannestad & Paldam, 2000), others take advantage of opinion surveys, which ask respondents for their own evaluation of the economy and analyze how it relates to government support (Duch & Stevenson, 2008; Feldman, 1984; Fiorina, 1981; Kiewiet & Rivers, 1984; Kinder & Kiewiet, 1979; Lewis-Beck, 1988). Many studies examining economic voting based on election surveys use four different dimensions: personal versus collective (usually labeled “ego-tropic” versus “socio-tropic”); retrospective versus prospective; simple versus mediated and cognitive versus affective (Lewis-Beck, 1986, p. 317). During the last decades, however, most individual level economic voting studies have focused on the socio-tropic and retrospective evaluations, which have become the standard item for economic voting models of voting behavior. The questions included in our survey allow for using retrospective and prospective socio-tropic economic evaluations as dependent variables, as well as evaluations of the current economic situation in Sweden.
Although few political science theories have been applied so frequently, the theory of economic voting has to our knowledge not been related to political extremism. Currently, the rise of radical right-wing and left-wing political parties has intensified simultaneously with the deepening of the economic crisis in Europe. Historically, distressful economic situations have often preceded revolutions or electoral victories of extreme politicians. These observations are consistent with psychological theories and findings assuming a link between uncertainty and radicalization (McGregor, 2006; see also Hogg, Meehan, & Farquharson, 2010; van Prooijen, Krouwel, Boiten, & Eendebak, 2015). Moreover, pessimistic economic views and low-income levels could be associated with conspiracy belief (Radnitz, 2012).
Therefore, we expect the increase in extreme political beliefs and support for radical political parties to be related to the economic and societal crisis, and ensuing feelings of uncertainty and insecurity. We expect extreme ideology to be positively related to unfavorable economic perceptions, expressed in the second hypothesis:
H2: Political ideology has a quadratic relation with economic evaluations, such that the political extremes are more likely to have negative economic assessments than political moderates.
Political Trust and Extreme Political Ideology
Interest in trust as a concept has produced voluminous literature in the social sciences and philosophy throughout the 1990s (Hardin, 1996; Misztal, 1996; Warren, 1999). According to Newton (1999, p. 179) political trust is being influenced by second-hand sources such as the media. The general meaning of political trust is seen as “citizens’ assessments of the core institutions of the polity” (Lipset & Schneider, 1983; Zmerli, Newton, & Montero, 2007, p. 41). Thus, being politically trustful means that one evaluates political institutions and politician’s performance positively. This entails reliability, legitimacy and transparency in their political decisions and acceptance of different views. Confidence in institutions and trust in politicians translates in the belief that they have a positive influence on the state and its citizens, that they treat everybody equally and with fairness, and are not detrimental to any particular societal group. In terms of behavior, political trust plays a decisive role in citizen’s adherence to governmental demands (Levi, 1997; Scholz & Lubell, 1998), which in turn fosters incumbent legitimacy. Moreover, political trust affects citizen’s propensity to vote for opposition candidates (Hetherington, 1999).
Political distrust is often strongly correlated with unfavorable policy evaluations. While it has been acknowledged that extreme ideological left and right positions correspond with lower trust levels (Inglehart, 1987), whether right-wing or left-wing individuals are “more distrustful varies from one country to another” (Newton, 1999, p. 181). Studies in the field of social psychology have shown that extreme conservative positions, for example, can come as a consequence of anxieties associated with uncertainties and fears and enforced by dogmatic political comprehensions, authoritarian tendencies and intolerance towards dissenting individuals (Jost et al., 2003). However, further empirical research has shown that some of these characteristics may need to be attributed to both left- and right-wing extreme ideological rigidity, as for instance suggested by findings that in the former Soviet Union, authoritarian tendencies are associated with increase in left-wing orientation (McFarland et al., 1992). Indeed, a recent large-scale sample in the Netherlands reveals that both the left and right extremes are more fearful of socio-economic developments than political moderates (van Prooijen, Krouwel, Boiten, & Eendebak, 2015).
Moreover, those who support the parties in power are more likely to express higher levels of political trust than those who voted for opposition parties. Thus, besides socio-economic factors, also political factors appear to cause political distrust. Whereas right-wing individuals are more likely to express political distrust when the Left is in power, left-wing ones are prone to trust less when the Right governs. If the government is relatively centrist, both the radical left and right individuals are likely to become less trustful (Miller, 1974). Aberbach and Walker (1970, p. 1202) argue that a distrustful person “should be more disposed to take part in violent activities or to endorse radical interpretations of social ills than a trusting one”.
According to political culture studies (Almond & Verba, 1989), countries with high levels of trust achieve better economic development and higher levels of support for democracy among their citizens. In their view, the reason why less developed countries have difficulties improving their economic performance and political system is their low levels of trust (Wiarda, 1992). Following this notion, one can argue that the political culture embedded in a society affects its political and economic achievements. Thus, if interpersonal and political trust is low in a given polity, there will be an increasing possibility for authoritarian rule which would impede positive political and economic development. Following this logic we present our third hypothesis:
H3: Political ideology has a quadratic relation with trust in politicians and interpersonal trust, such that the political extremes are likely to be less trustful than political moderates
Data and Method
In the previous sections, we proposed three hypotheses suggesting that individuals’ extreme political positions predict a higher likelihood of conspiracy theory beliefs, more negative sociotropic economic evaluations and finally, lower levels of trust in politicians on the one hand, and in people in general on the other. In order to test these hypotheses, we utilize survey data comprised of 3958 respondents collected in June and August 2013. The participants were recruited to an online panel when they took part in a 2010 online voting advice application (VAA) on the website of Aftonbladet – one of the most widely circulated Swedish newspapers, and voluntarily enclosed their e-mail addresses for future online surveys. Online VAA opt-in surveys generate non-probability samples, so results cannot be generalized to the total population. Datasets collected online generally suffer from problems of under-coverage and self-selection that can potentially bias estimates (Bethlehem, 2010; Hooghe & Teepe, 2007). However, there are also benefits to opt-in web surveys. First, the VAA for the 2010 Swedish election allowed us to gather unparalleled amounts of data from more than 400 thousand Swedish voters, combining wide-ranging information about their political behaviour, opinions and background characteristics. Another benefit is that computerized self-administration reduces measurement error relative to other modes of data collection, increasing both the level of reporting and the report accuracy of opinions and attitudes compared to more “conventional” surveys (Kreuter et al., 2008; Sakshaug et al., 2010). Moreover, online survey questions are answered more accurately since they are answered in private, and thus, free of peer pressure introduced by interviewers (Olson, 2006).
Left-right political ideology was measured as an eleven-point scale question which asked the respondents to indicate their political orientation on a scale ranging from 0 (far left) to 10 (far right). To test if ideological placement has a quadratic effect on conspiracy belief, economic evaluations and trust we calculated the product of this variable with itself after mean-centering it. In line with that, we also include a set of control variables concerning respondents’ background – age, gender and educational attainment. We expect to unravel a quadratic effect that would indicate that both the left and right political extremities are more prone to believe in conspiracy theories; more likely to have negative evaluations of the economy, and to be less trustful.
The first set of dependent variables consists of five conspiracy theory belief variables measured on a seven-point scale (for full wording of each variable used in this study see Table A.2 in the Appendix below). Moreover, in order to assess if there is also a general effect, which preponderates the individual dependent variables, we also include a variable which is composed of all five conspiracy theory variables compiled together (Cronbach’s α = .74).
The second set of variables is related to a particularly Swedish case and also permeates in a conspiracy theory debate: the unresolved murder of the social-democratic Prime Minister Olof Palme, who was assassinated in 1986. On a five-point scale ranging from 1 (very unlikely) to 5 (very likely), in separate questions, respondents were asked whether two organizations were responsible for the murder of the former Swedish leader. The possibilities include the Kurdistan Worker’s Party (Partiya Karkeren Kurdistan, PKK) and the Swedish police. We expect to reveal positive curvilinear effects which would indicate similar opinions of both extreme left-wing and extreme right-wing respondents.
The third set of variables deals with Swedes’ general economic perceptions. It includes three questions asking the respondents to evaluate the present Swedish economic situation, as well as retrospectively and prospectively. All three economic evaluations use five point scales where 1 means “very bad”/”has gotten much worse”/”will get much worse” and 5 means “very good”/”has gotten much better”/”will get much better”. Here, we expect to find a pattern in which extreme political orientations correspond with negative economic evaluations.
The fourth variable set contains dependent variables which deal with respondents’ trust in politicians, and people in general (interpersonal trust). The ‘trust in politicians’ question has a 4-point answer scale ranging from (1) do not trust them at all to 4 (trust them a lot). The ‘interpersonal trust’ question has an 11-point answer scale ranging from 0 (cannot trust people at all) to 10 (can trust people a lot).
A post-stratification weight accounting for sex, age and education was implemented to adjust for demographic deviations from the population (2 x 3 x 2 = 12 cells; ratio between highest and lowest weight: 9.8).
Results
We ran several hierarchical regression analyses with different sets of dependent variables, each corresponding to one of the aforementioned hypotheses. We test two separate models in order to assess if there are linear as well as curvilinear (quadratic) relationships between political ideology as the independent variable and conspiracy beliefs, economic perceptions and trust in politicians and people in general as the dependent variables. While the linear effects inform us whether the left or the right are more likely to agree or disagree with the questions or statements we use as independent variables, the quadratic effects reveal if there is a curvilinear pattern, indicating whether respondents at the extreme left and right are more strongly inclined to believe in conspiracy theories, to be less satisfied with economic conditions, and to be less trustful than politically moderate respondents. For a graphical representation of the linear and quadratic effects of political ideology alone on each of the dependent variables, please see the Appendix.
Conspiracy Beliefs
The results of hierarchical regression analyses with conspiracy belief statements compiled in a single dependent variable are presented in Table 1. The results indicate that political ideology has both linear and quadratic statistically significant effects on our conspiracy belief measure. In general, Swedes who associate with a far-left ideology are more likely to endorse conspiracy theories than those with a far-right ideology, albeit both groups are more prone to conspiracy beliefs than centrists, expressed by the statistical significance of the quadratic effect. The control variables all have negative statistically significant effects: older Swedes, males and higher educated people are less likely to believe in conspiracy theories.
Table 1
Determinants Explaining Belief in Conspiracy Theories (Index)
| Variables | R | R2 | ΔR2 | ΔF | df | β |
|---|---|---|---|---|---|---|
| Step 1 | .21 | .043 | .043*** | 38.72 | 3450 | |
| Political ideology | -0.04* | |||||
| Age | -0.09*** | |||||
| Male | -0.05* | |||||
| Education | -0.19*** | |||||
| Step 2 | .212 | .045 | .002* | 6.45 | 3449 | |
| Political ideology | -0.50** | |||||
| Age | -0.09*** | |||||
| Male | -0.05** | |||||
| Education | -0.19*** | |||||
| Quadratic term | 0.04* |
*p < . 05. **p < .01. ***p < .001.
We also test each of the conspiracy theories items separately, in order to assess whether the quadratic effect is persistent across all of them (see Table 2). For all but one statement, there is a negative statistically significant linear effect (p < .001), which indicates that Swedes who associate with far-right ideology are less likely to believe in conspiracies than their left-wing counterparts. The only exception is the statement which asserts that left politicians keep the population poor on purpose, with which a much higher proportion of far-right Swedes agree. Even though the quadratic term is significant, the effect is rather weak and is likely driven by only one extreme. Looking at Figure 6 in the Appendix, it is clear that individuals on the left are much less likely to agree with this proposition. In a nutshell, the results reported in Table 1 and 2 confirm our first hypothesis: politically extreme individuals are more likely to believe in conspiracy theories.
Table 2
Political Ideology as Determinant of Conspiracy Theory Belief
| Dependent variable | Model | R | R2 | ΔR2 | ΔF | df | β |
|---|---|---|---|---|---|---|---|
| Financial crisis planned by bankers | 3515 | ||||||
| Linear | .182 | .033 | .033*** | 120.91 | -0.19*** | ||
| Quadratic | .189 | .036 | .003** | 8.65 | 0.05** | ||
| Deliberate disesase spreading | 3517 | ||||||
| Linear | .096 | .009 | .009*** | 33.05 | -0.1*** | ||
| Quadratic | .104 | .011 | .002* | 5.37 | 0.04* | ||
| Politicians paid to wage wars | 3505 | ||||||
| Linear | .213 | .046 | .046*** | 167.24 | -0.22*** | ||
| Quadratic | .219 | .048 | .002** | 8.42 | 0.04** | ||
| Politicians linked to organized crime | 3507 | ||||||
| Linear | .100 | .010 | .010*** | 35.17 | -0.11*** | ||
| Quadratic | .108 | .012 | .002* | 6.19 | 0.19* | ||
| Left politicians keep people poor | 3496 | ||||||
| Linear | .341 | .117 | .117*** | 461.45 | 0.33*** | ||
| Quadratic | .347 | .120 | .004*** | 15.11 | 0.06*** | ||
*p < .05. **p < .01. ***p < .001.
Table 3 presents results that deal with a conspiracy theory specific to Sweden – the assassination of Olof Palme. The results reveal a statistically significant curvilinear pattern in both propositions for who is responsible for the murder.
Table 3
Determinants Explaining Conspiracy Belief Regarding the Murder of Olof Palme
| Variables | R | R2 | ΔR2 | ΔF | df | β |
|---|---|---|---|---|---|---|
| PKK was responsible for the murder of Olof Palme | ||||||
| Step 1 | .245 | .060 | .060*** | 51.63 | 3234 | |
| Political ideology | 0.09*** | |||||
| Age | -0.18*** | |||||
| Male | -0.13*** | |||||
| Education | -0.13*** | |||||
| Step 2 | .251 | .063 | .003** | 10.36 | 3233 | |
| Political ideology | 0.09*** | |||||
| Age | -0.19*** | |||||
| Male | -0.13*** | |||||
| Education | -0.13*** | |||||
| Quadratic term | -0.06*** | |||||
| The Police was responsible for the murder of Olof Palme | ||||||
| Step 1 | .224 | .050 | .050*** | 43.76 | 3312 | |
| Political ideology | -0.15*** | |||||
| Age | 0.06** | |||||
| Male | 0.04* | |||||
| Education | -0.14*** | |||||
| Step 2 | .228 | .052 | .002* | 6.2 | 3311 | |
| Political ideology | -0.16*** | |||||
| Age | 0.07** | |||||
| Male | 0.04* | |||||
| Education | -0.13*** | |||||
| Quadratic term | 0.04* | |||||
*p < .05. **p < .01. ***p < .001.
Olof Palme was a Swedish Social Democratic politician who served as prime minister of Sweden for two periods, 1969 – 1976 and 1982 – 1986, until he was killed on a street in Stockholm. The investigation that followed the murder is the third largest crime investigation in the world in terms of the volume of documentation (Åsard, 2006).
Palme was a politically divisive character in Sweden, as well as internationally, due to his outspoken criticism of American imperialism and colonialism in general, while also being an anti-communist. As a result he drew the ire of extremists on both ends of the political spectrum. This, coupled with the fact that the assassination was never officially solved and no suspect was ultimately incarcerated has led to various conspiracy theories. The most prominent theories imply that the PKK, an ultra left-wing organization that had carried out political assassinations in Europe, or a group of Swedish police officers with extreme right sympathies were responsible.
The positive statistically significant linear effect of the first proposition indicates that right-wing Swedes are more likely to assume that the PKK was involved in the assassination. Possibly due to an ideological bias, those on the right are more likely to consider that a left-wing organization is the likely culprit. The negative coefficient of the quadratic test, on the other hand, indicates that extreme political ideology corresponds to a decrease in the belief that the PKK was responsible for the political murder. All control variables also have statistically significant, negative effects: lower educated people, females, and the younger are more likely to agree that the PKK committed the political murder.
A positive quadratic statistically significant effect reveals that extreme political ideology correlates with increase in the belief that the Swedish police was involved in perpetrating the liquidation of the former Prime Minister. The strong negative linear effect of political ideology, however, shows that far right individuals are in general less likely to believe the proposition compared to far left ones. Similarly to the PKK item above, this is possibly caused by an ideological bias, where left-wing respondents being more inclined to believe in a fascist or nazi group within the Swedish police is responsible for the crime. In terms of background variables, age, gender and education have statistically significant effects on the dependent variable: males and older people are more likely to believe that the police have murdered Olof Palme, whereas those with higher education are more likely to disagree. These results also show the convergence of opinions of politically extreme groups, thus providing a further confirmation of our first hypothesis.
Economic Evaluations
We now move on to examine the relationship between extreme ideological positions and economic evaluations. Table 4 presents these results, and in line with expectations, Swedes who place themselves at the political extremes have more negative evaluations for the current state of the national economy compared to moderates. While the positive, statistically significant, linear effect of political ideology shows that right-wing individuals have generally more positive economic perceptions than leftists in 2013 (during the incumbency of a center-right coalition), the negative quadratic effect reveals that both extreme positions have lower evaluations of the present economic situation in Sweden. The effects of the background variables show that higher education and age are related to a more positive sociotropic economic evaluation. Males are also more likely than females to evaluate the economy positively.
Table 4
Determinants Explaining Economic Evaluations
| Variables | R | R2 | ΔR2 | ΔF | df | β |
|---|---|---|---|---|---|---|
| Evaluation of the state of the Swedish economy | ||||||
| Step 1 | .392 | .154 | .154*** | 158.78 | 3488 | |
| Political ideology | 0.35*** | |||||
| Age | 0.09*** | |||||
| Male | 0.07*** | |||||
| Education | 0.09*** | |||||
| Step 2 | .398 | .157 | .005*** | 19.25 | 3487 | |
| Political ideology | 0.36*** | |||||
| Age | 0.09*** | |||||
| Male | 0.06*** | |||||
| Education | 0.09*** | |||||
| Quadratic term | -0.07*** | |||||
| Retrospective Socio-tropic economic evaluation | ||||||
| Step 1 | .344 | .119 | .119*** | 113.50 | 3375 | |
| Political ideology | 0.33*** | |||||
| Age | -0.05*** | |||||
| Male | 0 | |||||
| Education | 0.08*** | |||||
| Step 2 | .346 | .121 | .002* | 5.80 | 3374 | |
| Political ideology | 0.33*** | |||||
| Age | -0.06*** | |||||
| Male | 0 | |||||
| Education | 0.08*** | |||||
| Quadratic term | -0.04* | |||||
| Prospective Socio-tropic economic evaluation | ||||||
| Step 1 | .335 | .112 | .112*** | 106.93 | 3380 | |
| Political ideology | 0.31*** | |||||
| Age | -0.05** | |||||
| Male | 0.02 | |||||
| Education | 0.10*** | |||||
| Step 2 | .338 | .114 | .002* | 7.10 | 3379 | |
| Political ideology | 0.32*** | |||||
| Age | -0.05** | |||||
| Male | 0.02 | |||||
| Education | 0.10*** | |||||
| Quadratic term | -0.04** | |||||
*p < .05. **p < .01. ***p < .001.
The results of the quadratic tests with the two other economic dependent variables are less clear, however. While there is a strong positive linear effect of political ideology on each of the dependent variables, which indicates rightists’ overall more positive economic evaluations compared to leftists, the strength of the quadratic effects is somewhat weaker, but still statistically significant. As for the background variables, higher education levels are associated with better economic perceptions in all cases, while age has a weak negative effect on prospective economic evaluations. We find a similar curvilinear pattern for all three economic variables. Thus, the results indicate that extreme ideological positions are related to negative economic evaluations. The results confirm the general thrust of Miller’s (1974) findings but add a new – economic – dimension to them.
Trust in Politicians and Interpersonal Trust
The last section deals with how ideological extremism predicts trust in politicians and interpersonal trust. The results in Table 5 do not provide support for our third hypothesis – that extreme political orientation leads to lower levels of trust in politicians and interpersonal trust. The inconsistency in effects for conspiracy belief and trust are somewhat surprising, given the evidence that distrust predicts conspiracy belief (see Abalakina-Paap et al., 1999). Possibly, these findings do not apply to the Swedish context, due to Swedish society’s overall high degree of trust (Ortiz-Ospina & Roser, 2016). The results indicate that right-wing ideology corresponds with higher levels of trust in politicians but with slightly lower levels of interpersonal trust. The quadratic terms, however, produce non-significant effects with regard to interpersonal trust and even a positive, albeit weak effect with regard to trust in politicians. With regard to education, the more educated Swedes are, the more likely they are to trust politicians, and to have a high degree of interpersonal trust. In terms of gender, males are less likely to trust politicians, whereas they tend to trust people more. Age is also a significant predictor of interpersonal trust, with older respondents more likely to trust fellow citizens.
Table 5
Determinants Explaining Trust in Politicians and Interpersonal Trust
| Variables | R | R2 | ΔR2 | ΔF | df | β |
|---|---|---|---|---|---|---|
| Trust in politicians | ||||||
| Step 1 | .189 | .036 | .036*** | 32.68 | 3522 | |
| Political ideology | 0.06*** | |||||
| Age | 0.03 | |||||
| Male | -0.07*** | |||||
| Education | 0.17*** | |||||
| Step 2 | .192 | .037 | .001* | 3.80 | 3521 | |
| Political ideology | 0.05*** | |||||
| Age | 0.03 | |||||
| Male | -0.06*** | |||||
| Education | 0.17*** | |||||
| Quadratic term | 0.03* | |||||
| Interpersonal trust | ||||||
| Step 1 | .225 | .05 | .05*** | 46.79 | 3522 | |
| Political ideology | -0.03* | |||||
| Age | 0.12*** | |||||
| Male | 0.06*** | |||||
| Education | 0.21*** | |||||
| Step 2 | .226 | .051 | .001 | 2.65 | 3521 | |
| Political ideology | -0.03 | |||||
| Age | 0.12*** | |||||
| Male | 0.06*** | |||||
| Education | 0.21*** | |||||
| Quadratic term | -0.03 | |||||
*p < .05. **p < .01. ***p < .001.
Discussion
The widespread similarities between the features and actions of political regimes, characterized with extreme ideology in the 20th century, have led to a debate in the social sciences, political science and political psychology. This debate suggests that while there are also many psychological differences between people at the political left vs. right, radical individuals at both sides of the spectrum are likely to share a similar pessimistic understanding of how the world and society functions.
In this study, we tested whether such similarities exist in the Swedish case, where the political system includes relevant political parties on both end of the political spectrum, as well as a substantial electorate that supports these parties. To do so, we proposed three hypotheses which served to test similar underlying psychological patterns between far left and far right individuals. By means of hierarchical regression analyses we tested whether there is a quadratic (curvilinear) relationship between extreme ideological positions and three sets of variables pertaining to conspiracy beliefs, economic evaluations, and political and interpersonal trust. The findings indicate that individuals who profess extreme ideology are indeed more likely to believe in conspiracies. As for economic evaluations, the findings also suggest that such similarities exist. Although, generally, citizens who oppose the incumbent government have more negative economic evaluations, both far left- and far right individuals tend to have a more negative economic evaluation than political moderates. However, the results show weak curvilinear relationship between trust in politicians and political ideology, as well as statistically insignificant relationship between interpersonal trust and ideological orientation. This implies that Swedes with radical or extreme ideology are not likely to have lower levels of political and interpersonal trust than others. Indeed, the finding that the extremes and moderates do not differ in interpersonal trust converges with a recent finding that they do not differ in terms of interpersonal paranoia (van Prooijen, Krouwel, & Pollet, 2015; Study 1), suggesting that the political extremes are less trustful only about the causes of societal and political events. Nevertheless, assessing variable to variable correlations, we find that conspiracy belief correlates negatively with positive economic perceptions, trust in politicians and interpersonal trust (see Appendix). This reveals that individuals who hold conspiracy beliefs are among the more pessimistic and distrustful fractions of society. In turn, political extremists are more likely to be conspiracy believers.
Interestingly, the results suggest that in terms of ideology, the linear terms are significant for all but one of the dependent variables. This shows that although there is curvilinear effect, it is not symmetrical, with the effects being steeper on one side of the curve than on the other. We find that individuals with far-left political orientation are more susceptible to conspiracy beliefs compared to their right-wing counterparts, are much more likely to be pessimistic about economic developments, and are less politically trustful. While the curvilinear effects are generally in line with previous research (e.g., van Prooijen, Krouwel, & Pollet, 2015), we find that in the Swedish case, individuals on the left are more likely to believe in conspiracies than those on the right. Future research efforts could explore whether this is still the case now, when a left-wing coalition is governing the Scandinavian nation. Prospective studies on the matter could also seek to replicate the findings by including variables measuring belief in additional conspiracy theories, and potentially taking into account other dependent variables such as authoritarianism and political cynicism, in order to further exemplify the similarities between citizens with extreme ideology at the two ends of the political divide.
While there are various conceptual differences between our main dependent variables, the overarching theoretical message that we offer is that both political extremes are more pessimistic about society than political moderates. One strength of the study presented here is that it yielded evidence of these effects of extremism on various conceptually distinct constructs that suggest such a negative worldview. Indeed, the moderate but significant correlations between our main dependent variables (see Table A.1 in Appendix) indicate related but distinct constructs, suggesting a broad indication of how the extremes perceive society. A limitation of our study, however, is that we did not have a direct measure of pessimism, and it therefore remains speculative whether societal pessimism is the core underlying factor that binds these findings.
The findings presented here are correlational, and future research may determine whether or not a pessimistic outlook on society causally increases political radicalism, or vice versa. Such a research question would require a longitudinal design that tracks participants’ ideology and perceptions of society over time. Given results of previous research, it stands to reason that societal pessimism leads to extremism. After all, research suggests that distressing societal circumstances increases the appeal of extremist political movements (Midlarsky, 2011), and that feelings of threat increase people’s support for radical leaders (Hogg et al., 2010). It is well-possible, however, that the relationship between radicalism and societal pessimism is bidirectional. Radicalization may alienate people by placing them at the fringes of society, which in turn may increase their negative perceptions of that society. These considerations suggest that it is worthwhile to examine both possible causal orders between the variables under investigation here.
Our findings allow for a certain confirmation that extreme ideology is related to belief in conspiracy theories. Political radicals also have somewhat more negative economic evaluations, although less clearly so, in prospective and retrospective terms, than for the general state of the Swedish economy. We find that, for all items we analyze, higher levels of education are related to a decrease of the likelihood to believe in conspiracy theories (cf. van Prooijen, 2017), better economic evaluations, and higher levels of political and interpersonal trust. Our analyses confirm the findings of previous research and supplement them with an economic dimension. Although the quadratic and linear effects of political ideology on some of the dependent variables are rather weak, given the considerably high levels of (political) trust in Sweden (Ortiz-Ospina & Roser, 2016), the existence of such effects is still telling. Therefore, we can conclude that both extremes converge in a mindset that is inclined to blame secret and evil conspiracies for the major problems that society faces, along with pessimistic expectations about contemporary economic developments. This is an important finding for those interested in political stability and system support within advanced democracies.