The Social Science Journal 51 (2014) 201211

Contents lists available at ScienceDirect

The Social Science Journal

journal homepage: www.elsevier.com/locate/soscij

What drives unconventional political participation?

A two level study
Daniel Stockemer
School of Political Studies, University of Ottawa, Canada

a r t i c l e

i n f o

Article history:
Received 16 October 2012
Received in revised form 21 October 2013
Accepted 21 October 2013
Available online 16 November 2013

Keywords:
Unconventional political behavior
Individual factors
Macro-level indicators

a b s t r a c t
There is little evidence on the factors that drive peaceful unconventional political participation. This study evaluates the impact of seven individual level constituents age, income,
education, gender, satisfaction with the government, engagement in civil society organizations and voting as well as ve macro-level factors economic development, democratic
experience, income inequalities, a countrys regime type and federalism on citizens participation in boycotts, demonstrations and petition signing activities. Participation in all
three protest activities hinges on education, voting, participation in civil society organizations, and lack of satisfaction with the government. Moreover, the inuence of some
macro-factors, such as democratic experience and economic development, and micro-level
factors, such as gender differs between the three forms of political engagement.
2013 Western Social Science Association. Published by Elsevier Inc. All rights reserved.

1. Introduction
Civil society provides an especially strong foundation
for a [political system] when it generates opportunities
for participation and inuence at all levels of governance. (Diamond, 1994, p. 9)
A participatory public is a dening feature of a
functioning political system (Tocqueville, 2000). Political
participation provides the link between the citizenry and
the governing process; it allows citizens to directly and
indirectly communicate their interests, concerns and complaints to the government (Tam & Rudolph, 2008). Aware
of the repercussions of citizens involvement in politics
for the well-being of a political system, numerous case,
comparative case, and large-N quantitative studies have
investigated several aspects of political participation (Van
der Meer & van Ingen, 2009). Most analyses focus on voting, the most conventional form of political engagement.
These studies have identied a plethora of macro-level

Tel.: +1 613 618 0425.

E-mail addresses: dstockem@uottawa.ca, d.stockemer@web.de

characteristics, such as compulsory voting, proportional

representation and the decisiveness of the election at the
heart of differences in turnout between countries (Dettrey
& Schwindt-Bayer, 2009; Franklin, 2004). The literature
has also identied an abundance of micro- or individuallevel attributes that impact citizens propensity to become
and remain both conventionally and unconventionally
engaged. Income, education, and participation in voluntary
organizations are commonly associated with higher levels of political engagement (Rosenstone & Hansen, 2002;
Verba, Schlozman, & Brady, 1995).
However, while recent research has disentangled
macro-level characteristics of turnout and micro-level
indicators for political participation more generally, few
studies examine the interplay between micro- and macrolevel factors on an individuals likelihood to participate
in peaceful forms of political protest. How does context inuence citizens propensity to engage in peaceful
unconventional political activities? Which individual and
contextual factors explain why some people sign petitions,
participate in boycotts, and demonstrate while others
do not? It is the goal of this article to answer these
puzzles. More precisely, this study analyzes the impact
of seven theoretically informed micro-level covariates:

0362-3319/$ see front matter 2013 Western Social Science Association. Published by Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.soscij.2013.10.012

202

D. Stockemer / The Social Science Journal 51 (2014) 201211

age, income, education, gender, satisfaction with the government, engagement in civil society organizations, and
voting as well as ve macro-level factors: economic
development, democratic experience, income inequalities,
federalism, and macro-level turnout on the three forms
of peaceful unconventional political actions: partaking in
boycotts, participating in demonstrations, and signing petitions. The results are derived from a large sample of more
than 41,000 individuals across 39 countries. By differentiating between the three most common forms of peaceful
political protest, this study also determines whether partaking in a boycott, participating in a demonstration, or
signing a petition can be explained by the same independent variables or not.
The results of this study are quite nuanced. First, an
individuals propensity to engage in peaceful protest activities is mainly driven by individual factors. Second, there
are some commonalities and differences with regard to
the constituents of the three peaceful protest activities. On
the one hand, the four micro-level indicators of education,
engagement in civic organizations, conventional political
participation, and dissatisfaction with the government are
individual drivers for high levels of all three forms of unconventional political engagement. On the other hand, there
are some differences in the effect of other micro-level factors, such as gender and income, as well as macro-level
indicators, such as economic development and democratic
government on the three forms of peaceful protest activities.
2. Dening unconventional political participation
Political participation, largely dened, comprises all
voluntary activities by individual citizens intended to
directly or indirectly inuence political choices at various
levels of the political system (Kaase & Marsh, 1979, p. 42).
One of the most common differentiations between different kinds of political participation is conventional versus
unconventional. Conventional political activities including
voting, party membership and participation in a political rally are actions that intend to inuence the political
process through the electoral arena (Verba & Nie, 1972).
Unconventional political participation aims to inuence
politics through non-institutionalized means. For example, holding a demonstration, participating in a boycott and
signing a petition are all forms of unconventional political
activities (Van der Meer & van Ingen, 2009).
Historically, unconventional political activities were
equated with all kinds of protests and comprised both
peaceful and violent forms of actions. In fact, prior to the
mid 20th century, revolutions, rebellions and revolts such
as the French and American Revolutions constituted a high
amount of contentious actions (Stockemer & Carbonetti,
2010). In the 20th and 21st centuries, deliberately resorting to violent forms of protest, such as wounding or killing
opponents or using weapons by protesters has become
less common, especially in Western countries (Roller &
Wessels, 1996). Rather, these types of protest have been
replaced by peaceful forms of political contention such
as participation in demonstrations, boycotts of products,
deliberate purchase of specic goods, and signing petitions.

Generally, these non-violent political acts are considered legitimate and within the bounds of accepted norms
(Meyer, 2004).
This study adopts the distinction between violent and
non-violent forms of political participation. Unconventional political activities are dened as all sorts of protest
behaviors that do not deliberately involve physical force
and operationalized by data from the World Value Survey. Consistent with the denition above, the concept
includes three measures of peaceful political protest: signing petitions, attending lawful demonstrations, and joining
boycotts. However, while these three forms are the most
common forms of peaceful protest activities, they differ in
at least three fundamental ways. First, individuals from the
sample of countries included in this analysis have varying
probabilities to partake in any of the three forms of political
action. For example, 31% of the sample respondents have
signed a petition; 18% have demonstrated; and 10% have
boycotted.
Second and relatedly, the three proxy variables of
protest activity involve different forms of protest. For
example, signing a petition is a very low risk form of
unconventional political activity; signing a petition takes
minutes, if not seconds, and the act remains largely invisible. In contrast, participation in a demonstration is a more
open act of political engagement that involves a relatively
high amount of time commitment. Third, the three forms
of political engagement might be used for different purposes. For instance, boycotts might be conducted for ethical
reasons, while demonstrations could be more overtly political. Because these three forms of contentious actions cover
different forms of unconventional political engagement,
each of these three forms of peaceful contentious actions
is treated individually in the analysis. Political crimes,
such as terrorism, sabotage, guerrilla warfare, hijacking,
assassinations, bombings, kidnappings, and armed attacks
are deliberately excluded from the denition of political
engagement.
Contrary to conventional political engagement, peaceful unconventional political involvement has been on the
rise over the past decades (Inglehart, 1999, p. 7). Table 1
presents the increase or decrease in all three forms of
unconventional participation in all 15 countries that participated in the 1981 or 1990 wave and the 2005 wave
of the World Value Survey. The rst number in each eld
presents the change in any of the three forms of political
engagement between 1981 or 1990 and 2005. The numbers in parentheses represent the percentage of citizens in
each country who self-identify as having ever participated
in any of these three forms of political engagement in 1981,
1990, and 2005 respectively. While all forms of unconventional political engagement have generally increased, there
is also some wide-ranging between-country variation. For
example, in Japan, all forms of protest behavior have more
than doubled between 1981 and 2005. However, in other
countries such as Turkey, unconventional engagement has
been stable over the past three decades. Finally, there are
even a few countries, including Chile, where protest actions
have actually declined.
For the remaining 60 countries that are included in any
of the ve waves of the World Value Survey, it is harder

D. Stockemer / The Social Science Journal 51 (2014) 201211

203

Table 1
Unconventional political engagement throughout time.

Argentina
Australia
Brazil
Chile
Finland
India
Japan
Mexico
Poland
Russia
South Africa
South Korea
Spain
Switzerland
Turkey
Average increase

Demonstrations

Petitions

Boycotts

+8.8 (22.631.4)
+7.9 (12.220.1)
.2 (18.518.3)
12.6 (29.516.9)
4.2 (14.510.3)
+4.1 (15.319.4)
+17.6 (8.726.3)
5.1 (8.93.8)
+.1 (10.110.2)
+14.7 (30.615.9)
+ 5.1 (7.712.8)
+ 4.3 (7.111.4)
+13.0 (22.935.9)
+12.9 (15.228.1)
+.6 (5.35.9)
4.5

6.5 (34.2 27.7)

+9.5 (69.779.2)
+5.9 (49.855.7)
5.1 (22.117)
+20.7 (29.850.5)
+6.6 (22.429)
+52.2 (8.759.9)
+7.5 (9.6 17.1)
+12.1 (11.523.5)
18.3 (26.58.3)
8.4 (20.111.7)
+14.1 (19.934)
+ 7.9 (15.523.4)
+16.1 (61.577.6)
.9 (12.811.9)
7.6

+.4 (2.63)
+10.5 (4.915.4)
1.8 (9.88)
1.2 (3.92.7)
7.5 (8.516)
.1 (15.215.1)
+3.9 (3.17.0)
+.7 (1.32.0)
+.2 (4.74.9)
1.2 (3.62.8)
+.6 (9.410)
+3.7 (2.36)
+1.5 (5.57)
+8.6 (10.619.2)
+.2 (5.25.4)
2.2

to establish similar trends. Some of them have only been

included once, such as Azerbaijan in 2005. Other states,
Italy and Germany for instance, are featured in two or, at
most, three adjacent waves (1990, 1995, and 2000 or 1995,
2000, and 2005) of the World Value Survey. With only a
maximum of 10 years of data, it is difcult to establish
strong trends, but the short term tendencies of all countries
which took part in three adjacent waves of the World Value
Survey, support Table 1. Most countries, such as Sweden or
Serbia, have seen an increase in unconventional political
participation.
Many scholars, including Zukin, Keeter, Andolina,
Jenkins, and Della Carpini (2006), argue that increases
in unconventional political activities compensate for
decreases in party and electoral politics. For example,
Ekiert and Kubik (2001) claim that contentious politics
offers citizens an opportunity to challenge and ameliorate
government policies. They not only allow people to directly
push their demands and demonstrate their grievances, they
also allow them to bring new topics to the agenda that
are often not treated by traditional politics; these new
themes include rights for minorities and lesbians and gays.
In addition, Tilly and Tarrow (2007) point out that doing
politics can be more satisfying than the typical elite-driven
behavior. In fact, peoples participation in social movements and collective action can create a healthy repertoire
of contention, a concept that has long been propagated by
the social movement literature (McAdam, Tarrow, & Tilly,
2001).
In fact, contentious actions have proven to be vital
for the advancement of many societies. In the 1960s and
1970s, great democratic reforms and societal transformations such as the US civil rights legislations would not
have materialized without the unrelenting dedication and
actions of committed individuals who fought desegregation and the Jim Crow laws through mainly peaceful means
(McAdam, 1982; Wei, 1993). In the 1980s, individual and
collective actions brought post-materialist topics such as
peace and the danger of nuclear power stations to the political agenda. More recently, actors in so-called new social
movements have transformed societies, rendering them
more tolerant toward groups such as gays and lesbians
(Alvarez & Escobar, 1992; De Pozo Alvino, 2006).

Given the importance of collective direct action, it is

central to comprehensively test the salience that microand macro-level indicators have on peaceful forms of political protest. This applies even more so because there is
so little research systematically testing the constituents
of various forms of unconventional political participation.
In fact, there appear to be only two studies examining the drivers of protest politics. By comparing western
and eastern democracies, the rst of these articles by
Bernhagen and Marsh (2007) nds that an individuals
likelihood to engage in protest activities is mainly driven
by socio-economic status and engagement in civil society organizations. The second article by Bellinger and Arce
(2011) is a macro-level study examining the effect of economic liberalization and a countrys level of democracy on
collective protest activity in 17 Latin American countries
from 1970 to 2003. The authors nd that more economic
and political freedoms trigger higher numbers of protests
in the aggregate.
The present study builds on both Bernhagen and Marsh
(2007) as well as Bellinger and Arce (2011) in several
ways. First, it is more comprehensive than both existing studies, looking not merely on either the micro-or
macro-level of unconventional political participation, but
examining how both individual factors and the macrocontext impact individuals propensity to engage in various
elite challenging activities. Second, it is also larger in
scope than previous research with the inclusion of 39
countries from various continents. Third, this research
examines whether the factors that explain one sort of
protest activity, such as signing petitions, are similar or
dissimilar to the indicators that account for another type
of unconventional political action, such as partaking in a
boycott.
The next section introduces the personal or microlevel constituents of peaceful political protest. Since recent
research (Van der Meer & van Ingen, 2009) strongly entertains the possibility that an individuals likelihood to
participate in the political process is not only inuenced
by personal characteristics but also by other factors. These
macro- or level two indicators are presented after the discussion of the micro-level constituents of unconventional
political participation.

204

D. Stockemer / The Social Science Journal 51 (2014) 201211

3. Micro-level constituents of peaceful protest

3.1. Income
Richer individuals will be more likely to engage in
unconventional forms of political actions. Not only do afuent individuals often have the political know-how and
expertise to participate in the political process, they also
have the necessary nancial resources to partake in actions.
In addition, richer citizens normally have various professional and social contacts, networks that can facilitate
contacts between activists and eventually participation in
protest politics (Barkan, 2004). Finally, and relatedly, nancial afuence is often positively correlated with increased
civic skills, which according to Burns, Schlozman, and
Verba (2001) are another stimulating factor for engagement in politics.
3.2. Education
Similar to nancial resources, education is frequently
cited as an important determinant for both increased conventional and unconventional political participation. The
literature (Comber, 2007; Flanagan, 2003) maintains that
a relatively high level of education is a prerequisite for
citizens to understand complex economic and political scenarios. In this sense, the capacity to think critically is likely
to be a facilitating link for individuals to engage in the
political process (Marien, Hooghe, & Quintelier, 2010). This
applies even more so, if we consider that parties, civil society organizations, and social movements all provide their
own expertise and educational material to convince other
activists and ordinary citizens of their propositions (Morse,
1993). Based on all these propositions, there is likely a positive link between higher education and stronger political
involvement.

studies (Burns et al., 2001) suggest that women should

have a particularly low tendency to participate in unconventional politics. According to Burns et al. (2001, p. 1),
the gap in political participation between men and women
increases if we go from voting to more radical forms of
political engagement such as boycotting or demonstrating.
Karp and Banducci (2008) further explain that this gap in
involvement can be explained by the tendency of women to
be less knowledgeable and passionate about politics than
men are. The rather solid evidence in favor of a participation gap for men renders it likely that men also outperform
women in the three forms of political engagement covered
by this study.
3.5. Voting
I assume that individuals that participate in the conventional political process through voting should have a higher
likelihood to engage in more unconventional actions than
citizens that are apolitical. For one, political engagement
often develops in steps. Individuals who are interested in
politics often start out voting. In this sense, it is likely
that individuals begin with conventional activities that
do not divert considerable resources and time. Participation in these common forms of involvement might further
politicize citizens by making them aware of the political
stakes or the failure of elected ofcials to respond to citizens demands. Simply voting on Election Day might not
be sufcient at some point for citizens who are politically informed and who want to change the environment
around them (Tilly & Tarrow, 2007). Zukin et al. (2006),
for example, argue that doing politics might become a satisfying endeavor for individuals who want to voice their
grievances and directly push their political convictions. The
empirical analysis will show whether conventional political engagement can trigger more unconventional forms of
political involvement.

3.3. Age
3.6. Satisfaction with the government
While there is strong support in the literature that voting increases monolithically with age, it is not clear if the
same relationship exists for age and unconventional political participation (Franklin, 2004; Pattie, Seyd, & Whitley,
2004). On the one hand, Putnam (2000) indirectly entertains the idea that younger generations are less likely to
participate unconventionally when he reports a decline in
civic awareness and social capital among the younger generations of a society. On the other hand, Marsh, OToole,
and Jones (2007) argue that young generations are political, but in unconventional rather than conventional ways.
According to this second perspective, younger generations
foster the transformation of norms and customs in society through non-electoral forms of political engagement
(Dalton, 2008). The empirical analysis will show if, at all,
which one of the two perspectives bears salience.
3.4. Gender
In the literature, there is some consensus that men are
more politically interested, informed and engaged than
women (Verba, Burns, & Schlozman, 1997). In particular,

It is likely that citizens who are satised with their government and politics in general, have no reason to engage
in more radical political activities. Rather, they can vote
to display their support. In contrast, disgruntled or dissatised citizens might feel that conventional actions are
inefcient and thus engage in more radical forms of politics. As Klandermans (1997) puts it, a demand for change
often begins with dissatisfaction, be it in the experience of
mediocre government performance, corruption in the state
apparatus, or feelings of injustice and moral indignation
about some state of affairs in politics or society.
3.7. Membership in civil society organizations
Finally, involvement in civil society organizations
should have a positive inuence on increased peaceful unconventional political engagement (Passy & Giugni,
2001, p. 143). For one, there should be a strong direct link
because unions, environmental or consumer organizations
frequently directly organize petition signing activities, boycotts, and demonstrations. On the other hand, there ought

D. Stockemer / The Social Science Journal 51 (2014) 201211

to also be an indirect inuence because networks among

activists are formed inside these civil society organizations
to help them connect to other protest structures (Cumber,
Routledge, & Nativel, 2008).
4. Macro-level constituents of voting
4.1. A countrys material wealth
Based on Ingleharts Value Change Theory (1981, 1990,
1997), I hypothesize that individuals, who live in a
developed country are more likely to engage in unconventional political activities than individuals in less afuent
countries. Inglehart (1981, p. 881) advances two hypotheses to push his claim. First, according to the scarcity
hypothesis people place the greatest value on those [goods]
that are in relatively short supply. That is, when some valued object is difcult to obtain, its worth is magnied. If
the supply increases to match the demand, the object is
taken for granted and attention shifts to other objects that
are still scarce (Abramson & Inglehart, 1987). The second
premise of Ingleharts value change approach is the socialization hypothesis, which states that to a large extent,
ones value priorities reect the conditions that prevailed
during ones pre-adult years (Inglehart, 1981, p. 881). In
this view, an individuals basic value priorities are initially
formed early in life in reaction to the conditions of this
formative period. These formative conditions include both
the social and political environment, as well as the individuals immediate family situation. Once developed, these
values tend to endure in the face of later changes in social
conditions.
Based on these two hypotheses, Inglehart (1990) argues
that economic development, accompanied by social and
cultural change, will cause an intergenerational shift from
material survival values to post-material quality of life
values. The rise of these post-material or post-modern values brings declining respect for authority and growing
emphasis on participation and self-expression (Inglehart
& Welzel, 2005). These trends ought to be conducive to
protest actions such as strikes, boycotts, or demonstrations.
4.2. Democracy
Second, peaceful protest politics should be more pronounced in a democracy. A democracy is a political system
that includes both political rights and civil liberties. According to Margolis (1979), it ought to promote individual
self-development and self-expression. In addition, the
concept of a healthy civil society encapsulates both conventional and unconventional forms of political participation.
Citizens can check those in power through both: participating in elections, meeting politicians, or engaging in political
campaigns; as well as boycotting, demonstrating, and collecting and signing petitions.
Under a democratic constitution, any segment of the
population has the right to form pressure groups and
social movement organizations, which can then facilitate
all sorts of protest behaviors. These forms of legitimate
collective actions can bring new topics to the political
agenda, place pressure on the government, and impact

205

public opinion; in short, direct actions are a way for citizens to ensure that their voices are heard; this applies
even more so if conventional channels of decision-making
are either blocked or non-existent (Molyneux, 2002). In
contrast, non-democracies do not promote the political
maturation of citizens. They ban civil society organizations
that do not conform to the governments agenda, and prohibit many or all forms of contentious politics. Based on
this rationale, I hypothesize that unconventional political
participation should be higher in democracies.
4.3. Income inequalities
The relative deprivation approach posits that the likelihood of protest increases with individuals feelings of
dissatisfaction with empirical realities (Seidman, 1994). A
persons level of frustration hinges by and large on how
that person sees the situation with regard to other individuals, either in the same country or in another state
or compared with the individuals own or collective past.
Income inequalities can be one major cause why people
feel deprived. Disparities in peoples disposable income
exacerbate material hardship for those at the bottom rung
of the income ladder; not only is their material situation
worsening over time, even more importantly, they are also
witnessing an increasing gap between what they have and
what others, who are at the top level of society, possess
(Choi, 1999). In particular, in situations where the regime
in power allies with the interests of those at the top rung of
society and does nothing to smooth existing inequalities,
popular anger might be considerable. In such a scenario,
citizens propensity to protest might be very high (Bellin,
2002). Given that income inequalities have been on the rise
in basically every country in the world in the past two or
three decades, I assume this factor to be directly related to
increased contentious actions.1
4.4. Federalism
A federation, a political system which divides power
between the central government and its subunits regions,
districts, municipalities has several advantages which
ought to be conducive to political engagement of all types.
First, federal arrangements can protect the preferences of
territorially-based groups, so that they are not subject to
decisions of the majority (Lijphart, 1999). Second, local
decision-makers have more information than centralized
bodies about the preferences and needs of their constituencies, which will most likely allow for better government
services (Olson, 1969). Third, federalism allows for local
variation in taxation, the provisions of goods and services,
and regulations variations that might be well-justied
based on different needs.
However, beyond protecting minority interests from
abuse by the majority and providing for an adequate

1
For example, Wade (2004) shows that between 1988 and 1993 alone
the ratio between the 10th richest percentile of the world population to
the 10th poorest percentile of the population has increased from 7.28 to
9.28.

206

D. Stockemer / The Social Science Journal 51 (2014) 201211

provision of goods and services, federations are also likely

to foster political participation (Levy, 2008). Furthermore,
because there are multiple channels of decision-making,
there should also be more deliberation between elected
ofcials and citizens. The fact that many decisions are also
made locally, under the direct supervision of the people
affected, should further foster a participatory political climate and increase both conventional and unconventional
political participation.
4.5. Turnout
Finally, I test whether macro-level conventional political participation trends inuence an individuals propensity to engage in peaceful unconventional actions. Previous
research (Burns et al., 2001) identies similar personal
characteristics among those involved in all kinds of political
activities and recognizes a large overlap between organizations involved in conventional and unconventional political
activities. Many former social movement organizations
from the far-right to the far-left have become social movement parties. The members of these groups, for example,
all Green Parties in Europe, participate in both the electoral arena, and in demonstrations and boycotts (Burstein
& Linton, 2002). Based on this rationale, there should be a
positive link between turnout and unconventional political
activities.
5. Data
The micro-level indicators, as well as the dependent variable, are retrieved from the World Values Survey
(2009). The three dependent variables gauge whether
respondents have ever participated in any of the three
protest activities: (1) signing petitions, (2) participating in
boycotts, or (3) partaking in demonstrations. The respective variable is coded 1, if the person has ever engaged in
any of these three actions and 0 otherwise. The micro-level
factors are operationalized as follows: rst, a scale from 1
to 10 measures a respondents income percentile in his or
her respective country. For example, somebody with the
value of 3 is in the 20th to 30th income percentile in his or
her respective country, and a person ranked 9 is situated in
the 80th to 90th income percentile. Second, on a scale from
0 to 9 0 stands for no formal education and 9 for a university education respondents educational level is recorded.
Third, age is operationalized by the actual age of the respondent at the time of the interview. Fourth, a dummy variable
for gender is included; men are coded 1; women are coded
0. The fth variable is also a dummy: individuals who selfidentify as having voted in the previous elections, are coded
1; citizens who did not vote, 0. Sixth, a four point scale
operationalizes the variable satisfaction with the government: a great deal is coded 1, and not at all is coded 4.
The nal micro-level indicator is an index that measures
individuals involvement in civil society organizations. The
index is an additive index of membership in the following 5 civil society organizations (unions, environmental
organizations, consumer organizations, professional organizations, and charitable organizations). For each of the

items, no membership is coded 0, passive membership 1,

and active membership 2.
The ve macro-level indicators are placed in the following coding scheme: the rst indicator, a countrys material
wealth, is determined by the states GDP per capita. Because
a linear relationship between higher macro-levels of material wealth and more individual protest activities cannot
be assumed, the variable GDP per capita is log transformed. The data were retrieved from the United Nations
(2011). The second indicator, democracy, is operationalized
by the number of years a country has been a democracy since the last transition.2 This indicator is a measure
of democratic experience. For one, it measures whether
democracies have more individuals that participate in
unconventional political activities than autocracies. Possibly even more importantly, it also gauges whether more
individuals engage in peaceful protest activities the more
a democracy consolidates (Schedler, 1998).
Third, income inequalities are operationalized by the
Gini coefcient. Gini scores range from 0 to 100. A score
of 0 stands for perfect income inequalities (all people in a
country have the same income) and a score of 100 signies that a single individual holds all wealth in a country.
Data for the Gini coefcient were collected from The Global
Peace Index Map of Gini data (2010). Federalism is operationalized by a dummy variable. All federations are coded
1, and all centralized states are coded 0. The data for this
variable come from the Forum of Federations (2009). The
nal variable, turnout, uses the standard denition and
measures the indicator by the percentage of eligible citizens who cast their ballot in their countrys most recent
legislative election.3 Data were collected from the International Institute for Democracy and Electoral Assistance
(IDEA, 2009).
6. Statistical procedures
Hierarchical Linear Modeling (HLM) is used to evaluate the inuence of the seven individual level covariates
and the ve macro-level indictors on the three dependent variables of unconventional political participation, (1)
participation in demonstrations, (2) boycotting, and (3)
signing a petition. One separate model for each dependent
variable measures whether the same or different factors
account for an individuals propensity to demonstrate,
partake in a boycott, or participate in a demonstration.
Several considerations justify the use of hierarchical linear modeling. First and most importantly, HLM accounts
for the clustering of individual data within its national
context (OConnell & McCoach, 2008). As such, this technique allows measuring how context inuences individual
behavior in this case, participation in protest activities.
Additionally, accounting for the two level structure of the
data prevents the overestimation of the statistical significance of the individual factors, thus providing a more
precise estimation of the level 1 indictors impact on the

2
According to the Polity IV guidelines a state is considered democratic
if it is ranked 6 or above on the Polity IV scale.
3
The data refer to 2009.

D. Stockemer / The Social Science Journal 51 (2014) 201211

207

Table 2
Descriptive statistics.
N

Mean

SD

Level 1 indicators
Signing petitions
Boycotting
Demonstrating
Income
Education
Age
Gender
Condence in the government
Voting
Membership in civil society organizations

41,890
41,890
41,890
41,890
41,890
41,890
41,890
41,890
41,890
41,890

.31
.10
.18
4.83
5.48
41.96
.49
2.57
.77
1.39

.46
.30
.38
2.31
2.39
16.45
.50
.90
.42
2.25

0
0
0
1
1
15
0
1
0
0

Level 2 indicators
Log GDP per capita
Years democracy
Gini coefcient
Macro-level turnout
Federalism

39
39
39
39
39

8.58
31.96
36.07
70.07
.28

1.61
45.1
9.05
15.51
.46

5.11
0
23.70
32.20
0

dependent variable. Because the dependent variables are

dummy variables, the three models are specied as binary
logistic HLMs comprised of more than 41,000 cases across
39 countries.4
In all models, there is wide variation in all three dependent variables. For example, the percentage of those citizens
who report that they have participated in demonstrations
ranges from below 5% for countries such as Thailand to
more than 30% in Spain. More pronounced is the betweencountry percentage differences of citizens who declare that
they have already signed a petition; it vacillates from less
than 10% for countries such as Ghana or Indonesia to over
70% or higher for Switzerland. The gap between countries
for participation in boycotts is less pronounced. It ranges
from less than 3% of all citizens in, among others, Moldova,
to more than 25% in Sweden. The mean engagement values
for the three forms of protest politics are: 18% for demonstrations, 31% for petitions, and 10% for boycotts. There is
similar variation in all independent variables (Table 2).
7. Results
The three HLM models trigger three distinct ndings
(Tables 35). First, the regression equations reveal four core
factors that are signicantly related to all three types of
unconventional political engagement. These four factors
include education, participation in civil society organizations, a lack of condence in the government, and
participation in elections. Second, there is variation in
some of the other micro-variables inuence on any of the
four protest activities. Third, the models indicate that the
inuence macro-level factors have on peaceful protest politics differs with the kind of activity. In short, macro-level
factors do not matter for demonstrations, are somewhat

4
The countries included in this analysis are: Italy, USA, Canada, Japan,
Mexico, South Africa, Australia, Norway, Sweden, Finland, South Korea,
Poland, Switzerland, Brazil, Chile, India, Slovenia, Bulgaria, Romania,
Turkey, Ukraine, Peru, Ghana, Moldova, Georgia, Thailand, Indonesia, Viet
Nam, Serbia, New Zealand, Morocco, Cyprus, Trinidad Tobacco, Andorra,
Malaysia, Burkina Faso, Mali, Zambia, and Germany.

Min

Max
1
1
1
10
9
94
1
4
1
12
11.88
196
67.76
99.64
1

important for boycotts, and carry noticeable weight for

explaining petition signing activities.
The results of Tables 35 indicate a high degree of consistency in the individual drivers of participation in protest
activity. Individuals are more likely to engage in any of
the three types of protest activities if they are highly educated, involved in civil society organizations, dissatised
with the government, and if they participate in the conventional realm through elections. Education has the strongest
bearing. For example, an average citizen in the average
sample country has a mere 6% chance to have ever signed
a petition if he has no formal education.5 The individuals
probability increases to 68% if he or she has a university
degree. Similar discrepancies between citizens of the lowest and the highest educational level occur for the other two
protest activities; the gap in boycotting and demonstrating is 5 percentage points for the former protest activity
and 35 percentage points for the latter form of unconventional activities. The variable membership in civil society
organizations has the second largest impact. For instance,
an individuals likelihood to sign a petition increases from
30% for a person who is not involved in any civil society
organizations to 54% for an individual actively engaged
in two organizations. For demonstrations, the gap is 20
percentage points (30% versus 50%) and for boycotts it
is 4 percentage points (6% versus 10%). The substantive
inuence of the other two common statistically signicant
variables a lack of condence in government and participation in elections is more moderate and does not
normally exceed 10 percentage points between individuals
situated at the extremes for the respective independent
variables.
Also, despite the fact that the coefcients for age are not
statistically signicant in any of the three models, age has
an inuence on all forms of political actions. The reason
for this effect lies in the measurement of the dependent
variables. The three dependent variables are based on the

5
For this probability transformation, as well as all other probability
transformations that ensue, all other variables are held constant at their
mean.

208

D. Stockemer / The Social Science Journal 51 (2014) 201211

Table 3
Result of the binary logistic HLM model featuring signing petitions as the dependent variable.
Fixed effects

SE

df

Model for an individuals mean likelihood of signing petitions (0 )

6.18
Intercept ( 00 )
.38
Log GDP per capita ( 01 )
.01
Years democracy ( 02 )
Gini coefcient ( 03 )
.02
National turnout ( 04 )
.004
Federalism ( 05 )
.48

Coefcient

(1.10)
(.092)
(.004)
(.013)
(.007)
(.36)

5.62
4.13
2.59
1.20
.52
1.34

(33)
(33)
(33)
(33)
(33)
(33)

<.001
<.001
.014
.240
.605
.189

Model for personal income slope (1 )

Intercept (y10 )

.039

(.011)

3.68

(41,844)

<.001

Model for personal education slope (2 )

Intercept (y20 )

.18

(.016)

11.06

(41,844)

<.001

Model for age slope (3 )

Intercept (y30 )

.0028

(.0027)

1.03

(41,844)

.303

Model for gender slope

Intercept (y40 )

.031

(.051)

(41,844)

.545

Model for voting slope

Intercept (y50 )

.606

.46

(.066)

6.97

(41,844)

<.001

Model for condence in government slope

Intercept (y60 )
.12

(.039)

3.68

(41,844)

<.001

Model for membership in civil society organization slope

.16
Intercept (y70 )

(.016)

9.74

(41,844)

<.001

questions: have you ever participated in a demonstration?

Have you ever participated in a boycott? Have you ever participated in a petition signing activity? By denition, older
people have a higher likelihood to have engaged in any
of these activities. Hence, we should see this positive age
effect in all three models if individuals at various ages participate at an equal rate. However, in none of the equations
is age a signicant indicator. The models show no age effect,
implying that younger individuals participate in the three
protest activities at a higher degree than older citizens.

While the inuence of education, voting, condence in

the government, membership in civil society organizations,
and age is consistent across the three models, both gender
and incomes inuence uctuates. The models indicate that
men are more likely to demonstrate and boycott, whereas
there is no gender difference in signing petitions. However,
in both the demonstration and the boycott models the indicators substantive inuence is relatively small: men are
5% more likely to demonstrate and 2% more likely to boycott. Higher personal income is also positively related to

Table 4
Result of the binary logistic HLM model featuring participating in boycotts as the dependent variable.
Fixed effects

Coefcient

SE

df

Model for an individuals mean likelihood of signing petitions (0 )

4.84
Intercept ( 00 )
.085
Log GDP per capita ( 01 )
.0073
Years democracy ( 02 )
.004
Gini coefcient ( 03 )
.011
National turnout ( 04 )
.14
Federalism ( 05 )

(1.20)
(.11)
(.0029)
(.016)
(.009)
(.28)

4.03
.76
2.46
.25
1.28
.49

(33)
(33)
(33)
(33)
(33)
(33)

<.001
.450
.019
.804
.211
.630

Model for personal income slope (1 )

Intercept (y10 )

.008

(.016)

.48

(41,844)

.632

Model for personal education slope (2 )

Intercept (y20 )

.18

(.016)

11.40

(41,844)

<.001

Model for age slope (3 )

Intercept (y30 )

.0001

(.002)

.05

(41,844)

.959

Model for gender slope

Intercept (y40 )

.23

(.007)

3.46

(41,844)

<.001

Model for voting slope

Intercept (y50 )

.46

(.09)

5.28

(41,844)

<.001

Model for condence in government slope

.17
Intercept (y60 )

(.054)

3.09

(41,844)

.002

Model for membership in civil society organization slope

.16
Intercept (y70 )

(.013)

12.72

(41,844)

<.001

D. Stockemer / The Social Science Journal 51 (2014) 201211

209

Table 5
Result of the binary logistic HLM model featuring partaking in demonstrations as the dependent variable.
Fixed effects

Coefcient

Model for an individuals mean likelihood of signing petitions (0 )

3.55
Intercept ( 00 )
.09
Log GDP per capita ( 01 )
.0009
Years democracy ( 02 )
.009
Gini coefcient ( 03 )
National turnout ( 04 )
.00009
Federalism ( 05 )
.096
Model for personal income slope (1 )
Intercept (y10 )
.02
Model for personal education slope (2 )
Intercept (y20 )
.18
Model for age slope (3 )
Intercept (y30 )
.0012
Model for gender slope
Intercept (y40 )
.27
Model for voting slope
Intercept (y50 )
.36
Model for condence in government slope
Intercept (y60 )
.13
Model for membership in civil society organization slope
Intercept (y70 )
.20

SE

df

(1.07)
(.12)
(.004)
(.015)
(.004)
(.40)

3.33
.77
.23
.588
.23
.24

(33)
(33)
(33)
(33)
(33)
(33)

.002
.448
.821
.561
.370
.813

(.013)

1.76

(41,844)

.079

(.015)

11.58

(41,844)

<.001

(.002)

.597

(41,844)

.551

(.038)

7.203

(41,844)

<.001

(.075)

4.74

(41,844)

<.001

(.044)

2.99

(41,844)

.003

(.011)

16.81

(41,844)

<.001

an individuals propensity to sign petitions.6 Model 3 predicts that somebody in the lowest income bracket is 9% less
likely to have ever signed a petition than someone in the
highest income bracket in his or her respective country.
However, personal income is unrelated to the two other
forms of peaceful protest activities.
Pertaining to the salience of macro-level indicators, the
three models provide a nuanced picture. First, none of the
macro-level factors impacts an individuals likelihood to
participate in a demonstration (Table 5). Hence, it seems
that context does not matter in an individuals decisions to
turn to the street for political reasons. Moreover, there is
only one contextual factor democratic experience that
positively inuences an individuals likelihood to partake
in boycotts (Table 4). The indicators inuence is rather
moderate. An average individuals likelihood to have ever
participated in a boycott increases from 6% in an autocracy to 10% in a country that has been democratic for
100 years. For the third form of unconventional political
engagement, an individuals likelihood to sign a petition,
macro-level indicators have a larger inuence. Petition
signing activities increase with the democratic experience
of a country, as well as its level of economic development
(Table 3). For example, an average individuals likelihood
to have ever signed a petition increases from 28% in a
non-democracy to 71% in a country with 100 years of democratic experience. Similarly, there is a more than 30 point
gap in citizens likelihood to sign a petition between poor
countries, such as Ghana with a GDP per capita of $1000,
and rich countries, such as Norway with a GDP per capita
of $50,000.

6
Running a correlation matrix between education and income reveals
that the two concepts are only weakly correlated (the Pearson Correlation
Coefcient is .335). Thus, the two concepts are rather distinct: the effects
of income on political participation are not captured by education.

8. Conclusion
The conclusions to draw from this study are threefold.
First, this research discovers that the individual factors
responsible for high protest activities are very stable for
all three forms of unconventional political participation,
especially education and to a somewhat lesser extent
membership in civil society organizations boosts citizens
likelihood to partake in demonstrations, boycotts, and
petition signing activities. The same applies, albeit less
pronounced, to dissatisfaction with the government and
conventional political participation in the form of voting.
Furthermore, younger individuals engage more strongly in
the three forms of protest politics than older individuals.
Second, and relatedly, the results indicate some differences
in the salience of other micro-level factors; that is, (1) men
are somewhat more likely than women to demonstrate and
partake in boycotts, but not so in signing petitions, and (2)
individual personal income only carries some weight in the
decision to sign petitions, but is unrelated to participation
in demonstrations and boycotts.
Third, this article brings to the fore some very interesting patterns pertaining to the importance of macro-level
indicators. I nd that economic development, as well as
democratic experience strongly inuences an individuals
decision to sign a petition. The latter of the two indicators,
democratic experience, also inuences citizens propensity
to participate in a boycott. However, none of these indicators impacts citizens likelihood to demonstrate. These
ndings allow me to draw some tentative conclusions. Long
lasting democratic experience and, to a lesser degree, economic modernization, strengthen citizens propensity to
participate in some, but not all forms of protest politics.
This also implies that petition signing becomes the more
widespread the longer a country is democratic and the
more highly it is developed.
The study at hand also allows for the formulation
of some predictions concerning the development of

210

D. Stockemer / The Social Science Journal 51 (2014) 201211

unconventional political activities in countries where contentious actions have not yet become the norm: further
democratization and to a lesser degree, economic development, should further weaken traditional values that are still
dominant in many African and Eastern European countries
such as Burkina Faso and Bulgaria, enticing populaces there
to display higher degrees of contentious action, especially
in the forms of petition signing activities. However, based
on the results here, only a small increase in the number of
demonstrations can be expected as countries democratize
and develop.
While this article provides a rst snapshot of the
salience of potentially relevant independent variables on
unconventional political participation, it could be further
rened. As more waves of the World Value Survey become
available, the study at hand could be replicated with more
countries and time periods. Future research could also try
to replicate this study with other data sources such as the
Afro- or Eurobarometer. In particular, the varying inuence of assorted macro-level indicators on an individuals
likelihood to protest should be further explored.
References
Abramson, P. R., & Inglehart, R. (1987). Generational replacement and the
future of post-materialist values. Journal of Politics, 49(2), 231241.
Alvarez, S., & Escobar, A. (1992). The Making of Social Movements in Latin
America: Identity, Strategy and Democracy. Boulder: Westview Press.
Barkan, S. E. (2004). Explaining public support for the environmental
movement: A civic voluntarism model. Social Science Quarterly, 85(4),
913937.
Bellin, E. (2002). Stalled Democracy: Capital. In Labor and the Paradox of
State-Sponsored Development. Ithaca: Cornell University Press.
Bellinger, P. T., Jr., & Arce, M. (2011). Protest and democracy in Latin Americas market era. Political Research Quarterly, 64(3), 688704.
Bernhagen, P., & Marsh, M. (2007). Voting and protesting: Explaining
citizens participation in new and old democracies. Democratization,
14(1), 4472.
Burns, N., Schlozman, L. K., & Verba, S. (2001). The private roots of public
action: Gender, equality and political participation. Cambridge: Harvard
University Press.
Burstein, P., & Linton, A. (2002). The impact of political parties, interest groups and social movement organizations on public policy:
Some recent evidence and theoretical concerns. Social Forces, 8(2),
381408.
Comber, M. K. (2007). Political participation and service learning: Civic
education as problem and solution. In D. M. Shea, & J. C. Green (Eds.),
Fountain of youth: Strategies and tactics for mobilizing Americas young
voters (pp. 115128). Lanham, MD: Rowman & Littleeld.
Choi, J. (1999). The Kwangju peoples uprising: Formation of the absolute
community. Korea Journal, 39(2), 238282.
Cumber, A., Routledge, P., & Nativel, P. (2008). The entangled geographies of global justice networks. Progress in Human Geography, 32(2),
183201.
Dalton, R. (2008). Citizens Politics: Public Opinion and Political Parties in
Advanced Industrial Democracies. Washington: CQ Press.
De Pozo Alvino, E. (2006). Social movements in a globalized world. In
S. Dasgupta, & R. Kiely (Eds.), Globalization and after (pp. 321348).
London: Sage Publications.
Dettrey, B. J., & Schwindt-Bayer, L. A. (2009). Voter turnout in presidential
democracies. Comparative Political Studies, 42(10), 13171338.
Diamond, L. J. (1994). Toward democratic consolidation. Journal of Democracy, 5(1), 417.
Ekiert, G., & Kubik, J. (2001). Rebellious civil society: Popular protest and
democratic consolidation in Poland 19891993. Ann-Arbor: Michigan
University Press.
Flanagan,
C.
A.
(2003).
Developmental
roots
of
political engagement. PS: Political Science & Politics, 36(2),
257261.
Forum of Federations: The Global Network of Federalism.
(2009). Federalism by country. http://www.forumfed.org/en/
federalism/by country/index.php

Franklin, M. N. (2004). Voter turnout and the dynamics of electoral competition in established democracies since 1945. Oxford, UK: Cambridge
University Press.
Inglehart, R. (1981). Post-materialism in an environment of insecurity.
American Political Science Review, 89(4), 12901304.
Inglehart, R. (1990). Culture shift in advanced industrialized societies. Princeton, NJ: Princeton University Press.
Inglehart, R. (1997). Modernization and postmodernization. Princeton:
Princeton University Press.
Inglehart, R. (1999). Postmodernization erodes respect for authority, but
increases support for democracy. In P. Norris (Ed.), Critical citizens:
Global support for democratic government (pp. 236256). New York:
Oxford University Press.
Inglehart, R., & Welzel, C. (2005). Modernization cultural change, and
democracy: The human development sequence. New York: Cambridge
University Press.
International Institute for Democracy and Electoral Assistance (IDEA).
(2009). Voter turnout. http://www.idea.int/vt/
Kaase, M., & Marsh, A. (1979). Political action: A theoretical perspective.
In S. H. Barnes (Ed.), Political action: Mass participation in ve western
democracies (pp. 2756). Beverly Hills: Sage Publication.
Karp, J., & Banducci, S. (2008). When politics is not just a mans game:
Womens representation and political engagement. Electoral Studies,
27(1), 105115.
Klandermans, B. (1997). The social psychology of protest. Blackwell:
Oxford/Cambridge, MA.
Levy, J. (2008). Self-determination, non-domination, and federalism.
Hypatia, 23(1), 6078.
Lijphart, A. (1999). Patterns of democracy: Government forms and performance in thirty-six countries. Yale University Press.
Marsh, D., OToole, T., & Jones, S. (2007). Young people and politics in the
UK: Apathy or alienation? New York: Palgrave Macmillan.
McAdam, D. (1982). Political process and the development of black insurgency 19301980. Chicago: University of Chicago Press.
McAdam, D., Tarrow, S., & Tilly, C. (2001). The dynamics of contention. New
York: Cambridge University Press.
Marien, S., Hooghe, M., & Quintelier, E. (2010). Inequalities
in non-institutionalized forms of political participation: A
multi-level analysis of 25 countries. Political Studies, 58(1),
187213.
Margolis, M. (1979). Viable democracy. New York: St. Martins Press.
Meyer, D. S. (2004). Protest and political opportunities. Annual Review of
Political Science, 30(1), 125145.
Molyneux, M. (2002). Gender and the silence of social capital:
Lessons from Latin America. Development and Change, 33(2),
167188.
Morse, S. W. (1993). The practice of citizenship: Learn by doing. The Social
Studies, 84(4), 164168.
OConnell, A. A., & McCoach, B. D. (2008). Multilevel modeling educational
data. Charlotte, NC: Information Age Publishing.
Olson, M. (1969). Strategic theory and its applications: The principle of scal equivalence: The division of responsibility among different levels
of government. American Economic Review, 59(2), 479532.
Passy, F., & Giugni, M. (2001). Social networks and individual perceptions:
Explaining differential participation in social movements. Sociological
Forum, 16(1), 123153.
Pattie, C., Seyd, P., & Whitley, P. (2004). Citizenship in Brita values participation democracy. Cambridge: Cambridge University Press.
Polity IV. (2009). Polity IV annual time series 1800 to 2008.
http://www.systemicpeace.org/inscr/inscr.htm.
Accessed
on
12.23.09
Putnam, R. (2000). Bowling alone: The collapse and revival of American
community. New York: Simon & Schuster.
Roller, E., & Wessels, B. (1996). Contexts of political protest in western
democracies: Organization and modernity. In F. D. Weil (Ed.), Extremism, protest, social movements, and democracy research on democracy
and society (pp. 91134). Greenwich, CT: JAI Press.
Rosenstone, S., & Hansen, J. N. (2002). Mobilization, participation and
democracy in America. New York: Longman.
Schedler, A. (1998). What is democratic consolidation? Journal of Democracy, 9(1), 91107.
Seidman, Gay. W. (1994). Manufacturing Militance: Workers Movements in
Brazil and South Africa. Berkeley: University of California Press.
Stockemer, D., & Carbonetti, B. (2010). Why do richer democracies survive? The non-effect of unconventional political participation. The
Social Science Journal, 47(2), 237251.
Tam, C. W. K., & Rudolph, T. J. (2008). Emanating political participation:
Untangling the social structure behind participation. British Journal of
Political Science, 38(2), 273289.

D. Stockemer / The Social Science Journal 51 (2014) 201211

The Global Peace Index Map of Gini data (2010). http://www.
visionofhumanity.org/gpi-data/#/2010/GINI
Tilly, C., & Tarrow, S. (2007). Contentious politics. London: Paradigm Publishers.
Tocqueville, A. de. (2000 [18351840]). Democracy in America. Chicago,
IL: University of Chicago Press.
United Nations. (2011). Per capita GDP at current prices US dollars.
http://data.un.org/Data.aspx?q=GDP+per+capita&d=SNAAMA&f=gr
ID%3a101%3bcurrID%3aUSD%3bpcFlag%3a1
Van der Meer, T., & van Ingen, E. (2009). Schools of democracy? Disentangling the relationship between civic participation and political action
in 17 European countries. European Journal of Political Research, 48(2),
281308.
Verba, S., & Nie, N. H. (1972). Participation in America: Political democracy
and social equality. New York: Harper and Row.

211

Verba, S., Schlozman, K. L., & Brady, H. (1995). Voice and equality: Civic
voluntarism in American politics. Cambridge: University of Harvard
Press.
Verba, S., Burns, N., & Schlozman, K. L. (1997). Knowing and caring about
politics: Gender and political engagement. The Journal of Politics, 59(4),
10511072.
World
Values
Survey
Online
data
analysis
(2009).
http://www.worldvaluessurvey.org/
Wade, R. H. (2004). Is globalization reducing poverty and income inequality? World Development, 32(4), 567589.
Wei, W. (1993). The Asian American Movement. Philadelphia: University of
Temple Press.
Zukin, C., Keeter, S., Andolina, M., Jenkins, K., & Della Carpini, M. (2006).
A new engagement: Political participation, civic Life and the changing
American citizens. New York: Oxford University Press.