Content uploaded by Homero Gil de Zúñiga
Author content
All content in this area was uploaded by Homero Gil de Zúñiga on Jul 27, 2020
Content may be subject to copyright.
British Journal of Social Psychology (2020)
©2020 The British Psychological Society
www.wileyonlinelibrary.com
The influence of conspiracy beliefs on conventional
and unconventional forms of political participation:
The mediating role of political efficacy
Alberto Ard
evol-Abreu
1
*, Homero Gil de Z
u~
niga
2,3
and
Elena G
amez
1
1
Department of Cognitive, Social and Organizational Psychology, Universidad de La
Laguna, Spain
2
Democracy Research Unit (DRU), Political Science, University of Salamanca, Spain
3
Media Effects Research Lab (MERL), Film/Video & Media Studies Department,
Pennsylvania State University, USA
Recent approaches from social psychology lend support to conspiracy beliefs as a
motivated form of social cognition, structured around and consistent with a higher-order
belief system, which may have an impact on the way people understand their political
environment and respond to it. Building on these accounts, this study examines the
influence of conspiracism on political efficacy and, indirectly, on conventional and
unconventional forms of political participation. Drawing on two-wave panel data
collected in five democracies (United States, Japan, United Kingdom, Poland, and Estonia;
n=5,428), results suggest that individuals who hold conspiracy beliefs tend to regard the
political system as less responsive to citizens’ demands –external dimension of political
efficacy. We also found a less clear and country-specific effect of conspiracy beliefs on
perceptions of being less equipped to partake in the political process –internal efficacy.
Furthermore, conspiratorial beliefs negatively affect conventional modes of political
participation, indirectly through reduced external efficacy. We finally examine group
differences by country that suggest that both individual- and contextual-level factors may
explain the observed pattern of influences. Our results emphasize the potential of
currently widespread conspiratorial narratives to undermine attitudes and behaviours
that lie at the heart of the democratic process.
Conspiratorial accounts of social reality have widespread appeal for many citizens,
especially when it comes to explaining complex situations, or when official explanations
are perceived as insufficient or inconsistent (Clarke, 2002). Take as an example the 9/11
terrorist attacks, one of the recent events that has stimulated conspiratorial thinking. A US
national telephone survey conducted in 2006 found that more than 36% of respondents
considered it ‘very’ or ‘somewhat likely’ that the government in Washington either
cooperated with the terrorists or did nothing to prevent the attacks (Stempel, Hargrobe, &
Stempell III, 2007). Similarly, significant numbers of people across the globe believe that
*Correspondence should be addressed to Alberto Ard
evol-Abreu, Departamento de Psicolog
ıa Cognitiva, Social y Organizacional,
Universidad de La Laguna, C/ Prof. Jos
e Luis Moreno Becerra, s/n, 38200 La Laguna, Spain (email: aardevol@ull.es).
DOI:10.1111/bjso.12366
1
climate change is a fraud by scientists and governments with obscure interests (Jolley &
Douglas, 2014b). One-third (33%) of respondents recently surveyed in the United States
think that the seriousness of climate change is exaggerated (Jones & Saad, 2018), and 15%
go even further and deny that global warming is either occurring at all or that is has any
connection with human activities (Smith, 2019).
Conspiratorial narratives are not equally alluring to everyone. Earlier studies proposed
that certain individuals may have a conspiracist cognitive style, connected to a
‘monological belief system’ (Goertzel, 1994, p. 741) in which previously endorsed
conspiracy theories underpin the acceptance of novel ones (Swami et al., 2011). More
recent, competing perspectives suggest that individuals endorse conspiracies theories
that cohere with their higher-order belief system, which ultimately makes conspiracism a
form of motivated cognition (Douglas, Sutton, & Cichocka, 2017). Among the motivations
to adhere to a conspiracist mindset is the need to regain a feeling of control over one’s non-
immediate social environment (Swami et al., 2016). In this vein, conspiracy theories offer
‘seemingly coherent’ (Hoftstadter, 2008, p. 37) and oversimplified explanations of
important events that might help meet these needs.
Having learned about conspiracies, believers may tend to think they are in possession
of special knowledge, which may give them ‘a position of independence and authenticity
outside the domain of the conspiracy and its world of ignorance’ (Mason, 2002, p. 50).
Alternatively, the hodgepodge of conventional and conspiracist explanations in conspir-
acy believers’ cognitive structures may make them cultivate feelings of contradiction and
uncertainty about the world, which may in turn undermine self-perceptions of
abilities to understand social and political issues. In another respect, conspiratorial
depictions of the social and political world tend to present Manichean dynamics in which
evil outgroups, frequently from political, economic, or scientific elites, engage in secret
plots to advance their own obscure interests at the expense of ordinary, ingroup citizens
(Kramer & Schaffer, 2014). This may have implications for attitudes towards these high-
power outgroups (see Imhoff & Bruder, 2014) and, we posit, for the self-perception of
‘powerfulness (or powerlessness) in the political realm’ (Morrell, 2003, p. 589), what is
known as political efficacy.
Based on these considerations, this study explores the impact of conspiracy beliefs on
the personal-regarding (internal) and system-regarding (external) dimensions of political
efficacy. The study then builds on the well-established role of efficacy as an antecedent of
participation to test a mediation model in which conspiracy beliefs directly and indirectly
affect different forms of participation (conventional and unconventional) through
efficacy.
Conspiracy beliefs: Effects on political efficacy
Conspiracy theories construct their narratives around malevolent and secret plots of
powerful groups or individuals, usually belonging to social elites: politicians, scientists,
governments, multinational companies, etc. These high-power groups are frequently
pictured as being responsible for the negative implications of the conspiracy theory in
question (Imhoff & Bruder, 2014). For example, conspiracy theories have blamed the
British Government and the MI6 for the death of Princess Diana; denounced world
governments and military for conducting secret ‘chemtrail’ spraying operations for mind
control and weather modification; or accused scientists for a massive cover-up regarding
the negative effects of vaccines (see Butler, Koopman, & Zimbardo, 1995; Jolley &
Douglas, 2014a, 2014b; Rose, 2017).
2Alberto Ard
evol-Abreu et al.
On these bases, it is no surprise that previous research has found a negative impact of
conspiracy beliefs on attitudes towards elites and powerful groups (e.g., politicians,
capitalists, or lobbyists) (Imhoff & Bruder, 2014). Specifically, the study by Imhoff and
Bruder suggested that conspiracy beliefs have ‘predictive validity for attributions blaming
authorities for intentional misconduct as well as unintentional error’ (p. 39). In a related
vein, Jolley and Douglas (2014b) observed that exposure to conspiracy theories increased
individual perceptions of ‘political powerlessness’. This latter study conceptualized and
measured political powerlessness as the perception that ‘the world is run by the few
people in power, and there is not much the little person can do about it’ (2014b, p. 39).
Although of opposite valence, political powerlessness refers to one of the dimensions of
political efficacy.
The concept of political efficacy was originally developed in political science, as a
cognitive domain of perceived control that contributes to explaining differences in
political engagement (Niemi, Craig, & Mattei, 1991; Wo lak, 2018). Political efficacy relates
to individual perceptions of ‘their ability to influence’ their political environment, in the
sense that their actions can affect political outcomes (Gamson, 1968, p. 42). Political
efficacy is traditionally regarded as a bidimensional belief that includes distinct external
and internal components. External efficacy is exactly the converse of political power-
lessness (Craig, 1980; Watts, 1973) and talks about perceptions of government and system
responsiveness to citizens’ demands: ‘The lack of external efficacy [...] indicates the
belief that the public cannot influence political outcomes because government leaders
and institutions are unresponsive to their needs’ (Miller, Miller, & Schneide r, 1980, p. 253,
italics added). Thus, an increased sense of powerlessness towards the government is
equivalent to a reduced perception of external efficacy. Based on these explanations, we
offer our first hypothesis:
H1 . Conspiracism is negatively related to external efficacy.
Conspiracy theories are attempts to bring order in a social environment that is
overloaded with information. Important events are often complicated to explain because
their development is contingent on the interaction of multiple actors, their persuasive or
coercive power, bureaucratic processes, human errors, natural and environmental
conditionings, or mere chance (Goldberg, 2004). Against this backdrop, conspiracy
narratives provide ‘archaeology in narrative form, locating causes and origins of the
conspiracy, piecing together random occurrences to organize a chronology or sequence
of sorts’, which may give believers a false sense of mastery (Mason, 2002, pp. 43–44). In a
related vein, recent research findings suggest that conspiracy theories connect with social
psychological needs such as the need to feel unique to others. If conspiracism helps satisfy
this need, believers may feel that they are better informed than others because they are in
‘possession of unconventional and scarce information’ (Lantian et al., 2017, p. 161, italics
added).
The above arguments would point to a positive relationship between conspiracy
beliefs and individuals’ self-assessment of their competence to understand political issues
and to ‘participate effectively in politics’ (Niemi et al., 1991, p. 1408). This is what the
political science literature refers to as the internal (personal-regarding) dimension of
political efficacy (Morrell, 2003), a variable that has been repeatedly found to correlate
with all forms of political participation (Craig, 1980; Jung, Kim, & Gil de Z
u~
niga, 2011).
Internal efficacy comprises individual perception of being appropriately informed about
Conspiracism and political participation 3
social and political issues (Dyck & Lascher, 2009, p. 4), and one may therefore expect that
the perceived possession of privileged information is associated with higher levels of
internal efficacy.
Yet, there are also indications that conspiracy beliefs constitute a breeding ground for
the adherence to multiple, partial, unrelated,and even contradictory explanations of social
life (Swami & Coles, 2010; Wood, Douglas, & Sutton, 2012). Some people might turn to
conspiracy theories in pursuit of simpler explanations to make sense of the world, or
because they doubt official explanations of events (van Prooijen, 2017), but they might end
up getting lost in a sea of contradictions and uncertainties that might reduce their self-
perceptions of political competence. Taking the death of the Princess Diana as an example,
conspiracy-prone personalities have been offered a menu of competing theories including
her assassination by MI6, a revenge from some of Mohammed al-Fayed enemies,a settling of
accounts by arms traffickers for her involvement in international campaigns against anti-
personnel mines, or the simulation of her own death, among others (Wood et al., 2012).
Conspiracy believers may not be able to stomachthis ‘indissoluble’mixture of explanations,
which may feed the perception that they cannot understand political issues.
A complementary argument is that believers may ultimately not be sure that the
alternative realities depicted in conspiracy theories are actually true. Many of them may be
aware that their ‘rare’ information contradicts institutionally validated explanations and
may therefore harbour doubts about their personal skills to understand and participate in
politics. Previous empirical findings offer some support for this interpretation. In a natural
experiment on the effects of watching the film JFK by Oliver Stone, the authors found that
post-view participants were more confident about the correctness of their conspiracy
beliefs (Butler et al., 1995). Thus, compared to the pre-view group, participants in the
post-view condition showed higher levels of confidence in their beliefs that the Dallas
police, CIA/Pentagon officials, and Vice President Lyndon B. Johnson were all involved in
the assassination of Kennedy. However, the post-view group also believed –at least to
some extent –in the government’s official version that Lee Harvey Oswald acted alone to
kill Kennedy. These findings suggest that an amalgamation of official and conspiratorial
accounts may be shaping conspiracy believers’ worldviews, which would add complexity
to the already challenging task of understanding contemporary reality. These conflicting
explanations make it difficult to predict the sign of the influence of conspiracy beliefs on
internal efficacy. Also, we cannot rule out the possibility that positive and negative
influences of conspiracism on internal efficacy cancel each other out. We therefore ask
our first research question:
RQ1: What is the direct effect of conspiracism on internal efficacy?
Conspiracism and participation: Conventional and unconventional activities
Research exploring the individual and collective impact of conspiratorial thinking
suggests consequences beyond attitudes and beliefs. In the area of public health, general
conspiracism appears to be connected to important risk behaviours such as avoiding
conventional medicine and engaging in pseudoscientific medical practices (Oliver &
Wood, 2014). Similarly, anti-vaccine conspiracy beliefs seem to have a detrimental effect
on future vaccination intentions (Jolley & Douglas, 2014a). Among certain sociodemo-
graphic groups of African Americans, black genocide beliefs –that is the idea that elites are
deploying strategies to reduce the black population –could have increased their unsafe
sex practices (Bogart & Thorburn, 2006).
4Alberto Ard
evol-Abreu et al.
In the political realm, several studies have examined the influence of exposure to
specific conspiracy narratives on future intentions to participate in politics. Since many
conspiracy theories describe the political system and its institutions as part of a wider
network of conspirators engaged in malevolent activities, it seems reasonable to expect a
direct negative association between conspiracism and (at least conventional) participa-
tion. Some empirical support has been found for this idea. The above-mentioned
experiment involving viewers of JFK showed that exposure to conspiracism (i.e.,
watching JFK) had a deleterious effect on future intentions to participate in conventional
participation activities (e.g., voting, donating money, and campaigning) (Butler et al.,
1995). While the study did not test for indirect effects, the authors suggested that this
negative response could have been channelled through certain political attitudes (e.g.,
reduced external efficacy) or emotions (anger and hopefulness). Using a different
methodological approach (online experiments involving university students), the study
by Jolley and Douglas (2014b) assessed the effects of exposure to pro- or anti-conspiracy
written information on participants’ post-reading intentions to engage in institutionalized
forms of participation (voting, donating money to a party or organization). The authors
found no direct effect, but they did observe a mediated negative relationship between
exposure to specific conspiracy theories and conventional participation through political
powerlessness. Finally, a nationally representative online survey conducted among
twelve hundred Americans suggests that conspiracy theorists are particularly disposed to
believe that elections are affected by fraud, which may partially explain their lower levels
of conventional political engagement –from voting to donating money, putting up
political signs, attending local meetings, and running for office –(Uscinski & Parent,
2014).
These theoretical arguments and empirical findings lend us theoretical support to
argue that conspiracism may only have negative effects over conventional forms of
participation. According to Inglehart (1977), conventional participation comprises the
institutional and bureaucratized mobilization of public support aimed at creating and
maintaining legitimacy in democracies. These hierarchical institutions are used to lead ‘a
mass of disciplined troops’, whose only task is to choose ‘between two or more sets of
decision-makers’ (Inglehart, 1977, pp. 3, 299). At this stage, it is however not clear
whether the negative association between conspiracism and conventional participation
can be fully explained by a reduced sense of political efficacy. If this were the case, then
we may expect no direct association between conspiracism and conventional participa-
tion. But there may be other, still unexplored, direct and indirect mechanisms that lead
conspiracists to reduce their levels of conventional political engagement. We therefore
ask our second research question:
RQ2: What is the direct effect of conspiracism on conventional political participation?
In addition to the above-described institutionalized forms of political action, citizens in
modern ‘post-materialist’ societies frequently engage in unconventional participation
behaviours outside the institutional framework (Inglehart & Catterberg, 2002). These
include a wide range of protest-oriented activities aimed at showing discontent and
achieving social change (Ekman & Amn
a, 2012; Norris, 2002). Thus, unconventional
participation can come in the form of legal (e. g., creating a petition in an online platform
such as change.org) or illegal (squatting buildings or damaging private property);
individual (boycotting products or companies for political reasons), or collective
(participating in demonstrations or marches) actions (Ekman & Amn
a, 2012).
Conspiracism and political participation 5
The above theoretical distinctions are relevant in the context of this study because
conspiracy believers may view unconventional activities as more appropriate to redress
their perceived grievances. Accordingly, conspiracy believers may have subjective
reasons to protest because nearly all conspiracy theories include narratives about
injustices and individual or group deprivation. In addition, conspiracy theories tend to
interpret social and political events in ‘dualistic, categorical’ terms (Sapountzis & Condor,
2013), which may help conspiracists develop a certain feeling of collective identity, one
based on their common perception of being exploited, victimized, or harmed by powerful
outgroup elites (see Kramer & Schaffer, 2014).
These feelings of identification with disadvantaged groups suffering from the tyranny
of elites may stimulate collective protest on behalf of those groups (van Zomeren,
Postmes, & Spears, 2008). Uscinski and Parent’s survey found that conspiracy believers –
particularly those in the top 4% of conspiratorial predispositions –show higher levels of
agreement with the statement: ‘Violence is sometimes an acceptable way to express
disagreement with the government’. Although the present article does not address violent
participation, we expect a direct positive effect of conspiracy beliefs on unconventional
participation. However, as for the case of conventional participation, we cannot predict if
this positive influence can be entirely explained by reduced levels of political efficacy –
which would eliminate any direct effect of conspiracism on unconventional participation
–or there are also other mechanisms of influence not considered here. Accordingly, we
pose the following research question:
RQ3: What is the direct effect of conspiracism on unconventional (non-violent) political
participation?
An efficacy-mediated model of political participation
The associations between political efficacy and participation have been well established
in the literature. Broadly speaking, efficacy can be said to be a direct and positive
antecedent of political participation: those who feel that their actions do ‘have, or can
have, an impact on the political process’ (external dimension) and that ‘the individual
citizen can play a part in bringing about this change’ (internal dimension) are more likely
to participate (Campbell, Gurin, & Miller, 1954, p. 187). External efficacy has shown to be
a particularly decisive factor in explaining changing levels of participation in recent
history. In a study assessing reasons for the downward voting trend in U.S. presidential
elections, Shaffer (1981) found that the system-regarding dimension of political efficacy
accounted by itself ‘for 67 per cent of the turnout decline [from 1960 through 1976]’ (p.
78) (see Abramson & Aldrich, 1982 for a discussion of Shaffer’s findings).
Yet things become less straightforward when we jointly consider the relationships
between both dimensions of efficacy and different forms of participation (e.g.,
conventional and unconventional). Thus, those who perceive themselves as competent
and able to influence politics (high internal effica cy) are generally more likely to take every
opportunity to participate, whether it is voting, contacting an official, or joining a
demonstration (Zhang & Lin, 2018). Differently, the perception that governments and
institutions are responsive to citizens’ demands expressed through conventional
channels may foster institutionalized participation, but disincentivize protest-based
actions (Lee, 2010). Put differently, those who perceive that conventional channels of
influence are not promising because their government does not listen to ‘people like
them’ (low external efficacy) may seek for unconventional alternatives to change the
current state of affairs (Pollock, 1983; Reicher, 2004; van Stekelenburg & Klandermans,
6Alberto Ard
evol-Abreu et al.
2013). We therefore expect conspiracy beliefs to follow multiple indirect routes to
influencing conventional and unconventional forms of participation. First, we hypoth-
esize that conspiracy beliefs will have opposite effects on both types of participation
through reduced levels of external efficacy:
H2 . Conspiracism will negatively affect conventional participation through external
efficacy.
H3 . Conspiracism will positively affect unconventional political participation
through external efficacy.
Finally, we also consider possible indirect effects of conspiracism on conven-
tional and unconventional participation through the personal-regarding (internal)
dimension of efficacy. Since we could not predict the sign of the relationship between
conspiracism and internal efficacy, we ask a final set of research questions about the
possible mediating role of internal efficacy in the relationships between conspiracism and
participation:
RQ4: What is the indirect effect of conspiracism on conventional participation through
internal efficacy?
RQ5: What is the indirect effect of conspiracism on unconventional participation through
internal efficacy?
Methods
Data
The data used in this study come from an original online survey conducted in two waves,
the first (W
1
) in September 2015 and the second (W
2
) in March of the next year. As part of
the larger ‘Digital Influence World Project,’ researchers at the time based at Massey
University (New Zealand) and the University of Vienna (Austria) collaboratively collected
survey data from countries in Europe, Asia, the Americas, Oceania, and one city in South
Africa (see Gil de Z
u~
niga & Liu, 2017 for details on the country sample and methods). The
research teams jointly designed the questionnaire and collected data about media uses,
personality traits, patterns of social trust, and political behaviours, among others. For this
study, we used concurrently collected data from five well-established democracies in
three continents: United States, Japan, United Kingdom, Poland, and Estonia (see Gil de
Z
u~
niga & Liu, 2017 for details).
This country sample represents different societies and geographical regions of the
world, which increases the generalizability and external validity of the findings. At the
same time, all five countries show certain uniformity in terms of economic performance,
human development, and political rights. For example, according to Human Develop-
ment Index, all countries in our sample enjoy ‘very high’ human development (UNDP,
2015). Similarly, the countries in our study have open media environments and acceptable
levels of freedom of information and expression. These macro-level attributes allow us to
pool the five surveys and create a large sample that is culturally diverse and relatively
homogeneous in terms of democratic quality.
The research teams translated the questionnaires into the principal language of each
country and built equivalent web-based questionnaires in Qualtrics. We then hired the
Conspiracism and political participation 7
services of the media research company Nielsen, which was tasked with the adminis-
tration of the survey. Nielsen was asked to get national adult samples whose demographic
composition closely mirrored that of the selected countries (in term of key variables such
as age, sex, income, ethnicity, or education level; see Gil de Z
u~
niga & Liu, 2017). Nielsen
and their international partners used quota sampling techniques to select specific
respondents among those previously registered on their panels, and not necessarily
interested in our study.
From W
1
, we received valid responses from a total of 5,428 individuals (United States,
n
W1
=1,161; Japan, n
W1
=975; United Kingdom, n
W1
=1,064; Poland, n
W1
=1,060;
Estonia n
W1
=1,168). Six months later, Nielsen re-contacted the original respon-
dents and obtained usable information from 3,146 panellists (averaged retention rate
58%; United States, 45%, n
W2
=528; Japan, 59%, n
W2
=577; United Kingdom, 64%,
n
W2
=679; Poland, 59%, n
W2
=629; Estonia, 63%, n
W2
=733). Seventy-three of these
cases could not be matched with their W
1
responses, and were therefore excluded
from further analyses. Compared to W
1
,ourW
2
sample is slightly older (M
W1
=47.43;
M
W2
=50.16 years, respondents were legal adults); contains more whites (W
1
, 94.6%;
W
2
, 96.4%) and less women (W
1
,52%;W
2
, 49.3% female); and is somewhat more
educated (‘1’ =elementary school to ‘5’ =postgraduate degree;M
W1
=3.42;
M
W2
=3.45).
Variables of interest
Because the survey was translated to four different languages and distributed in five
different countries, reliability values of most of the constructs are acceptable, but not
excellent.
Unconventional (but non-violent) participation
This variable taps the extent to which respondents engage in non-violent political protest
and actions outside the institutional structures of the system, aimed at achieving social
change (Norris, 2002). Building in previous approaches (Inglehart & Catterberg, 2002;
Ekman & Amn
a, 2012), we asked respondents about their frequency of participation
(1 =never to 7 =all the time) in the following activities during the previous 3 months:
‘attending a political rally, participating in any demonstrations, protests, or marches’,
‘boycotting a certain product or service because of the social or political values of the
company’, or ‘creating an online petition’ (averaged scale, W
1
Cronbach’s a=.72;
M=1.69; SD =1.05; W
2
Cronbach’s a=.71; M=1.61; SD =0.98).
Conventional participation
We asked respondents about the frequency with which they take part in electoral and
non-electoral forms of formal political participation (see Ekman & Amn
a, 2012; Gil de
Z
u~
niga, Ard
evol-Abreu, & Casero-Ripoll
es, 2019; Verba & Nie, 1972/1987). Specifically,
we asked respondents to report how frequently (1 =never to 7 =all the time) they voted
in ‘national or presidential elections,’ and ‘local or statewide elections’. Respondents
were also asked about their frequency of engagement, during the 3 months prior to the
survey, in activities such as ‘contacting an elected public official’ and ‘participating in an
online question and answer session with a politician or public official’ (averaged scale, W
1
Cronbach’s a=.62; M=3.53; SD =1.18; W
2
Cronbach’s a=.62; M=3.59; SD =1.13).
8Alberto Ard
evol-Abreu et al.
External political efficacy
We asked respondents to rate their degree of agreement (1 =disagree completely to
7=agree completely) with the following statements: ‘No matter whom I vote for, it won’t
make a difference’ (coded reversely) and ‘People like me don’t have any say in what the
government does’ (coded reversely). The resulting two item scales showed good
reliability: W
1
Spearman–Brown coefficient =.69; M=4.07; SD =1.58; W
2
Spearman–
Brown coefficient =.74; M=4.10; SD =1.56.
Internal political efficacy
We asked respondents about the extent of their agreement (1 =disagree completely to
7=agree completely) with the following assertions: ‘I consider myself well qualified to
participate in politics’ and ‘People like me can influence the government’ (W
1
Spearman–
Brown coefficient =.68; M=3.34; SD =1.42; W
2
Spearman–Brown coefficient =.67;
M=3.42; SD =1.42).
Generalized conspiracism
Based on previous operationalizations of the construct (Gil de Z
u~
niga & Liu, 2017; Rose,
2017), we asked respondents to indicate their level of agreement with the following four
statements: ‘Seeing a conspiracy behind many events is the result of an overly active
imagination’ (coded reversely), ‘Many significant world events have occurred as a result of
a conspiracy’, ‘Despite what the authorities say, large business and/or government
routinely engage in sinister, secret activities in the name of profit’, and ‘When one looks at
the bigger picture, it is easy to see that many seemingly unrelated events form part of a
larger plan, orchestrated by powerful others acting in secrecy’ (W
1
Cronbach’s a=.74;
M=4.12; SD =1.13; W
2
Cronbach’s a=.77; M=4.08; SD =1.12.
Control variables
In order to minimize confounding effects of the predictor variables, all our regression-
based models include a block of demographic characteristics as controls: age, gender,
education, income perception (‘1’ =people who are the least well off in society to ‘10’ =
people who are the most well off;M=5.17; SD =1.86), and ethnicity (94.2% whites, or
Nipponjin in Japan). A second block of variables controls for the effects of respondents’
sociopolitical antecedents that have been associated to our mediating or dependent
variables in previous studies (see, e.g., Eveland & Hively, 2009; Jung, et al., 2011; Shah
et al., 2007): size of the discussion network (transformed with the natural logarithm prior
to analysis), political discussion frequency, strength of ideological affiliation, political
knowledge, and political interest.
Statistical analyses
With the aim to model the structural relationships among conspiracism, efficacy, and
participation, we used path analysis. All observed variables were residualized for
demographics and sociopolitical controls prior to model fitting (see Keum, Devanathan,
Deshpande, Nelson, & Shah, 2004). Our models are autoregressive and therefore consider
the temporal stability of a construct by controlling for previous levels of the dependent
variable (Y
1
). As we only collected two waves of data, we estimated
Conspiracism and political participation 9
conspiracism ?efficacy paths (apaths in Figure 1) using W
1
measures of conspiracism
and W
2
measures of efficacy, while controlling for W
1
levels of efficacy. Following Cole
and Maxwell (2003), we assumed stationary and calculated efficacy ?participation (b
paths) using W
1
measures of efficacy and W
2
measures of participation, while controlling
for W
1
levels of participation. Under this assumption, we computed the products axbto
determine the effects of conspiracism on participation through efficacy (Cole & Maxwell,
2003, p. 562). The residualization of the variables of interest was performed using SPSS
software, version 25. Path analysis was conducted with Mplus, version 8.4.
Results
Based on our theoretical assumptions, we first developed an autoregressive path model
that tested for direct influences of conspiracism (W
1
) on conventional and unconven-
tional participation (W
2
), as well as indirect effects through internal and external efficacy
–assuming stationary, as explained above. This model included the previous
Unconventional
Participation (W1)
Unconventional
Participation (W2)
Conventional
Participation (W2)
Conspiracism
(W1)
External
Efficacy (W1)
Conspiracism
(W2)
Internal
Efficacy (W2)
External
Efficacy (W2)
Conventional
Participation (W1)
Internal
Efficacy (W1)
Figure 1. Autoregressive path model of conspiracy beliefs (W
1
) on internal and external efficacy (W
2
),
and conventional and unconventional participation (W
2
). Note. Maximum-likelihood estimation. Con-
tinuous path entries are standardized coefficients (StdYX standardization). **p<.01; ***p<.001.
Dashed paths indicate non-significant relationships. Stability paths are represented in grey. The key paths
for indirect effects are shown in black. Other paths are omitted for clarity (see methods section). The
effects of demographic and sociopolitical controls were residualized in all observed variables prior to
model fitting. W
1
variables were brought into the model by mentioning their variance in the MODEL
command. The model also includes indirect effects of conspiracism on conventional and unconventional
participation through political efficacy (represented in Table 3). Explained variance of criterion variables:
Conspiracism (W
2
), R
2
= .427; Internal Efficacy (W
2
), R
2
= .277; External Efficacy (W
2
), R
2
= .275;
Conventional Participation (W
2
), R
2
= .472; Unconventional Participation (W
2
), R
2
= .243.
10 Alberto Ard
evol-Abreu et al.
measurement in time one (W
1
) for each observed variable in W
2
(stability paths). We also
modelled concurrent residual covariances among all W
2
variables. The fit of this initial
model was not good enough (Table 1).
We then followed a partially exploratory strategy, based on theoretical considerations
and the examination of model modifications that would improve the model fit (see Roth &
MacKinnon, 2012). We first deleted non-significant time-lagged paths between conspir-
acy and both types of participation (non-mediated effects that were not part of our
theoretical formulation). To capture the lagged associations between the two dimensions
of political efficacy, we regressed W
2
internal efficacy on W
1
external efficacy, and vice
versa. We also modelled time-delayed effects of W
1
participation (both modes) on W
2
efficacy (both dimensions; see Finkel’s ‘reciprocal effects of participation and political
efficacy,’ 1985). The model is outlined in Figure 1. This revised structure provides a better
fit to our data, both in the overall sample and in each of the country subsamples (Table 1).
H1 predicted a negative effect of conspiracy beliefs on the external dimension of
political efficacy. An examination of lagged zero-order correlations reveals that W
1
conspiracism and W
2
external efficacy are negatively correlated: r=.186, p<.001
(Table 2). This effect remains significant and strong in the autoregressive path model
(Figure 1, path a2; b=.068, p<.001), which confirms H1. To test if the estimate of this
path coefficient was equal across countries, we conducted multi-group analyses and
calculated the chi-square difference test between a partially constrained model –where
the path a2 is restricted invariant across countries –and a fully unconstrained model –
where all path coefficients are allowed to vary across groups. The non-significance of this
difference test, v
2
=4.00, df [4], p= .406, suggests that the negative effect of W
1
conspiracism on W
2
external efficacy remains equivalent across the five country
subsamples.
RQ1 asked about the possible influence of conspiracy beliefs on internal efficacy. The
time-lagged (W
1
↔W
2
) correlation between conspiracism and internal efficacy is
significant negative in sign: r=.098, p<.001 (Table 2). However, once one accounts
for W
1
levels of efficacy and includes demographic and sociopolitical controls, path
analysis does not provide empirical support for an effect of conspiracism on internal
efficacy (path a1 in Figure 1; b=.021, p=.226). Multi-group analyses indicate that the
a1 path coefficient changes across countries, difference test v
2
=9.84, df [4], p= .004.
At the level of each individual country, this path coefficient shows statistical significance
only among the British subsample (b=.100, p=.006). This suggests a not generalized,
but country-specific negative relationship between conspiracism and internal efficacy.
In order to answer RQ2 and RQ3, the model also tested possible direct influences of
conspiracism on both forms of participation. The uncontrolled and time-lagged
Table 1. Fit statistics for path models of direct and indirect effects
Model n v
2
df p-value RMSEA CFI TLI SRMR
Initial 4,379 183.05 12 <.001 .057 .968 .908 .038
Revised (Figure 1) 4,379 15.30 8 .053 .014 .999 .994 .009
United States 962 9.73 8 .284 .015 .999 .993 .020
Japan 784 8.12 8 .422 .004 1.000 .999 .016
United Kingdom 849 7.41 8 .493 <.001 1.000 1.002 .013
Poland 858 18.14 8 .020 .038 .988 .949 .025
Estonia 926 14.55 8 .069 .030 .996 .981 .017
Conspiracism and political participation 11
Table 2. Zero-order correlations among all independent and dependent variables in the study
Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1. Age –
2. Gender
(1 = female)
.11
c
–
3. Education .00 .04
b
–
4. Income .01 .04
b
.19
c
–
5. Race (1 = White) .13
c
.02 .06
c
.03
a
–
6. Discuss.
Network Size W1
.05
c
.08
c
.19
c
.11
c
.01 –
7. Discussion
Frequency W1
.09
c
.07
c
.18
c
.20
c
.02 .65
c
–
8. Strength of
Ideology W1
.13
c
.09
c
.12
c
.07
c
.02 .16
c
.16
c
–
9. Political
Knowledge W1
.18
c
.33
c
.17
c
.07
c
.09
c
.18
c
.09
c
.12
c
–
10. Political Interest
W1
.19
c
.24
c
.15
c
.14
c
.03
a
.37
c
.39
c
.29
c
.40
c
–
11. Internal Efficacy
W1
.04
b
.22
c
.18
c
.23
c
.05
c
.33
c
.36
c
.19
c
.22
c
.49
c
–
12. External Efficacy
W1
.01 .05
c
.09
c
.15
c
.03
a
.10
c
.07
c
.08
c
.10
c
.16
c
.30
c
–
13. Internal Efficacy
W2
.04
a
.18
c
.20
c
.23
c
.05
b
.29
c
.32
c
.22
c
.19
c
.46
c
.64
c
.34
c
–
14. External Efficacy
W2
.01 .02 .12
c
.15
c
.00 .15
c
.13
c
.13
c
.10
c
.21
c
.34
c
.53
c
.36
c
–
15. Conventional
Part. W1
.22
c
.14
c
.20
c
.17
c
.06
c
.34
c
.40
c
.21
c
.24
c
.44
c
.38
c
.13
c
.35
c
.22
c
–
16. Unconventional
Part. W1
.16
c
.09
c
.15
c
.11
c
.06
c
.36
c
.51
c
.11
c
.01 .18
c
.30
c
.01 .25
c
.08
c
.38
c
–
Continued
12 Alberto Ard
evol-Abreu et al.
Table 2. (Continued)
Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
17. Conventional
Part. W2
.22
c
.14
c
.20
c
.16
c
.06
b
.32
c
.35
c
.20
c
.20
c
.40
c
.31
c
.17
c
.37
c
.19
c
.75
c
.25
c
–
18. Unconventional
Part. W2
.13
c
.09
c
.14
c
.10
c
.04
a
.30
c
.40
c
.12
c
.03 .16
c
.24
c
.08
c
.32
c
.05
b
.23
c
.58
c
.37
c
–
19. Conspiracism
W1
.01 .02 .15
c
.14
c
.02 .03
a
.09
c
.03
a
.09
c
.02 .07
c
.18
c
.10
c
.19
c
.03
a
.08
c
.05
a
.05
b
–
20. Conspiracism
W2
.03 .06
b
.17
c
.16
c
.04
a
.01 .06
b
.02 .11
c
.04
a
.10
c
.17
c
.11
c
.20
c
.05
a
.06
c
.03 .04
a
.66
c
–
Note. Cell entries are two-tailed pairwise correlation coefficients. Superscript a =p<.05, Superscript b =p<01, Superscript c =p<.001.
Conspiracism and political participation 13
correlation between conspiracism and participation is significant for both types of
participation, but opposite in sign: W
1
conspiracism - W
2
conventional participation,
r=.047, p=.011; W
1
conspiracism - W
2
unconventional participation, r=.049,
p=.008 (Table 2). However, as explained above, the non-mediated paths W
1
conspir-
acism ?W
2
conventional participation, and W
1
conspiracism ?W
2
unconventional
participation were not significant in the initial model. We deleted these direct effects
paths in the revised model because they were not part of our theoretical predictions. If
these paths are re-included in the revised structure, model fit decreases (v
2
=14.63,
df =6, p=.023, RMSEA =.018, CFI =.998, TLI =.991, SRMR =.009) and both direct
paths are still non-significant (W
1
conspiracism ?W
2
conventional participation:
b=.002, p=.919; W
1
conspiracism ?W
2
unconventional participation: b=.014,
p=.427). Therefore, the data do not support direct effects of W
1
conspiracism on W
2
conventional (RQ2) or unconventional (RQ3) forms of participation.
Hypotheses 2 and 3, and research questions 4 and 5 evaluated indirect effects of
conspiracism on political participation. These indirect influences are commonly
calculated as the product of path atimes path b. Focusing on the autoregressive paths
(aand bin Figure 1), only a2 (conspiracism ?external efficacy) and b3 (external
efficacy ?conventional participation) coefficients are significant, which makes indirect
influences of conspiracism more likely to flow via reduced external efficacy and to
negatively affect conventional forms of participation. Multi-group analyses show that,
among bpaths, only b2 remains unchanged across countries, difference test v
2
=9.18, df
[4], p= .056. We also found that a constrained model with both a2 and b2 paths
constrained to be equal did not result in a decrease in model fit, difference test
v
2
=13.177, df [8], p=.105. All the remaining bpaths differ across countries: b1,
difference test v
2
=17.24, df (4), p= .001; b3, difference test v
2
=12.28, df (4),
p= .015; b4, difference test v
2
=15.41, df (4), p= .003.
H2 stated a negative influence of W
1
conspiracism on W
2
conventional participation
via decreased levels of external efficacy. We found empirical support for H2 in the
autoregressive mediation analysis (Table 3): unstandardized point estimate =.003
[.001], p=.015. Differently, our third hypothesis predicted an indirect route of influence
from conspiracism to unconventional participation through external efficacy, positive in
sign. This mediated effect was not observed in the longitudinal mediation test (point
estimate =.002 [.001], p=.122).
Finally, RQ4 and RQ5 queried whether internal efficacy has a mediating role in linking
W
1
conspiracy beliefs with W
2
conventional (RQ4) or unconventional (RQ5) forms of
participation. After controlling for initial (W
1
) levels of efficacy and participation, we did
not find empirical support for the indirect link conspiracy beliefs ?internal effi-
cacy ?conventional participation (RQ4, point estimate =.000 [.000], p=.401).
Similarly, our data did not provide support for an effect of W
1
conspiracy beliefs on W
2
unconventional participation mediated by internal efficacy (point estimate =.001 [.000],
p=.368) (Table 3).
Longitudinal indirect effects are not equivalent across countries, which does not come
as a surprise considering the group variance of most aand bpaths. Multi-group analyses
show that most longitudinal mediation effects fall outside the range of significance once
we split the sample into five country subsamples. There are however five indirect
influence pathways in the United Kingdom, Estonia, and Japan that reach or approach
statistical significance and are worth mentioning. In the British subsample, conspiracism
has a negative effect on unconventional participation via reduced internal efficacy (RQ5,
point estimate =.009 [.004], p=.039). Also in the United Kingdom, conspiracism
14 Alberto Ard
evol-Abreu et al.
shows borderline significant indirect effects on: a) conventional participation through
reduced internal efficacy (RQ4, point estimate =.007 [.004], p=.076) and b)
unconventional participation through reduced external efficacy (H3, point esti-
mate =.007 [.004], p=.062). In Estonia and Japan, the path conspiracy beliefs ?ex-
ternal efficacy ?conventional participation reaches or approaches conventional
significance levels (H2; Estonia, point estimate =006 [.003], p=.040; Japan, point
estimate =.011 [.006], p=.080).
Discussion
This study used a direct and indirect efficacy-mediated model to examine the effects of
conspiracism on political participation. Overall, our results place conspiracy beliefs in a
negative light regarding their implications for democratic citizenship. The most relevant
and consistent result of our analyses is that conspiracy beliefs negatively influence the
external, system-regarding dimension of political efficacy. This is in line with Jolley and
Douglas (2014b) findings regarding feelings of ‘political powerlessness’ and means that
those who endorse conspiracy beliefs tend to develop more negative perceptions about
governmental and institutional responsiveness to citizens’ demands. This detrimental
effect follows logically from some of the common storylines of conspiracy theories:
powerful networks of elites (including governments and political leaders) that secretly
collude to pursue their agenda at the expense of ordinary people. This effect was robust,
statistically significant, and equivalent across countries, and remained so even after
controlling for sociodemographic characteristics and W
1
levels of external efficacy. Being
external efficacy a well-documented antecedent of political participation, we believe this
negative influence to be highly relevant in the context of representative democracies.
Second, our results show a less clear and country-specific (only in the United
Kingdom) influence of conspiracy beliefs on the persona l-regarding dimension of political
efficacy. In the pooled sample, the time-lagged zero-order correlation between these two
variables is negative and strong, but this effect disappears in the path model with
sociodemographic, personality, and autoregressive controls. One of the reasons for this
weak or absent effect may be that the relationship between conspiracism and internal
efficacy varies among individuals or groups, and the effects cancel each other in the overall
Table 3. Indirect effects of conspiracism on conventional and unconventional participation
Indirect effects Estimate (SE)
Two-tailed
p-value
Conspiracism (W
1
)?Internal Efficacy (W
1, 2
)?
Conventional Participation (W
2
)
.000 (.000) .401
Conspiracism (W
1
)?Internal Efficacy (W
1, 2
)?
Unconventional Participation (W
2
)
.000 (.000) .368
Conspiracism (W
1
)?External Efficacy (W
1, 2
)?
Conventional Participation (W
2
)
.003 (.001) .015
Conspiracism (W
1
)?External Efficacy (W
1, 2
)?
Unconventional Participation (W
2
)
.002 (.001) .122
Note. Autoregressive (W
1
X?W
1, 2
M?W
1
Y) indirect effects calculated as the product of a9b
paths in Figure 1, assuming stationarity of the causal structure. All coefficients are unstandardized.
W
1
=Wave 1, W
2
=Wave 2.
Conspiracism and political participation 15
sample. As proposed in the theory section, there could be individuals or subpopulations in
which conspiracy beliefs increase internal efficacy because they perceive they have
access to ‘special’ and ‘privileged’ forms of knowledge that makes them better informed
than others. Conversely, other conspiracy-minded individuals or subpopulations may
predominantly experience a reduction in their levels of internal efficacy, because the
aggregation of these unconventional ideas cultivates feelings of contradiction and
uncertainty about the political world. Future research should study country-level
characteristics (media and political systems, types of conspiracy theories that generate
interest, etc.) that may interact with individual variables and explain the observed country
differences. Also, future designs should analyse whether different time frames between
waves of data collection (i.e., more or <6 months) lead to stronger or weaker relationships
between conspiracism, efficacy, and participation.
Also of interest are the links between conspiracy beliefs and participation. Concerning
direct influences, our findings do not support an immediate effect of conspiracism on
conventional or unconventional participation. These associations seem to be mediated by
conspiracists’ reduced levels of political efficacy. On the one hand, we found that
conspiracy beliefs are indirectly and negatively related to conventional participation
through external efficacy, even after accounting for previous levels of efficacy and
participation. This makes good theoretical sense if we consider the above-described
deleterious effect of conspiracism on the external dimension of political efficacy and the
well-established connection between external efficacy and political participation (see
Abramson & Aldrich, 1982; Shaffer, 1981).
On the other hand, we expected that diminished external efficacy arising from
endorsement of conspiratorial narratives would act as a catalyst for protest against
political elites. Thus, we hypothesized that in the face of perceived political injustice –
governments and elites involved in conspiracies –against which conventional channels of
influence are considered to be inefficient –low external efficacy –believers might channel
their grievances into different forms of protest: demonstrations marches, boycotts,
petitions, etc. In our sample, however, external efficacy (b4 path) does not show any
relationship with unconventional participation. This is different to other studies with a
cross-sectional design (e.g., Lee, 2010) and may explain the lack of mediated effect of
conspiracy beliefs on protest participation via the system-regarding dimension of efficacy.
The study also explored indirect influences of conspiracism on participation through
internal efficacy. Given the general lack of relationship betwee n internal efficacy (W
1
) and
any type of participation (W
2
), it was difficult to find any indirect effect channelled by the
personal-regarding dimension of efficacy. We only found one small and country-specific
(in the British subsample) negative effect conspiracism ?internal efficacy ?uncon-
ventional participation. Future studies should attempt to replicate these findings using
more than two waves of data in order to avoid having to assume stationarity of the indirect
effects, and –similar as above –change the time intervals between waves of data
collection.
Albeit important, these findings need to be considered in light of their shortcomings.
One limitation relates to the testing of mediation models using two waves of non-
experimental data. The mediating effects in this study operate through two causal paths:
(1) conspiracism ?efficacy and (2) efficacy ?participation. To incorporate the
temporal dimension of causality to both paths of the mediated effect, we should have
analysed the independent variable (X, conspiracism) in wave 1, the mediators (M,
efficacy) in wave 2, and the dependent variables (Y, participation) in a third wave that we
do not have (Kline, 2015; see also Hamaker, Kuiper, & Grasman, 2015; Roth &
16 Alberto Ard
evol-Abreu et al.
MacKinnon, 2012). Instead, we assumed stationarity of the effects and calculated both a
and bpaths using the same time points (W
1
and W
2
). In order to strengthen causal
interpretations and assumptions about directionality, future longitudinal studies should
collect at least three waves of data and examine both paths of the mediated effect in
connection with appropriate time lags (Kline, 2015).
Another caveat of using non-experimental data for mediation analyses is the possibility
that the estimation of the indirect effect may be biased towards overstatement –although,
as explained by Bullock, Green & Ha, 2010, these threats to inference are also problematic
in experimental designs. This overestimation bias is particularly likely to occur when
omitted variables influence M and Y ‘in the same direction’, or when the mediator is
correlated with another unobserved mediator/s (Bullock et al., 2010, pp. 552–555). We
have tried to minimize this source of error by employing a number of control variables (10
demographic and sociopolitical antecedents) that may affect both political efficacy and
participation, or act as mediators of the conspiracy–participation relationship. To address
omitted variable bias, future studies should experimentally manipulate both mediators
(internal and external efficacy), making sure that the intervention is targeting each
specific mediator (see Bullock et al., 2010).
Another caveat to consider is our measurement of unconventional political
behaviours. Non-institutionalized, protest-oriented activities can be legal or illegal,
depending on the nature of the action and the legal/institutional environment (Ekman &
Amn
a, 2012). Our measurement of unconventional participation focuses on legal actions,
except for the item referring to ‘demonstrations, protests, or marches’, which may be
legal or illegal. We deliberately excluded violent activities such as confronting the police,
damaging of public or private property, or squatting buildings because their occurrence
tends to be very infrequent in the general population (Ekman & Amn
a, 2012; Muller,
2015). Future studies should focus on more specific subsamples of the population to
include violent and illegal forms of protest, and examine how conspiracism relate to these
behaviours.
Future research should also explore country differences that may affect the substance
of the model developed in this study. This is an important point, because most of the time-
lagged relationships between our variables of interest –especially the indirect paths –
could not be reproduced at the level of individual countries. This may be in part a problem
of statistical power and sample size: our models include five variables of interest, four
autoregressive terms, and 10 controls, with relatively small country subsamples ranging
between n=784 and n=962. However, it is also likely that society- and country-level
characteristics show a direct influence on some of the outcomes assessed in this study or
moderate individual-level effects. We collected a cross-country sample with the purpose
of increasing the generalizability and external validity of our findings, but our analyses do
not use a multi-level approach.
Despite these considerations, the present article contributes the current debate on the
social consequences of exposure to conspiracy theories and endorsement of conspiracy
beliefs. All in all, the study clarifies the so far unexplored effects of conspiracy beliefs on
the different dimensions –personal- and system-regarding –of political efficacy, and how
they translate into people’s political behaviours. Our results showcase the negative
impact of conspiracism on democracy, especially in the light of what Barkun (2016) has
termed as ‘the mainstreaming of the fringe’, referred to the generalization, and acceptance
of previously stigmatized forms of knowledge (i.e., conspiracy theories). If conspiratorial
thinking can taint people’s efficacy perceptions and indirectly discourage conventional
forms of participation, the voice of the citizens will have less influence over their
Conspiracism and political participation 17
governments and political elites. Although it is true that most of the indirect effects found
herein are small and country-specific, it is also true that seemingly minor effects may exert
important negative influences when considering a large population (e.g., voters in a
national election).
Acknowledgements
This research was supported by Grant FA2386-15-1-0003 from the Asian Office of Aerospace
Research and Development. Alberto Ard
evol-Abreu is funded by the ‘Viera y Clavijo’ Program
from the Agencia Canaria de Investigaci
on, Innovaci
on y Sociedad de la Informaci
on (ACIISI),
and the Universidad de La Laguna (ULL). The authors are grateful to James H. Liu and all
participants of the Digital Influence World Project for their help with the data collection for this
study.
Conflicts of interest
All authors declare no conflict of interest.
Author contributions
Alberto Ard
evol-Abreu, PhD (Conceptualization; Data curation; Formal analysis; Investi-
gation; Methodology; Software; Supervision; Validation; Writing –original draft; Writing –
review & editing) Homero Gil de Z
u~
niga (Conceptualization; Funding acquisition;
Investigation; Project administration; Resources; Writing –review & editing) Elena G
amez
(Conceptualization; Writing –review & editing).
Data Availability Statement
The data that support the findings of this study will be made openly available in an open
repository.
References
Abramson, P. R., & Aldrich, J. H. (1982). The decline of electoral participation in America. American
Political Science Review,76, 502–521. https://doi.org/10.1017/S0003055400188379
Barkun, M. (2016). Conspiracy theories as stigmatized knowledge. Diogenes,1–7. https://doi.org/
10.1177/0392192116669288
Bogart, L. M., & Thorburn, S. T. (2006). Relationship of African Americans’ socio demographic
characteristics to belief in conspiracies about HIV/AIDS and birth control. Journal of the
National Medical Association,98, 1144–1150.
Bullock, J. G., Green, D. P., & Ha, S. E. (2010). Yes, but what’s the mechanism? (don’t expect an easy
answer). Journal of Personality and Social Psychology,98(4), 550. https://doi.org/10.1037/
a0018933
Butler, L. D., Koopman, C., & Zimbardo, P. G. (1995). The psychological impact of viewing the film
JFK: Emotions, beliefs, and political behavioral intentions. Political Psychology,16, 237–257.
https://doi.org/10.2307/3791831
Campbell, A., Gurin, G., & Miller, W. E. (1954). The voter decides. Evanston, IL: Row, Peterson and
Company.
18 Alberto Ard
evol-Abreu et al.
Clarke, S. (2002). Conspiracy theories and conspiracy theorizing. Philosophy of the Social Sciences,
32, 131–150. https://doi.org/10.1177/004931032002001
Cole, D. A., & Maxwell, S. E. (2003). Testing mediational models with longitudinal data: questions
and tips in the use of structural equation modeling. Journal of Abnormal Psychology,112(4),
558–577. https://doi.org/10.1037/0021-843x.112.4.558
Craig, S. C. (1980). The mobilization of political discontent. Political Behavior,2, 189–209.
https://doi.org/10.1007/BF00989890
Douglas, K. M., Sutton, R. M., & Cichocka, A. (2017). The psychology of conspiracy theories.
Current directions in Psychological Science,26(6), 538–542. https://doi.org/10.1177/
0963721417718261
Dyck, J. J., & Lascher, E. L. (2009). Direct democracy and political efficacy reconsidered. Political
Behavior,31(3), 401–427. https://doi.org/10.1007/s11109-008-9081-x
Ekman, J., & Amn
a, E. (2012). Political participation and civic engagement: Towards a new typology.
Human Affairs,22, 283–300. https://doi.org/10.2478/s13374-012-0024-1
Eveland, Jr, W. P., & Hively, M. H. (2009). Political discussion frequency, network size, and
“heterogeneity” of discussion as predictors of political knowledge and participation. Journal of
Communication,59, 205–224. https://doi.org/10.1111/j.1460-2466.2009.01412.x
Finkel, S. E. (1985). Reciprocal effects of participation and political efficacy: A panel analysis.
American Journal of Political Science,29, 891–913. https://doi.org/10.2307/2111186
Gamson, W. A. (1968). Power and discontent. Homewood, IL: Dorsey.
Gil de Z
u~
niga, H., Ard
evol-Abreu, A., & Casero-Ripoll
es, A. (2019). WhatsApp political discussion,
conventional participation and activism: exploring direct, indirect and generational effects.
Information, Communication & Society,1–18. https://doi.org/10.1080/1369118X.2019.
1642933
Gil de Z
u~
niga, H., & Liu, J. H. (2017). Second screening politics in the social media sphere:
Advancing research on dual screen use in political communication with evidence from 20
countries. Journal of Broadcasting & Electronic Media,61(2), 193–219. https://doi.org/10.
1080/08838151.2017.1309420
Goertzel, T. (1994). Belief in conspiracy theories. Political Psychology,15, 731–742. https://doi.
org/10.2307/3791630
Goldberg, R. A. (2004). ‘Who profited from the crime?’ Intelligence failure, conspiracy theories and
the case of September 11. Intelligence and National Security,19, 249–261. https://doi.org/10.
1080/0268452042000302985
Hamaker, E. L., Kuiper, R. M., & Grasman, R. P. P. P. (2015). A critique of the cross-lagged panel
model. Psychological Methods,20(1), 102–116. https://doi.org/10.1037/a0038889
Hoftstadter, R. (2008). The paranoid style in American politics (5th ed.). New York, NY: Vintage
Books-Random House.
Imhoff, R., & Bruder, M. (2014). Speaking (un-)truth to power: Conspiracy mentality as a generalised
political attitude. European Journal of Personality,28,25–43. https://doi.org/10.1002/per.
1930
Inglehart, R. (1977). The silent revolution. Princeton, NJ: Princeton University Press.
Jolley, D., & Douglas, K. M. (2014a). The effects of anti-vaccine conspiracy theories on vaccination
intentions. PLoS ONE,9,1–9. https://doi.org/10.1371/journal.pone.0089177
Jolley, D., & Douglas, K. M. (2014b). The social consequences of conspiracism: Exposure to
conspiracy theories decreases intentions to engage in politics and to reduce one’s carbon
footprint. British Journal of Psychology,105,35–56. https://doi.org/10.1111/bjop.12018
Jones, J., & Saad, L. (2018). Gallup poll social service: Environment. Gallup News Service. Retrieved
from http://news.gallup.com/file/poll/231575/180327ClimateChange.pdf
Jung, N., Kim, Y., & Gil de Z
uniga, H. (2011). The mediating role of knowledge and efficacy in the
effects of communication on political participation. Mass Communication and Society,14(4),
407–430. https://doi.org/10.1080/15205436.2010.496135
Conspiracism and political participation 19
Keum, H., Devanathan, N., Deshpande, S., Nelson, M. R., & Shah, D. V. (2004). The citizen-
consumer: Media effects at the intersection of consumer and civic culture. Political
Communication,21(3), 369–391. https://doi.org/10.1080/10584600490481479
Kline, R. B. (2015). The mediation myth. Basic and Applied Social Psychology,37, 202–213.
https://doi.org/10.1080/01973533.2015.1049349
Kramer, R. M., & Schaffer, J. (2014). Misconnecting the dots: origins and dynamics of out-group
paranoia. In J. W. van Prooijen, P. A. M. van Lange (Eds.), Power, politics, and paranoia. Why
people are suspicious of their leaders (pp. 199–217). Cambridge, UK: Cambridge University
Press.
Lantian, A., Muller, D., Nurra, C., & Douglas, K. M. (2017). I know things they don’t know! Social
Psychology,48(3), 160–173. https://doi.org/10.1027/1864-9335/a000306
Lee, F. L. (2010). The perceptual bases of collective efficacy and protest participation: The case of
pro-democracy protests in Hong Kong. International Journal of Public Opinion Research,
22(3), 392–411. https://doi.org/10.1093/ijpor/edq023
Mason, F. (2002). A poor person’s cognitive mapping. In P. Knight (Ed.), Conspiracy nation: the
politics of paranoia in postwar America (pp. 40–56). New York, NY: New York University
Press.
Miller, W. E., Miller, A. H., & Schneider, E. J. (1980). American National Election Studies data
source-book: 1952–1978. Cambridge, MA: Harvard University Press.
Morrell, M. (2003). Survey and experimental evidence for a reliable and valid measure of internal
political efficacy. Public Opinion Quarterly,67(4), 589–602. https://doi.org/10.1086/378965
Muller, E. N. (2015). Aggressive political participation. Princeton, NJ: Princeton University Press.
Niemi, R. G., Craig, S. C., & Mattei, F. (1991). Measuring internal political efficacy in the 1988
National Election Study. The American Political Science Review,85, 1407–1413. https://doi.
org/10.2307/1963953
Oliver, J. E., & Wood, T. (2014). Medical conspiracy theories and health behaviors in the United
States. JAMA Internal Medicine,174, 817–818. https://doi.org/10.1001/jamainternmed.2014.
190
Pollock, III, P. H. (1983). The participatory consequences of internal and external political efficacy:
A research note. Western Political Quarterly,36(3), 400–409. https://doi.org/10.1177/
106591298303600306
Reicher, S. D. (2004). The context of social identity: Domination, resistance and change. Political
Psychology,25, 921–946. https://doi.org/10.1111/j.1467-9221.2004.00403.x
Rose, C. L. (2017). The measurement and prediction of conspiracy beliefs (unpublished doctoral
dissertation). Victoria University of Wellington, New Zealand. Retrieved from http://researcha
rchive.vuw.ac.nz/handle/10063/6420
Roth, D. L., & MacKinnon, D. P. (2012). Mediation analysis with longitudinal data. In J. T. Newson, R.
N. Jones & S. M. Hofer (Eds.), Longitudinal data analysis: A practical guide for researchers in
aging, health, and social sciences (pp. 181–216). New York, NY: Routledge.
Sapountzis, A., & Condor, S. (2013). Conspiracy accounts as intergroup theories: Challenging
dominant understandings of social power and political legitimacy. Political Psychology,34,
731–752. https://doi.org/10.1111/pops.12015
Shah, D. V., Cho, J., Nah, S., Gotlieb, M. R., Hwang, H., Lee, N. J., ...McLeod, D. M. (2007). Campaign
ads, online messaging, and participation: Extending the communication mediation model.
Journal of Communication,57(4), 676–703. https://doi.org/10.1111/j.1460-2466.2007.
00363.x
Shaffer, S. D. (1981). A multivariate explanation of decreasing turnout in presidential elections,
1960–1976. American Journal of Political Science,25(1), 68–95. https://doi.org/10.2307/
2110913
Smith, M. (2019). Most people expect to feel the effects of climate change, and many think it will
make us extinct. YouGov. Retrieved from: https://today.yougov.com/topics/science/article
sreports/2019/09/16/global-climate-change-poll
20 Alberto Ard
evol-Abreu et al.
Stempel, C., Hargrove, T., & Stempel, III, G. H. (2007). Media use, social structure, and belief in 9/11
conspiracy theories. Journalism and Mass Communication Quarterly,84, 353–372. https://
doi.org/10.1177/107769900708400210
Swami, V., & Coles, R. (2010). The truth is out there: Belief in conspiracy theories. The Psychologist,
23, 560–563.
Swami, V., Coles, R., Stieger, S., Pietschnig, J., Furnham, A., Rehim, S., & Voracek, M. (2011).
Conspiracist ideation in Britain and Austria: Evidence of a monological belief system and
associations between individual psychological differences and real-world and fictitious
conspiracy theories. British Journal of Psychology,102, 443–463. https://doi.org/10.1111/
j.2044-8295.2010.02004.x
Swami, V., Furnham, A., Smyth, N., Weis, L., Lay, A., & Clow, A. (2016). Putting the stress on
conspiracy theories: Examining associations between psychological stress, anxiety, and belief in
conspiracy theories. Personality and Individual Differences,99,72–76. https://doi.org/10.
1016/j.paid.2016.04.084
United Nations Development Programme. (2015). Human development report 2015: Work for
human development. Retrieved from http://hdr.undp.org/sites/default/files/2015_huma
n_development_report.pdf
Uscinski, J. E., & Parent, J. M. (2014). American conspiracy theories. Oxford, UK: Oxford University
Press.
van Prooijen, J. W. (2017). Why education predicts decreased belief in conspiracy theories. Applied
Cognitive Psychology,31,50–58. https://doi.org/10.1002/acp.3301
van Stekelenburg, J., & Klandermans, B. (2013). The social psychology of protest. Current
Sociology,61, 886–905. https://doi.org/10.1177/0011392113479314
van Zomeren, M., Postmes, T., & Spears, R. (2008). Toward an integrative social identity model of
collective action: A quantitative research synthesis of three socio-psychological perspectives.
Psychological Bulletin,134, 504–535. https://doi.org/10.1037/0033-2909.134.4.504
Verba, S., & Nie, N. H. (1987). Participation in America: Political democracy and social equality.
Chicago, IL: University of Chicago Press (Original work published in 1972).
Watts, M. W. (1973). Efficacy, trust, and orientation toward socio-political authority: Students’
support for the university. American Journal of Political Science,17(2), 282–301, https://doi.
org/10.2307/2110521
Wolak, J. (2018). Feelings of political efficacy in the fifty states. Political Behavior,40(3), 763–784.
https://doi.org/10.1007/s11109-017-9421-9
Wood, M. J., Douglas, K. M., & Sutton, R. M. (2012). Dead and alive beliefs in contradictory
conspiracy theories. Social Psychological and Personality Science,3, 767–773. https://doi.org/
10.1177/1948550611434786
Zhang, X., & Lin, W. Y. (2018). Stoking the fires of participation: extending the Gamson hypothesis
on social media use and elite-challenging political engagement. Computers in Human
Behavior,79, 217–226. https://doi.org/10.1016/j.chb.2017.10.036
Received 19 July 2019; revised version received 16 January 2020
Conspiracism and political participation 21
