Joost de Moor
University of Leuven, Center for Citizenship and Democracy, Parkstraat 45,
B-3000 Leuven (Belgium), [email protected]
Reinterpreting the relation between external efficacy and political
participation
The role of output structures for state- and non-state oriented action forms
Paper presented at the 5th ECPR General Conference
Glasgow, 4-6 September 2014
Abstract. From an instrumental point of view, political efficacy is one of the most important
predictors of political participation. The more effective people consider certain forms of
action to be, the more likely they are to engage in them. Numerous studies have stressed the
role of citizens’ perceptions of the state as a political context to explain (the perceived
effectiveness of) political participation, yet in doing so they have mainly focused on
perceptions of the state’s willingness to be responsive – i.e., citizens’ external input efficacy.
Perceptions of the state’s ability to be responsive – i.e., external output efficacy – are largely
overlooked. The main goal of this paper is to address this gap in the literature by analyzing
the link between external output efficacy and different forms of state- and non-state oriented
forms of participation. To do so, data are used from the 2014 PARTIREP Belgian election
survey (N = 2015). In line with the hypotheses, it is found that high levels of external output
efficacy indirectly increase state-oriented political participation, while non-state oriented
political participation is unaffected.
Key words: Political participation, political efficacy, external efficacy.
Introduction
Citizens’ political participation ‘beyond the vote’ is often considered to be an essential
condition for a good rule of, by, and for the people (most recently: della Porta, 2013). For
one, increased levels of political participation improve citizen-elite congruency, as well as
citizens’ trust in, and satisfaction with, democracy (e.g., Putnam, 1993; Verba & Nie, 1972).
Because of the importance of political participation to democracy, an important body of
literature has sought to explain why citizens become politically active, understanding political
participation as a means for citizens to change political outcomes and to foster social change
(e.g., van Stekelenburg & Klandermans, 2013; van Zomeren, Postmes, & Spears, 2008;
Verba, Schlozman, & Brady, 2002). From such an instrumental point of view, citizens are
considered to be most likely to engage in political participation if they consider doing so to be
an effective means to these goals (Marien, Hooghe, & Quintelier, 2010; van Zomeren et al.,
2008; Verba et al., 2002). In line with these studies, this paper aims to contribute to the
literature explaining political participation by further analyzing what determines citizens’
expectations of the effectiveness of participating, and by analyzing whether this affects their
actual political participation.
To explain such expectations of effectiveness, this study looks at citizens’ perceptions
of state responsiveness – i.e., at their external efficacy (Niemi, Craig, & Mattei, 1991).
Various studies have indicated that citizens’ perceptions of the political context play a major
role in determining their (perceived effectiveness of) political participation (Bandura, 1986;
Lee, 2010; McAdam, 1982). In particular, it is expected that if citizens consider state
authorities to be responsive to their demands, they will consider participation to be more
effective, and consequently, they will become more likely to participate (Corcoran,
Pettinicchio, & Young, 2011; Karp & Banducci, 2008; Kriesi, Koopmans, Duyvendak, &
Giugni, 1995). If they consider the authorities to be irresponsive instead, participation will
appear ineffective, and hence, unappealing. In short, in order to understand (the perceived
effectiveness of) political participation, it is important to take into account citizens’ external
efficacy.
However, it is often overlooked that state responsiveness consists of at least two
elements: 1) whether authorities are willing to take citizens’ demands into account, and 2)
whether or not they are able to effectively produce political output (Hutter, 2014; Kitschelt,
1986; Kriesi et al., 1995). Citizens’ perceptions of the willingness of authorities to take their
demands into account – i.e., external input efficacy – has often been studied, and it has been
1
found to relate positively to political participation (Karp & Banducci, 2008; Niemi et al.,
1991). In contrast, citizens' perceptions of the ability of the political system to produce
political outcomes – i.e., external output efficacy – has rarely been studied. As a result, it
remains largely unknown whether and how external output efficacy affects (the perceived
effectiveness of) political participation. Therefore, this study aims to address this gap in the
literature by answering the following research question: What is the indirect effect of external
output efficacy on certain forms of political participation, as mediated by the perceived
effectiveness of those forms of participation?
To address this question, this paper uses data from the 2014 PARTIREP election
survey. This survey is the first large-N survey to include a measure of external output
efficacy. These data are therefore uniquely appropriate to address the research question. As
the research question concerns an indirect effect, the data will be analyzed using mediation
analyses. The structure of the remainder of this paper is as follows: I will first discuss the
literature on the link between political participation, political efficacy, and the perception of
input and output structures. Next, I will introduce the data, measurements, and methods used,
after which I will present the results of the analyses. Finally, I will conclude with a discussion
of the theoretical implications of the findings.
Theoretical framework
State- and non-state oriented political participation
People are using an increasingly wide range of activities in order to pursue political goals or
to express their political views. In reaction to this ongoing trend, scholars have begun to reconceptualize political participation in order to include this expanding variety of activities
(e.g., Norris, 2002; Zukin, Keeter, Andolina, Jenkins, & Delli Carpini, 2006). Most recently,
Van Deth has conceptualized political participation as “citizens’ activities affecting politics”
(2014, p. 3). Building on this conceptualization, actions can be identified as political
participation if they are carried out on a voluntary and amateurish basis, and if they apply to
one of the following criteria: 1) they are carried out within the institutional arena of state
politics, 2) they are targeted at the state, its personnel or its institutions, 3) or they are targeted
at
community
problems
by
targeting
non-state
actors.
Thus,
while
previous
conceptualizations of political participation focused mainly on the first two categories of
2
state-oriented activities (e.g., Verba, Nie, & Kim, 1978), more recent conceptualizations of
political participation have increasingly underlined the importance of both state- and nonstate oriented activities (Norris, 2002; van Deth, 2014).
This study aims to increase our understanding of political participation by analyzing
people’s perceptions of state responsiveness. However, taking the above into account, it
becomes clear that perceptions of the state have different implications for different types of
participation, depending on the degree to which they are state-oriented. Therefore, this study
focuses on three different types of action that represent the varying degrees to which common
types of political participation can be orientated at the state (Norris, 2002; van Deth, 2014;
Zukin et al., 2006). Firstly, it looks at citizens contacting politicians through mail or email as
a form of political participation that is acted out within the institutional arena of state politics
and that is targeted at state authorities. Secondly, it looks at citizens joining a demonstration
as a form of participation that is located outside the institutional arena of state politics and
that may be targeted either at state authorities or at non-state actors. Thirdly, it looks at
citizens boycotting products as a form of participation that is located outside the institutional
arena of state politics, and that is mainly targeted at non-state actors. These variations will be
taken into account when analyzing the effect of people’s perceptions of state responsiveness.
In the remainder of this theoretical framework, I will first discuss how feelings of
political efficacy relate to political participation. I will then discuss the role of the perceived
political context in explaining feelings of efficacy regarding certain forms of participation.
Here, I wish to make the main argument of this paper, namely, that perceptions of the state’s
output structure, i.e., external output efficacy, are essential in this regard. Finally, I will
formulate a number of hypotheses concerning how external output efficacy may affect (the
perceived effectiveness of) different types of political participation, depending on the degree
to which these types of participation are targeted at the state.
Efficacy and political participation
The antecedents of political participation may vary, but often citizens become politically
active in order to alter social conditions they consider unjust or unfair (Klandermans, 1984;
van Zomeren et al., 2008; Verba et al., 2002). In other words, political participation is often
instrumentally motivated as it advances goals that are external to the action (Rucht, 1990;
Walgrave, Van Laer, Verhulst, & Wouters, 2010). For instance, people contact politicians to
encourage them to take into account environmental concerns, they protest to undo austerity
3
measures, or they boycott products because they want to reverse the unethical production of
certain products (e.g., Balsiger, n.d.; Graziano & Forno, 2012; Rüdig & Karyotis, 2013).
From an instrumental perspective, people are considered to be more likely to become
politically active if they consider doing so to be effective (Gamson, 1968; Klandermans,
1997; van Zomeren et al., 2008). Therefore, the literature explaining political participation
has paid great attention to the concept of political efficacy. Political efficacy refers to “the
feeling that individual political action does have, or can have, an impact upon the political
process, (…) the feeling that political and social change is possible, and that the individual
citizen can play a part in bringing about this change” (Campbell, Gurin, & Miller, 1954, p.
187). Unsurprisingly, political efficacy is often found to predict political participation: the
more effective people believe political participation to be, the more likely they are to engage
in it (Hooghe & Marien, 2010; Verba et al., 2002).
However, it is important to keep in mind that the link between political efficacy and
political participation may vary regarding different forms of participation, leading to
corresponding variations in people’s engagement in specific forms of participation (Hooghe
& Marien, 2013b; Morrell, 2005). For instance, someone may feel that joining a
demonstration is highly effective, but that signing a petition is useless. S/he consequently
becomes more likely to engage in the former activity, while abstaining from the latter. In
short, the relation between political efficacy and political participation depends on the form
of participation at hand, and existing research has shown that feelings of efficacy regarding
specific forms of participation are better predictors of those forms of political participation
than general feelings of political efficacy (Morrell, 2005; Wollman & Strouder, 1991).
Building on these arguments, this study aims to explain specific forms of political
participation by looking at how effective people consider those particular forms of
participation to be.
H1: The more effective respondents consider a certain form of political participation
to be, the more likely they are to engage in this form of political participation.
External efficacy and output structures
In order to explain people’s beliefs regarding the effectiveness of certain forms of political
participation, it is instructive to analyze their perceptions of opportunities (or constraints)
within the political context. This argument has been put forward primarily in the literature on
4
external efficacy (Karp & Banducci, 2008; Niemi et al., 1991). External efficacy refers “to
beliefs about the responsiveness of governmental authorities and institutions to citizen
demands” (Niemi et al., 1991, pp. 1407–8). People with higher external efficacy are thus
inclined to believe that their demands will be taken into account by state authorities, and they
will therefore consider political participation to be more effective (Niemi et al., 1991).
Consequently, they have a higher propensity for political participation (e.g., Copeland, 2014;
Karp & Banducci, 2008).
However, whether or not the state appears to be responsive depends on two aspects.
On the one hand, it is determined by whether state authorities are willing to take citizens
claims into account, or in other words, whether its input structure is open (as opposed to
closed). On the other hand, it depends on its ability to effectively produce political output, or
in other words, whether its output structure is strong, as opposed to weak (Hutter, 2014;
Kriesi et al., 1995; Micheletti, 2003). In order for the state to be responsive, it needs to be
willing, as well as able to translate citizens demands into political output. Hence, both
perceptions of input structures – i.e., external input efficacy – and output structures – i.e.,
external output efficacy – are expected to affect people’s (perceived effectiveness of) political
participation (Kriesi et al., 1995; Stolle, Hooghe, & Micheletti, 2005).
Nonetheless, existing studies have generally only measured external efficacy as
individuals’ perception of the state’s willingness to take citizens’ demands into account
(Balch, 1974; Niemi et al., 1991). Individuals’ perceptions of the state’s ability to produce
political output are rarely taken into account. Consequently, the literature on external efficacy
and political participation has painted a one-sided picture, that leaves unanswered the
question how external output efficacy affects (the perceived effectiveness of) political
participation. It is the main goal of this study to address this gap in the literature.
Hypotheses
To sum up, the current study builds on two common theoretical arguments to explain
variations in people’s political participation. Firstly, to understand why some citizens engage
in certain forms of political participation (while other do not), it is important to take into
account whether or not they believe such forms of participation can be effective (Marien et
al., 2010; Verba & Nie, 1972; Verba et al., 2002). Secondly, to understand why some citizens
believe certain forms of political participation can be effective (while others do not), we need
to take into account their perceptions of the political context (Karp & Banducci, 2008; Lee,
5
2010; McAdam, 1982). Combined these suggestions propose that there is an indirect effect of
perceptions of the political context – i.e., external efficacy – on certain forms of political
participation, as mediated by perceptions of the effectiveness of those participation forms.
While existing studies have mainly focused on external input efficacy, this study aims to
extend this argument to include external output efficacy.
However, in doing so, it is important to recall that this argument applies to different
forms of participation in different ways. That is, depending on the degree to which specific
types of participation are state-oriented, they may be more or less affected by external output
efficacy. As argued above, political participation includes various activities that vary with
regard to their orientation towards the state (Norris, 2002; van Deth, 2014). The more stateoriented an activity is, the more it will depend on state responsiveness in order to be effective.
State-oriented forms of participation advance social change by making claims towards state
authorities to demand political change. If the state remains irresponsive to such claims, they
remain ineffective. To state the obvious, non-state oriented forms of participation do not
depend on state responsiveness in such a way. Hence, depending on the degree to which a
form of participation is state-oriented, the hypothesized indirect effect of external output
efficacy will vary, resulting in three different effects. Firstly, it is expected that the effect will
be strongest in the case of types of political participation that mainly target state actors (e.g.,
contacting politicians through mail or email). Secondly, it is expected that the effect will be
present, yet weaker, in the case of types of participation that can be targeted either at state, or
non-state actors (e.g., joining a demonstration). Finally, it is expected that the effect will be
absent in the case of non-state oriented types of participation (e.g., boycotting a product).
H2: There is a positive indirect effect of external output efficacy on state-oriented
forms of political participation, as mediated by the perceived effectiveness of those
forms.
H3: The more state oriented the form of participation, the stronger the indirect
effect.
H4: There is no indirect effect of external output efficacy on non-state-oriented
forms of political participation.
6
Data and measures
In order to test the hypotheses proposed above, data is used from the 2014 PARTIREP preelection survey. This telephone survey contains data from representative sample of 2015
Belgian adults, including both French and Dutch speaking citizens. The survey was
conducted prior to the three-level (regional, national, and European) elections of May 25,
2014. The fact that the PARTIREP survey contains data from only the Belgian population
may have important implications for the generalizability of the findings, yet there is sufficient
evidence that the Belgian case is representative of most other developed European
democracies. As Hooghe and Marien (2012) have observed, “analyses of the European Social
Survey have shown repeatedly that it [Belgium] is not exceptional with regard to
participation patterns or political attitudes in the European context” (p. 7). Thus, although
further comparative analyses would merit the generalizability of this study, the Belgian case
is interesting beyond its own borders as well.
Dependent variables: state and non-state oriented political participation
This study aims to explain why people engage in, or abstain from, various common forms of
state and non-state oriented political participation. As a typical example of state-oriented
participation, contacting politicians through mail or email is used. As a typical example of
non-state oriented participation, boycotting products is used. Joining in a demonstration is
used as an example of a ‘mixed’ form, targeting both state- and non-state actors. For each of
these types of participation, respondents were asked whether they had made use of it during
the last 12 months. For each form of participation a binary measure was created where
respondents who had participated were coded as 1, while others were coded as 0.i
Independent variable: external output efficacy
Unlike external input efficacy, external output efficacy has not yet been measured in large-N
survey research. The PARTIREP survey is the first large-N survey that includes a measure of
citizens’ external output efficacy. Respondents were asked to what extent they agree or
disagree (1 = completely disagree, 5 = completely agree) with the following statement:
“Politicians in my country are capable of acting upon problems”. This item taps in on
respondents’ beliefs concerning the state’s ability to act, or in other words, their external
7
output efficacy. In the question wording, the national level is emphasized because this study
is concerned with respondents’ perceptions of the state’s ability to act. The survey question
was used and tested previously in a mixed-methods case-study on Belgian environmental
activists (de Moor, Marie, & Hooghe, 2013). This study confirmed the usefulness of the
question in terms of understandability and construct validity. Thus, respondents who score
higher on this question believe more strongly that the national political system is capable of
acting. In other words, they perceive the state’s output structure as strong.
Mediating variables: the perceived effectiveness of political participation
The PARTIREP survey contains detailed information on how effective respondents consider
several individual forms of political participation, including contacting politicians, joining in
a demonstration, and boycotting. Concerning each of these forms of participation,
respondents were asked the following question: “Citizens can do various things to affect
political decisions. Can you indicate how effective you think each of these actions are in
affecting political decisions?”. Answers were given on a Likert-scale, ranging from 1 (“very
ineffective”) to 7 (“very effective”).
Control variables
It is hypothesized that each form of participation is affected by the perceived effectiveness of
that specific form of participation, but not by the perceived effectiveness of other forms of
participation. Therefore, the perceived effectiveness of the those other forms of participation,
respectively, will be included in the analyses of each form of participation as control
variables.
Moreover, as argued above, the existing literature has often focused on external input
efficacy, and it has been found to affect (the perceived effectiveness of) political
participation. Therefore, this variable will be included in the analyses as well. It is measured
using a single item where responds were asked to what extent they agreed or disagreed (1 =
completely disagree, 5 = completely agree) with the following statement: “The average
citizen does affect political decisions and the actions government takes”.
Additionally, several personal characteristics and political attitudes have repeatedly
been linked to (the perceived effectiveness of) political participation and therefore need to be
controlled for in our analyses. As argued above, men and older people generally feel more
8
efficacious about politics, and they are more inclined to engage in institutional forms of
participation (Burns, Schlozman, & Verba, 2001; Marien et al., 2010; Stolle & Hooghe,
2011). Women and young people have a stronger tendency to engage in non-institutional
forms of participation (Marien et al., 2010; Stolle & Hooghe, 2011). People with higher
education generally feel more efficacious, and overall they participate in politics more often
(Blais & Rubenson, 2013; Niemi et al., 1991). As for political attitudes, in general, people
with higher political interest and internal efficacy – i.e., the feeling that one can understand
and participate in politics (Niemi et al., 1991) – are more inclined to participate in politics
(Burns et al., 2001; Verba et al., 2002). Political trust is found to have a positive relation with
institutional participation, whereas it relates negatively to non-institutional participation
(Hooghe & Marien, 2013a). Thus age, sex (0 = men, 1 = women), and a categorical variable
for level of education (recoded to 1 = low, 2 = middle, 3 = high)ii are included as background
variables in all analyses. Political interest is measured using a single item where 0 means very
low political interest and 10 very high political interest. Political trust is measured using a
scale of eight items with an Eigenvalue of 4.17 and a Cronbach’s α of .86. Internal efficacy is
also measured using a scale of four items with an Eigenvalue of 2.03 and a Cronbach’s α of
.67.
Methods
The hypotheses formulated above suggest that the effect of external output efficacy on
political participation is mediated by perceptions of effectiveness. Mediation effects are
usually modeled using a method of decomposition, where the total effect of the independent
variable X on the dependent variable Y is decomposed into a direct effect and an indirect
effect. This is done by comparing regression parameters between the reduced model (i.e.,
without controlling for the mediator Z) and the full model (i.e., including the mediator Z)
(Alwin & Hauser, 1975; Baron & Kenny, 1986; Karlson, Holm, & Breen, 2010). Three steps
are used to decompose an effect. Firstly, separate regressions are used to predict the effect of
an independent variable X on a mediator Z (effect a), of a mediator Z on a dependent variable
Y (effect b), and of an independent X on dependent variable Y (c, or the total effect of X on
Y). An indirect effect (ab) occurs when a and b are both significant (Baron & Kenny, 1986;
Kohler, Karlson, & Holm, 2011). In a second step, the full model is predicted, where the
9
effect of X on Y is controlled for Z, hence predicting the direct effect (c`). Finally, by
subtracting c` from c, the size of the indirect effect (ab) is obtained.
The decomposition method builds on the assumption that the mediation effect is built
up of linear effects between the independent, mediating and dependent variables (Baron &
Kenny, 1986; Karlson et al., 2010). However, this assumption cannot be met in the current
study, because unlike the mediating variables, the dependent variables are dichotomous and
predicted using logistic regression analysis. As a result, the decomposition method cannot be
applied in the straightforward fashion described above. The problem is that the inclusion of
an additional (mediating) variable in a logistic regression not only affects the effect sizes, but
also in the scaling of the parameters (for more information see: Karlson et al., 2010; Kohler et
al., 2011). Therefore, comparing c and c’ using logistic regressions would conflate mediation
and rescaling. In this scenario, we cannot assess whether or not a mediating effect occurs, nor
how large it is.
To address this problem, Karlson, Holm and Breen (2010) propose the KHB-method
that corrects this limitation of the decomposition method for nonlinear probability models. It
includes the standardized residuals of the regression of X on Z in the reduced model, thereby
ensuring that the coefficients in the different models are measured on the same scale.
Consequently, the KHB-method warrants against the conflation of mediation and rescaling,
and coefficients can be standardized and compared across different linear and non-linear
regression models, thereby enabling decomposition. This method will be applied using the
‘khb’ program in Stata 12.
Throughout all regression analyses, robust standard errors will be used.
Results
Before turning to the analyses, it is useful to take into account some descriptive statistics that
give us a broad picture of (the perceived effectiveness of) political participation within our
sample (see Appendix for further descriptive statistics and bivariate correlations). On
average, people feel more efficacious about participating in a public demonstration (xˉ = 3.68,
SD = 1.57) than about boycotting products (xˉ = 3.53, SD = 1.63) or contacting politicians
(xˉ = 3.12, SD = 1.87). Boycotting products was the most prevalent form of participation (36
percent of the respondents indicated to have done so), followed by contacting politicians (18
10
percent), and joining a demonstration (14 percent). T-tests indicate that all the differences
between these means are statistically significant.
Figure 1: the hypothesized mediation effects
Perceived effectiveness
of:
a
External output efficacy
1: contacting politicians
2: demonstrating
3: boycott product
c'
b
Political participation:
1: contacting politicians
2: demonstrating
3: boycotting products
External output efficacy
c
Political participation:
1: contacting politicians
2: demonstrating
3: boycotting products
The hypotheses formulated above concern the indirect effect of external output efficacy on
different forms of political participation, as mediated by the perceived effectiveness of each
of these forms. To test this mediation effect, the direct effects between the independent,
mediating and dependent variables will be analyzed, i.e., effects a, b, and c in Figure 1
(Alwin & Hauser, 1975; Baron & Kenny, 1986). If both effects a and b are significant, a
mediation effect occurs. The KHB method will then be applied to decompose the total effect
of external output efficacy into a direct (c’) and an indirect (ab) effect on political
participation (Karlson et al., 2010; Kohler et al., 2011). These analyses will be repeated
separately for each of the three individual forms of political participation, starting with
contacting politicians, followed by participating in a demonstration, and finally, boycotting
products.
11
Contacting politicians
Table 1 contains the results from the logistic regressions on contacting politicians. In Model I
we see that people with higher education, higher political interest and higher internal efficacy
are significantly more likely to contact politicians. This is in line with previous findings in the
literature. External output efficacy is negatively associated with contacting politicians (path
c), but this effect is not significant.
Table 1: Logistic regression of contacting a politician
Predictor
Control Variables
Intercept
Age
Sex (1 = female)
Education (ref. = low)
Middle
Higher
Political Interest
Political Trust
Internal Efficacy
External input efficacy
Model I
Model II
Model III
Model IV
.024*** (.011)
1.002 (.004)
1.075 (.136)
.006*** (.003)
1.001 (.004)
.998 (.133)
.006*** (.003)
1.000 (.004)
1.003 (.135)
.010*** (.005)
1.002 (.004)
.991 (.133)
1.760** (.309)
2.459*** (.424)
1.755** (.314)
2.234*** (.388)
1.714** (.312)
2.185*** (.384)
1.7221** (.309)
2.158*** (.375)
1.064** (.024)
.987 (.043)
1.628***(.136)
1.114 (.063)
1.077** (.026)
.948 (.051)
1.595*** (.137)
1.002 (.061)
1.076** (.027)
.943 (.050)
1.584*** (.138)
1.000 (.062)
1.073** (.026)
.976 (.049)
1.589*** (.135)
1.042 (.063)
Variables of Interest
External output efficacy
PE Contacting politiciansa
PE Demonstratinga
PE Boycotting producta
.882 (.059)
.772** (.056)
1.564*** (.070)
1.561*** (.074)
.987 (.046)
1.050 (.039)
1.605*** (.073)
McKelvey & Zavoina’s pseudo R2
.12
.22
.22
.23
N
1910
1901
1878
1896
Note: *p < .10, **p < . 05, ***p < .01. Coefficients are odds ratios. Robust standard errors between brackets. a:
PE = Perceived effectiveness.
Model XIII in Table 4 (p. 17) presents the results from an OLS regression on the perceived
effectiveness of contacting politicians. We see that people with high education perceive
contacting politicians as more effective than people with low education. Moreover, a positive
effect exists for people with higher internal efficacy and for people with higher external input
efficacy. Again, these findings are in line with the literature discussed above. Finally, in
support of hypothesis 2, we see that there is a significant positive effect of external output
efficacy on the perceived effectiveness of contacting politicians (effect a). An increase of one
12
unit on external output efficacy is predicted to increase the mediating variable by .25. It is
interesting to note that this effect is stronger than that of external input efficacy (.17). This
supports the argument made earlier that while the literature has mainly focused on the latter,
the former is at least as important in understanding why people feel more or less efficacious
about state-oriented political participation.
As for path b, we see in Model II (Table 1) that perceiving contacting politicians as
effective positively affects one’s likelihood of contacting a politician. More precisely, a one
unit increase on the mediating variable increases one’s odds of having contacted a politician
by 56 percent. In order to test whether this link is specific for the perceived effectiveness of
this form of political participation, and not for the perceived effectiveness of political
participation in general, the perceived effectiveness of joining a demonstration and of
boycotting a product are included in Model III. We see that there is only a significant effect
for the perceived effectiveness of contacting politicians, and not for the perceived
effectiveness of the other forms of participation. These results support the hypothesis (H1)
that there is not one general feeling of efficacy leading to political participation, but that the
effect of efficacy beliefs are specific to the type of participation.
In Model IV the full model is presented, thus showing the effect of the independent
variable on the dependent variable under the control of the mediating variable. We see that
the direct negative effect of external output efficacy has become stronger and is now
significant (OR = .772**). This suggests that the mediating variable performs the role of
suppressor (Baron & Kenny, 1986; MacKinnon, Fairchild, & Fritz, 2007). That is, when the
positive association between the independent variable and the mediator are controlled for, the
negative effect of the independent variable becomes isolated and thereby significant.
Although no assumptions were made about the direct effect of external output efficacy in this
study, this negative effect is somewhat surprising, and it will therefore be touched upon
further in the discussion below. However, it does not contradict the hypothesized indirect
effect that is found, as we are in fact dealing with an ‘inconsistent mediation’ (i.e., the direct
effect of a predictor has the opposite sign from its indirect effect) (Alwin & Hauser, 1975;
Baron & Kenny, 1986; MacKinnon et al., 2007). This inconsistent mediation suggests that,
although people who perceive the state’s output structure as strong are generally less inclined
to contact politicians, they are more likely to perceive contacting politicians as effective. If
they do so, this increases the likelihood that they will contact politicians, thereby balancing
out the negative direct effect.
13
As hypothesized (H2), there is thus a significant indirect effect of external output
efficacy on contacting politicians. In order to assess how strong this effect is, and how much
of the total effect is mediated, the KHB-method is applied. The KHB method provides
average marginal effects (AME) of the total (c), direct (c’) and indirect effect (ab), where the
indirect effect is the difference between the total and the direct effect (ab = c-c’). The AME
of external output efficacy in the reduced model (c) is -.033, which implies that on average,
the probability of someone contacting a politician decreases by 1.3 percent points for one
standard deviation change of the independent variable. Under the control of the
mediator/suppressor, this effect increases (c’ = -.036). The indirect of external output efficacy
is thus .013, suggesting that an increase of one standard deviation of the independent variable
increases the probability someone contacts a politician by 1.3 percent points. In order to
calculate the share of the total effect that is mediated in an inconsistent mediation, ab should
be expressed as a percentage of the sum of the absolute values of ab and c’ (Alwin & Hauser,
1975; MacKinnon et al., 2007): |.013|/(|.013|+|.036|)=.265. Put differently, 27 percent of the
total effect of external output efficacy on contacting politicians is mediated by the perceived
effectiveness of doing so.
Joining a demonstration
Demonstrations can be targeted at both state and non-state actors. Therefore it was
hypothesized (H3) that although there is still a mediated positive effect of external output
efficacy on joining a demonstration, this effect is weaker than in the case of the essentially
more state-oriented act of contacting politicians. Again, we use the method of decomposition,
after which the indirect effect will be calculated using the KHB method.
In Model V (Table 2) we observe that higher education, internal efficacy and external
input efficacy are positively associated with joining in a demonstration. There is a negative,
but non-significant direct effect of external output efficacy (c). However, in Model XIV
(Table 4) we see that there is a significant positive effect of external output efficacy on the
perceived effectiveness of joining a demonstration (a), while in Model VI (Table 2) we see
that people who perceive protesting as more effective are more likely to join a demonstration
(b). Again, Model VII suggests that joining a demonstration is only affected by the perceived
effectiveness of this form of action, and not of the others, thus offering further support for
hypothesis 1.
14
There is thus a significant positive indirect effect (ab). This finding is further
supported when we look at the full model, where the effect of external output efficacy is
controlled for the perceived effectiveness of joining a demonstration. Here, we see that
although effect becomes stronger and significant. This suggests that again we are again
dealing with a mediator that functions as a suppressor, constituting an inconsistent mediation.
Table 2: Logistic regression of joining a demonstration
Predictor
Control Variables
Intercept
Age
Sex (1 = female)
Education (ref. = low)
Middle
Higher
Political Interest
Political Trust
Internal Efficacy
External input efficacy
Model V
Model VI
Model VII
Model VIII
.067*** (.032)
.991* (.004)
.781 (.111)
.014*** (.007)
.993 (.004)
.754 (.109)
.013*** (.007)
.993 (.004)
.768 (.112)
.019*** (.009)
.994 (.004)
.747* (.108)
1.469* (.270)
1.517* (.279)
1.434 (.268)
1.453* (.273)
1.421 (.269)
1.451 (.277)
1.418 (.265)
1.424 (.267)
1.023 (.023)
.987 (.049)
1.498*** (.135)
1.115 (.068)
1.034 (.025)
.979 (.057)
1.450*** (.131)
1.036 (.067)
1.034 (.025)
.992 (.058)
1.475*** (.135)
1.051 (.068)
1.031 (.025)
.996 (.056)
1.441*** (.129)
1.063 (.068)
Variables of Interest
External output efficacy
PE Contacting politicians
PE Demonstrating
PE Boycotting product
.895 (.063)
.857* (.063)
1.430*** (.065)
.930 (.046)
1.466*** (.073)
.996 (.040)
1.440*** (.065)
McKelvey & Zavoina’s pseudo R2
.08
.15
.16
.16
N
1911
1913
1879
1908
Note: *p < .10, **p < . 05, ***p < .01. Coefficients are odds ratios. Robust standard errors between brackets. PE
= Perceived effectiveness.
The average marginal effects of the total and the direct effect are again obtained using the
KHB-method to calculate how much of the total effect is mediated. The difference between
the total effect (c = -.013) and the direct effect (c’ = -.016) is .003. In other words, an increase
of one standard deviation of external output efficacy significantly increases the probability
that someone participated in a demonstration with 0.3 percent. This indirect effect ab
constitutes 16 percent of the total effect, and is very small. However, this is in line with the
hypothesis that although there should be an indirect effect, this effect should be smaller than
in the case of the more state oriented action form of contacting politicians. Comparing effect
a and b between contacting politicians and joining a demonstration further indicates that the
15
indirect effect for the latter is smaller because external output efficacy less strongly predicts
the perceived effectiveness of joining a demonstration, whereas perceptions of effectiveness
of each form of participation are equally predictive of whether an individual engages in them.
Boycotting products
Whereas a positive mediation effect was hypothesized of external output efficacy on
contacting politicians, and to a lesser extent, on joining a demonstration, no such effect was
hypothesized (H4) to exist for non-state oriented forms of participation like boycotting
products. That is, although it is expected that people who consider boycotts to be effective
will be more likely to engage in them, their perceptions of effectiveness are most likely not to
be affected by their perceptions of the state. Hence, it is expected that path b is significant
and positive, but that path a is not, in which case, no significant indirect effect (ab) can be
established.
Table 3: Logistic regression of boycotting products
Predictor
Control Variables
Intercept
Age
Sex (1 = female)
Education (ref. = low)
Middle
Higher
Political Interest
Political Trust
Internal Efficacy
External input efficacy
Model IX
Model X
Model XI
Model XII
.280*** (.094)
.998 (.003)
1.063 (.108)
.086*** (.029)
.994 (.003)
1.051 (.112)
.094*** (.033)
.994* (.003)
1.054 (.112)
.111*** (.039)
.994 (.003)
1.045 (.111)
1.362* (.177)
2.099*** (.270)
1.329* (.182)
1.888*** (.252)
1.323* (.181)
1.870*** (.252)
1.306 (.179)
1.858*** (.249)
.991 (.018)
.942 (.033)
1.390*** (.090)
1.050* (.047)
.984 (.019)
.950 (.037)
1.340*** (.093)
.978 (.045)
.983 (.019)
.952 (.037)
1.352*** (.094)
.980 (.046)
.982 (.019)
.962 (.037)
1.344*** (.093)
1.000 (.047)
Variables of Interest
External output efficacy
PE Contacting politicians
PE Demonstrating
PE Boycotting product
.900* (.047)
.884* (.048)
1.448*** (.043)
.989 (.036)
.980 (.036)
1.456*** (.046)
1.448*** (.043)
McKelvey
&
Zavoina’s
pseudo R2
.06
.17
.17
.18
N
1906
1887
1876
1883
Note: *p < .10, **p < . 05, ***p < .01. Coefficients are odds ratios. Robust standard errors between brackets. a:
PE = Perceived effectiveness.
16
In Model IX (Table 3) we see that people with higher education are again more likely to
participate. People with higher internal efficacy and higher external input efficacy are also
more likely to participate. Again, these findings are in line with previous studies. External
output efficacy has a significant negative direct effect, which will be discussed further below.
In Model XV (Table 4) we see that there is no significant effect (a) of external output
efficacy on the perceived effectiveness of boycotting product. In Model X (Table 3) we see
that there is a significant positive effect of perceiving boycotting products as effective on
actually participating in this form of action. In Model XI we see that this effect is exclusively
significant for the perceived effectiveness of boycotting products, and not for the perceived
effectiveness of other forms of participation (in support of H1). Finally, in the full model
(Model XII, Table 3) we see that the direct effect of external output efficacy is not affected
by controlling for the perceived effectiveness of boycotting products. In short, as
hypothesized, there is no indirect effect of external output efficacy on the probability that
someone will engage in a boycott.
Table 4: OLS regression of the perceived effectiveness (PE) of individual forms of political
participation
Model XIII
Model XIV
Model XV
PE contacting
politicians
PE joining a
demonstration
PE boycotting a
product
1.449*** (.238)
-.001 (.002)
.159* (.073)
3.273*** (.250)
-.011*** (.002)
.058 (.077)
2.352*** (.292)
.008** (.003)
.066 (.088)
.055 (.089)
.394*** (.091)
.080 (.096)
.284*** (.094)
.150 (.109)
.483*** (.110)
-.018 (.013)
.021 (.027)
.101* (.048)
.166*** (.033)
-.039** (.015)
-.032 (.029)
.113 (.051)
.149*** (.034)
.012 (.016)
-.079* (.034)
.122* (.059)
.041** (.038)
.248*** (.038)
.143** (.041)
.044 (.048)
R2
.07
.05
N
1897
1908
Note: *p < .10, **p < . 05, ***p < .01. Robust standard errors between brackets.
.03
1886
Predictor
Control Variables
Intercept
Age
Sex (1 = female)
Education (ref. = low)
Middle
Higher
Political Interest
Political Trust
Internal Efficacy
External input efficacy
Variables of Interest
External output efficacy
17
In sum, we have seen that each specific form of participation is affected by the perceived
effectiveness of that form of participation, and not by the perceived effectiveness of the other
forms of participation. These findings support hypothesis 1, and are in line with previous
studies that have suggested that political participation becomes more likely when people
believe that specific action forms are effective (Marien et al., 2010; Morrell, 2005; van
Stekelenburg & Klandermans, 2013; Verba et al., 2002).
This study has primarily aimed to increase our understanding of why people do, or do
not, believe that certain forms of participation can be effective, thereby indirectly explaining
actual participation. More precisely, it has sought to explain political participation by looking
at people’s perception of the state’s ability to act – i.e., their external output efficacy. It was
hypothesized that this mediation effect should exist mainly for state-oriented forms of
participation, and to a lesser extent for forms of participation that can be targeted at both state
and non-state actors. No such effect was expected to exist for non-state oriented forms of
participation. In support of Hypothesis 2, 3 and 4, the results indicate that there is a
significant indirect effect of external output efficacy on contacting politicians and joining in a
demonstration. The effect is indeed stronger for the former. In further support of the
hypotheses, no such effect is found for the non-state oriented act of boycotting products.
Discussion
In line with most political participation literature, this study indicates that citizens are more
likely to engage in political participation if they consider doing so to be effective (Marien et
al., 2010; van Stekelenburg & Klandermans, 2013; Verba et al., 2002). This is the case for all
three forms of participation analyzed in this study: contacting politicians, joining a
demonstration, and boycotting products. Moreover, it suggests that perceptions of the
political context are significant predictors of the perceived effectiveness of political
participation, as well as actual political participation. In particular, it proposes that if citizens
consider the state to have a strong output structure, they are more likely to perceive stateoriented political participation as an effective means to social change. Consequently, there is
a positive indirect effect of external output efficacy on state-oriented forms of political
participation. No such effect occurs in the case of non-state oriented forms of participation
like boycotting products.
Whereas there is thus a positive indirect effect of external output efficacy on stateoriented political participation, the direct effect of external output efficacy on political
18
participation is in all cases negative. If citizens perceive the state to have a strong output
structure, it is found that they become less likely to act. This surprising finding does not
contradict our hypothesis, yet it still begs further reflection. In the case of state-oriented
forms of action, it may be that citizens who feel that the state is capable of addressing
problems in society consider political participation as less urgent or less necessary. In their
eyes, the state is effectively dealing with society’s challenges, and therefore, it does not need
correction through political action. In contrast, if citizens feel the state is not capable of
solving society’s main problems, citizens act to correct. As such these findings resonate an
argument that William Gamson made in 1968: “high trust in authorities implies some lack of
necessity for influencing them” (p. 7). In a similar vein, Almond and Verba (1963) described
the satisfied citizen, who may well feel efficacious, but who considers corrective political
participation to be unnecessary. In line with these descriptions, the results in this paper
suggest that if citizens trust in the authorities’ ability to act, they will feel less urged to
perform corrective pressure through state-oriented political participation. As for non-state
oriented action, it may be that if citizens perceive the output structure as weak, state-oriented
action may appear ineffective, and non-state oriented action becomes a reasonable
alternative.
Taking the above into account, the state’s perceived output structure thus performs an
important role in establishing a democratic linkage between citizens and the state in at least
two ways. On the one hand, when citizens perceive the output structure as strong, it appears
they feel it is less necessary to ‘correct’ authorities through political action. On the other
hand, a strong output structure presents an important perquisite for citizens to believe that
political participation can be effective, thus inciting them to participate in case they believe
doing so is necessary. Given this dual role of the state’s output structure, it is remarkable that
it has received such limited attention in the literature on political participation. Thus,
although the current study does support the common understanding that an open input
structure facilitates political participation, it stresses that perceptions of the output structure
clearly affect political participation as well.
Finally, it is important to note certain limitations of this study, and possible venues for
future research as well. Firstly, it needs to be recognized that the structure of the used data is
cross-sectional, and that this limits our ability to make strong causal claims regarding the link
between political attitudes (like efficacy) and political behavior. Most studies on this subject
indicate that political attitudes predict political behavior – e.g., higher political efficacy leads
to higher political participation (van Stekelenburg & Klandermans, 2013; van Zomeren et al.,
19
2008; Verba et al., 2002). However, other studies stress that the effect may actually go in the
opposite direction as well, arguing that political attitudes are shaped by the experience of
participation (Klandermans, van der Toorn, & van Stekelenburg, 2008; Quintelier & van
Deth, 2014). Notwithstanding the importance of this argument, however, most political
participation literature suggests that the direction of the effect goes mainly from attitudes to
political behavior. For one, people who have never engaged in political activities have
political attitudes nonetheless. In other words, political attitudes precede political
participation. Moreover, as Inglehart (2008) has suggested, political attitudes remain
relatively stable throughout people’s lives, rather than being changed after each individual
experience of political participation. Still, the findings in this study would merit from a
longitudinal or experimental study that could more strongly assess questions of causality.
Secondly, the findings in this study are limited to one case. Although the Belgian case
is often considered to be representative of other developed European democracies, assessing
whether this study’s findings will hold in different national contexts would advance the
generalizability of the findings. Moreover, a comparative approach would allow to assess
whether differences between countries in terms of output strength are reflected in citizens’
political attitudes and behavior. As for now, this study strongly suggests that the effect of the
perceived output structure on political participation will be supported by such a comparative
study.
20
References
Almond, G., & Verba, S. (1963). The Civic Culture. Political Attitudes and Democracy in
Five Nations. Newbury Park: Sage Publications.
Alwin, D. F., & Hauser, R. M. (1975). The Decomposition of Effects in Path Analysis.
American Sociological Review, 40(1), 37–47.
Balch, G. (1974). Multiple indicators in survey research: the concept “sense of political
efficacy.” Political Methodology, 1, 1–43.
Balsiger, P. (n.d.). Between shaming corporations and promoting alternatives. An in-depth
analysis of the tactical repertoire of a campaign for ethical fashion in Switzerland.
Journal of Consumer Culture.
Bandura, A. (1986). Social Foundations of Thought and Action: Social Cognitive Theory.
Englewood Cliffs: Prentice Hall.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social
psychological research: conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51(6), 1173–82.
Blais, A., & Rubenson, D. (2013). The source of turnout decline: New values or new
contexts? Comparative Political Studies, 46(1), 95–117.
doi:10.1177/0010414012453032
Burns, N., Schlozman, K. L., & Verba, S. (2001). The Private Roots of Public Action.
Cambridge: Harvard University Press.
Campbell, A., Gurin, G., & Miller, W. E. (1954). The Voter Decides. Evanston: Row &
Peterson.
Copeland, L. (2014). Conceptualizing political consumerism: How citizenship norms
differentiate boycotting from buycotting. Political Studies, 62(1), 172–186.
doi:10.1111/1467-9248.12067
Corcoran, K. E., Pettinicchio, D., & Young, J. T. N. (2011). The Context of Control: a CrossNational Investigation of the Link Between Political Institutions, Efficacy, and
Collective Action. The British Journal of Social Psychology, 50(4), 575–605.
doi:10.1111/j.2044-8309.2011.02076.x
De Moor, J., Marie, S., & Hooghe, M. (2013). Linking Lifestyle Politics and State Oriented
Action. A mixed Methods Inquiry into forms of engagement among lifestyle activitst in a
a Belgian Environmental Movement. Leuven.
Della Porta, D. (2013). Can Democracy Be Saved? Participation, Deliberation, and Social
Movements. Cambridge: Polity.
21
Gamson, W. A. (1968). Power and Discontent. Homewood: The Dorsey Press.
Graziano, P. R., & Forno, F. (2012). Political Consumerism and New Forms of Political
Participation: The Gruppi di Acquisto Solidale in Italy. The ANNALS of the American
Academy of Political and Social Science, 644(1), 121–133.
doi:10.1177/0002716212454839
Hooghe, M., & Marien, S. (2010). Political Trust, Political Efficacy and Forms of
Participation. Oxford: Nuffield College.
Hooghe, M., & Marien, S. (2012). How to reach Members of Parliament ? Citizens and
Members of Parliament on the Effectiveness of Political Participation Repertoires, 1–25.
Hooghe, M., & Marien, S. (2013a). A Comparative Analysis of the Relation Between
Political Trust and Forms of Political Particiaption in Europe. European Societies, 15(1),
131–52.
Hooghe, M., & Marien, S. (2013b). How to reach Members of Parliament? Citizens and
Members of Parliament on the Effectiveness of Political Participation Repertoires.
Parliamentary Affairs, 66, 1–25. doi:10.1093/pa/gss057
Hutter, S. (2014). New Cleavages and Protest Politics in Western Europe. Minneapolis:
University of Minnesota Press.
Inglehart, R. F. (2008). Changing values among Western publics from 1970 to 2006. West
European Politics, 31(1-2), 130–146. doi:10.1080/01402380701834747
Karlson, K. B., Holm, A., & Breen, R. (2010). Comparing Regression Coefficients Between
Models using Logit and Probit : A New Method. Retrieved July 07, 2014, from
http://www.yale.edu/ciqle/Breen_Scaling effects.pdf
Karp, J. A., & Banducci, S. A. (2008). Political Efficacy and Participation in Twenty-Seven
Democracies: How Electoral Systems Shape Political Behaviour. British Journal of
Political Science, 38(2), 311–334. doi:10.1017/S0007123408000161
Kitschelt, H. (1986). Political Opportunity Structures an Political Protest: Anti-Nuclear
Movements in Four Democracies. British Journal of Political Science, 16, 57–85.
Klandermans, B. (1984). Mobilization and participation: social-psychological expansions of
resource mobilization theory. American Sociological Review, 49(5), 583–600.
Klandermans, B. (1997). The Social Psychology of Protest. Oxford: Blackwell.
Klandermans, B., van der Toorn, J., & van Stekelenburg, J. (2008). Embeddedness and
identity: How immigrants turn grievances into action. American Sociological Review,
73(6), 992–1012. doi:10.1177/000312240807300606
Kohler, U., Karlson, K. B., & Holm, A. (2011). The Stata Journal. The Stata Journal, 11(3),
420–38.
22
Kriesi, H., Koopmans, R., Duyvendak, J. W., & Giugni, M. (1995). New Social Movements in
Western Europe: a Comparative Analysis. Minneapolis: University of Minnesota Press.
Lee, F. (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. doi:10.1093/ijpor/edq023
MacKinnon, D. P., Fairchild, A. J., & Fritz, M. S. (2007). NIH Public Access. Annual Review
of Psychology, 58(Hebb 1966), 1–22.
doi:10.1146/annurev.psych.58.110405.085542.Mediation
Marien, S., Hooghe, M., & Quintelier, E. (2010). Inequalities in Non-institutionalised Forms
of Political Participation: A Multi-level Analysis of 25 countries. Political Studies,
58(1), 187–213. doi:10.1111/j.1467-9248.2009.00801.x
McAdam, D. (1982). Political Process and the Development of Black Insurgency, 1930-1970.
Chicago: University of Chicago Press.
Micheletti, M. (2003). Political Virtue and Shopping: Individuals, Consumerism, and
Collective Action. New York: Palgrave.
Morrell, M. E. (2005). Deliberation, democratic decision-making and inernal political
efficacy. Political Behavior, 27(1), 49–69. doi:10.1007/s
Niemi, R. G., Craig, S. C., & Mattei, F. (1991). Measuring internal political efficacy in the
1988 National Election Study. American Political Science Review, 85(4), 1407–1413.
Norris, P. (2002). Democratic Phoenix. Reinventing Political Activism. Cambridge:
Cambridge University Press.
Putnam, R. D. (1993). Making Democracy Work: Civic Traditions in Modern Italy.
Princeton: Princeton University Press.
Quintelier, E., & van Deth, J. W. (2014). Supporting democracy: Political participation and
political attitudes. Exploring causality using panel data. Political Studies, 62(1934),
153–171. doi:10.1111/1467-9248.12097
Rucht, D. (1990). The strategies and action repertoires of new movements. In R. J. Dalton &
M. Kuechler (Eds.), Challenging the Political Order: New Social and Political
Movements in Western Democracies (pp. 156–75). Cambridge: Polity.
Rüdig, W., & Karyotis, G. (2013). Who Protests in Greece? Mass Opposition to Austerity.
British Journal of Political Science, 44(03), 487–513. doi:10.1017/S0007123413000112
Stolle, D., & Hooghe, M. (2011). Shifting inequalities. Patterns of exclusion and inclusion in
emerging forms of political participation. European Societies, 13(1), 119–142.
Stolle, D., Hooghe, M., & Micheletti, M. (2005). Politics in the Supermarket: Political
Consumerism as a Form of Political Participation. International Political Science
Review, 26(3), 245–269. doi:10.1177/0192512105053784
23
Van Deth, J. W. (2014). A conceptual map of political participation. Acta Politica, 1–19.
doi:10.1057/ap.2014.6
Van Stekelenburg, J., & Klandermans, B. (2013). The social psychology of protest. Current
Sociology, 0(0), 1–20. doi: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(4), 504–35. doi:10.1037/0033-2909.134.4.504
Verba, S., & Nie, N. H. (1972). Participation in America. Political Democracy and Social
Equality. Chicago, London: The University of Chicago Press.
Verba, S., Nie, N. H., & Kim, J. (1978). Participation and Political Equality: A Seven-Nation
Comparison. New York: Cambridge University Press.
Verba, S., Schlozman, K. L., & Brady, H. E. (2002). Voice and Equality. Civic Voluntarism
in American Politics. Cambridge: Harvard University Press.
Walgrave, S., Van Laer, J., Verhulst, J., & Wouters, R. (2010). Why People Protest.
Comparing Demonstrators’ Motivations Across Issues and Nations.
Wollman, N., & Strouder, R. (1991). Believed efficacy and political activity: a test of the
specificity hypothesis. The Journal of Social Psychology, 131, 557–566.
Zukin, C., Keeter, S., Andolina, M., Jenkins, K., & Delli Carpini, M. X. (2006). A New
Engagement? Political Participation, Civic Life, and the Changing American Citizen.
Oxford: Oxford University Press.
24
APPENDIX: Descriptives of survey items
Pearson’s correlations with variables of interest
N
Min.
max.
Mean
1. PEIPPa
2. PENIPPb
3. IPP
4. NIPP
1981
1966
2011
2007
1-7
1-7
0-1
0-1
3.79
3.35
.21
.63
5. External efficacy
6. Perceived output structure
1998
1998
1-4.67
1-5
2015
2015
2015
2013
1945
1989
18-84
0-1
1-3
0-10
0-10
1-5
Item
Std. dev.
1.
2.
3.
4.
5.
6.
1.29
1.18
.41
.48
.503***
.232***
.144***
.110***
.243***
.282***
-
2.43
3.12
.87
.98
.396***
.246***
.168***
.102***
.120***
-.024
.105***
-.077***
.267***
-
47.96
.50
6.09
4.42
2.72
17.32
.50
2.80
1.35
.80
-.052*
.023
-.030
.083***
.101***
-.050*
.021
-.053*
-.025
.089***
.008
-.048*
.099***
.038
.216***
-.146***
-.014
-.031
-.021
.177***
-.118***
-.053*
-.003
.177***
.103***
.074**
-.038
-.034
.128***
.013
Control variables
5. Age
6. Sex (1 = female)
7. Education
8. Political interest
9. Political trust
10. Internal efficacy
Note: *p < .05, **p < .01, ***p < .001.
a: perceived effectiveness of institutional political participation. b: perceived effectiveness of non-institutional political
participation.
25
Notes
i
The original questionnaire contained 4 possible answers as to whether someone had engaged in a form of
political participation: 1) often, 2) sometimes, 3) rarely, 4) never. For reasons of distribution (all items are
heavily skewed toward the ‘never’ category, with only few respondents indicating the ‘often’ and ‘sometimes’
categories), the items were recoded into two categories. The first three categories were recoded into
‘participated’, while the fourth category became ‘did not participate’.
ii
The original questionnaire contained 10 categories of educational level that cover the complexities of the
Belgian educational system. For reasons of comparability and clarity, these categories were recoded. People
with academic or non-academic higher education were coded as ‘high’. People who finished secondary
education were coded as ‘middle’. Otherwise people were coded as ‘low’.
26