I need a 6 page research paper

greta89

HostetterMidtermPaper.pdf

Home >Political Science homework help >I need a 6 page research paper

Risk and Resistance: Risk Acceptance and Protesting Behavior in Democratic and Non-Democratic Countries

Abstract

Kam’s (2012) theoretical framework argues that risk-accepting individuals participate in politics

because they enjoy exciting and novel activities. Given that nondemocracies are more repressive

than democracies, how might individuals’ acceptance of risk and system of government influence

the decision to protest? Using data from the 2005-2014 World Values Survey, I find that highly risk-

accepting individuals in democratic countries are much more likely to report a willingness to

participate in future political boycotts than their less risk-accepting counterparts. Substantively, the

results indicate that highly risk-accepting individuals are 52% and 41% more likely to boycott in

median democratic countries compared to other members of society depending on whether one

uses Freedom House or Polity IV scores. Further, I find no evidence that risk acceptance influences

demonstrating or petitioning. Low risk-accepting individuals are more hesitant in their willingness to

risk life and limb by challenging the status quo in democratic and non-democratic countries.

1 Introduction

Why are some citizens willing to protest their government while others passively turn a blind eye

and abstain? Scholars offer three schools of thought regarding individual protest behavior. The first,

disaffected radicalism, argues that protesters are unsatisfied with or alienated from traditional

representative channels (Gurr 1970). Second, strategic resource scholars suggest that protests are a

function of civic expression rather than disaffection with the political process (Inglehart 1977; 1997).

A third school argues the decision to protest is dependent upon the context of the political

environment rather than any generalizable motivation.

Scholars dedicate a plethora of resources to better understand why individuals participate in

politics generally, and protests specifically, because political participation increases democratic

satisfaction (Anderson et al. 2005; Blais & Gélineau 2007) and political equality (Rosenstone &

Hansen 1993). Further, protests are often successful, lead to political change, and allow citizens to

express their grievances and policy preferences to political elites (Celestino & Gleditsch 2013;

Hooghe & Marien 2014; Stephan & Chenoweth 2008). By considering additional explanations of

protest behavior, scholars can better understand how and why some governments are more

responsive to citizen preferences than others.

Psychology scholars offer valuable insight into this debate by considering individuals’ risk

acceptance, defined as the extent to which individuals seek out risky behaviors and uncertain

outcomes (Ehrlich & Maestas 2010; Weber, Blais, & Betz 2002). Risk-accepting individuals are

generally comfortable with uncertainty (Ehrlich & Maestas 2010; Levy 2003) and they are more likely

to pursue exciting activities compared to other members of society (Kam 2012; Zuckerman 1979;

2007). An important missing link in the literature is how risk acceptance influences protesting

behavior in different political systems.1

I use the 2005 to 2014 World Values Survey to test whether and how risk acceptance

influences the propensity to protest in democratic and non-democratic countries. I consider two

important research questions: (1) do the risk accepting protest at different rates in democratic and

non-democratic countries? (2) does increasing political freedoms motivate or repel participation for

the risk accepting? I measure democracy and nondemocracy using the Polity IV and Freedom House

scores. The findings indicate that risk acceptance is a robust predictor of protest activity, but the

substantive findings are limited to only boycotting. Highly risk-accepting individuals are not only

much more likely to participate in boycotts than low risk-accepting individuals, they are also more

likely to report a willingness to boycott in countries with more freedom. On the other hand, the

findings also suggest that the risk accepting do not provide any more internal pressure on non-

democratic regimes than low risk-accepting individuals. Nondemocracies, therefore may continue to

harbor undemocratic principles or, at a minimum, quell any such rebellion beforehand with impunity.

The willingness of the risk accepting to participate in protests is an important theoretical and

practical implication that furthers scholarly understanding of the causal mechanisms associated with

individual protesting behavior. Supplementary analyses lend support to the following proposed

causal mechanisms: risk-accepting individuals participate in politics because they find it novel and

exciting (Kam 2012), democracies have higher rates of political participation compared to

nondemocracies (Dalton, van Sickle, and Weldon 2010), and democracies respond to protestors in

1 By protest behavior I mean demonstrating, boycotting, and petitioning. By political systems I mean democracies and nondemocracies. I use political systems and regimes interchangeably throughout the remainder of the paper.

systematically different ways compared to non-democratic regimes (Davenport 1999; Poe et al.

1999).

2 Risk Acceptance and General Protest Participation

Political scholars find strong evidence that risk acceptance influences American political participation

(Kam 2012) and candidate vote choice (Kam & Simas 2012). However, little work has considered the

role of risk acceptance at the international level. To develop a broader theory of protesting behavior

between countries, we must first determine whether the risk accepting are generally more likely to

protest in countries outside of the United States. The causal mechanisms for these studies are three-

fold; risk-accepting individuals participate in politics because they find it exciting, novel, and an

opportunity to challenge the political status quo (Ehrlich & Maestas 2010; Kam 2012; Kam & Simas

2012; Levy 2003). While countries vary in their political institutions and citizens vary in their degree

of participation around the world (Dalton, van Sickle, & Weldon 2010), it does not strain credulity to

assume that the causal mechanisms of risk acceptance and participation in the United States are any

different outside of the United States. Indeed, risk-accepting individuals are generally comfortable

with uncertainty (Ehrlich & Maestas 2010; Levy 2003) and may be more likely to challenge the status

quo and accept losses (Kahneman & Tversky 1979) that may arise from protesting.

Following the work of Kam (2012), I argue the risk accepting generally participate in politics

because they find it exciting, novel, and an opportunity to challenge the political status quo. While I

argue the causal mechanisms of risk acceptance and protesting are the same across countries, the

institutional and political differences between countries introduces several important theoretical

caveats beyond what can be studied in the United States. One such caveat is the role of political

freedoms and liberties.

3 Risk Acceptance, Political Freedom, and Protesting Behavior

An interesting paradox in the literature finds that those who are the most satisfied with democracy

are also the most likely to engage in political protests against the government (Dalton, van Sickle,

and Weldon 2010; Passini and Morselli 2011; Przeworski et al. 2000; Vanhuysse 2006). Explanations

for this paradox include growing economic conditions and positive life satisfaction that allow

democratic citizens more resources to protest their government (Dalton, van Sickle, and Weldon

2010), civic duty minded citizens that are willing to challenge authority (Passini and Morselli 2011),

and a desire for social change (Moscovici, 1976; Nemeth, 2003). If individuals are more likely to

protest in democracies compared to nondemocracies, then the influence of risk acceptance may

further enhance the democratic protesting paradox. In other words, the risk accepting may be even

more likely to protest in democracies compared to their less risk-accepting counterparts. On the

other hand, while the risk accepting may participate in politics at higher rates than other members

of society (Kam 2012), the repressive policies in nondemocracies may discourage participation for

high and low risk-accepting individuals.

4 Government Institutions and Political Freedom

Protestors achieve their goals by making threats against the government and disrupting the political

environment (DeNardo 1985; Piven & Cloward 1977). Governments have four potential responses

to protests. They can: (1) concede to the demands of challengers without repression, (2) repress the

challengers without concessions, (3) tolerate the challenge without concession or repression, or (4)

repress the challenges but grant concessions (Franklin 2009). Protesters win concessions when

governments concede to their demands, while repressions are the negative consequences

governments enact in response to protests. I focus on the theoretical implications of concessions

and repressions on the decision to protest.

I identify three important factors why we may observe different protesting behavior from the

risk accepting in democratic and non-democratic countries. First, government responses to

protesting are systematically different between democracies and nondemocracies. Scholars find that

higher levels of democracy lead to lower levels of political repression and higher levels of

government concession (Davenport 1995; Henderson 1991). In other words, democracies inherently

respond to the wishes of its citizens. In democracies, the risk accepting (1) may overestimate the

chance of concessions or (2) have greater confidence that the government will not respond with

repressions. Conversely, non-democratic countries are more likely to “subjugate citizens to prevent

any kind of threat” (Regan & Henderson 2002 p. 3). Nondemocracies provide few rights and are less

beholden to institutional constraints which, in turn, enables their leaders to repress impending

threats (Davenport 1999; Poe et al. 1999). In nondemocracies, the risk accepting (1) may

underestimate the chance of repressions or (2) overestimate the chance of concession. Second,

elected political elites are constrained by democratic principles that unelected leaders need not

follow. Indeed, democratic voters may punish elected leaders in future elections (Gartner & Regan

1996). This suggests greater uncertainty when protesting in nondemocracies, as non-democratic

leaders are less likely to grant concessions to their citizens and more likely to repress those who

engage in protests (Davenport 1995; Henderson 1991). Protests are far more likely to take place in

democracies than nondemocracies (Przeworski et al. 2000), in part because democracies “allow their

citizens greater freedoms of expression” (Vanhuysse 2006 p. 1). Finally, democratic citizens possess

numerous avenues to influence their government that non-democratic citizens do not have, such as

voting, petitioning, attending rallies and meetings, and writing letters. These additional avenues

minimize the use of violent methods by citizens against the government (Regan & Henderson 2002).

These additional avenues are further magnified in electoral authoritarian regimes. While electoral

authoritarian regimes conduct elections, they often limit the number of parties, use voter

intimidation tactics, and may go to the extreme of manipulating election results (Schedler 2013). In

these electoral authoritarian regimes, elections are symbolic.

Differences in granting government concessions, use of government constraint in democratic

systems, and manipulation of electoral institutions by authoritarian non-democratic regimes

suggests citizens may approach political participation in general, and protesting specifically, in

systematically varying ways depending upon the country in which they live. On the one hand, the

risk accepting, compared to other members of society, may be more likely to protest in

nondemocracies because it provides an opportunity to challenge the status quo. On the other hand,

the lack of government concessions in nondemocracies may deter participation because the risk

accepting may view protesting as a futile endeavor. However, theory would suggest that their higher

willingness to challenge the status quo and accept losses, and their comfortability with uncertainty

may lead the risk accepting to protest at higher rates than other members of society. While I make

no directional hypothesis concerning participation, we should expect varying relationships between

risk acceptance and protesting behavior depending on country freedoms and liberties. This leads to

the following hypothesis: Political freedoms influence protesting activity for the risk accepting.

5 Data and Method of Analysis

I test the hypothesis using 2005-2014 data from the World Values Survey (WVS), which has a large

sample size (N= 91,204) over multiple years. The WVS contains a variety of political systems,

geographic regions, and control variables that scholars find influence protest behavior (Blais 2006;

Kam 2012; Lijphart 2012; Norris, Walgrave, & Van Aelst 2005). More importantly for the present

analysis, the WVS provides a multiple item index of risk behavior.

The WVS asks respondents “Have you ever participated in a peaceful and lawful

demonstration?”, “Have you ever participated in a political boycott?”, and “Have you ever signed a

petition?”2 Although I make no hypotheses differentiating between boycotts, demonstrations, or

signing petitions, the literature indicates substantive differences between the three activities (Norris,

Walgrave, & Van Aelst 2005). We should expect different substantive participation rates for these

protest activities, and the inclusion of Risk Acceptance may further exaggerate the differences.

Demonstrate, Boycott, and Petition are three-point scales coded 0 if respondents said they would

“never do this,” 1 if they “might do this,” and 2 if they “have done this.” Because the dependent

variables are categorical, I use a two-level multinomial logit generalized structural equation model

set to countries as the first stage and individuals as the second to test the hypothesis.

I measure risk acceptance using a similar approach to that used previously by other scholars

(Berinsky & Lewis 2007; Freese 2004; Hoyle et al. 2002; Kam 2012; Miller 2000). Kam (2012) uses the

following statements in her risk scale: “I like new and exciting experiences, even if I have to break

the rules” and “In general, how easy or difficult is it for you to accept taking risks?” Similarly, the

2 Respondents can choose from “I would never do this” (boycott= 63.67%, demonstrate= 48.81%, petition= 43.55%), “I might do this” (boycott= 28.58%, demonstrate= 35.71%, petition= 29.32%), or “I have done this” (boycott= 7.75%, demonstrate= 15.48%, petition= 27.13%).

WVS includes an 8-item index about risk-accepting, risk-seeking, and risk-tolerant attitudes.3 For

example, the WVS asks respondents whether “Adventure and taking risks are important to this

person; to have an exciting life” and “It is important to this person to have a good time; to “spoil”

oneself.” These questions capture favorability towards risk similar to Kam’s measure. Responses

range from “very much like me” to “not at all like me.” I code Risk Acceptance as a 41-point scale

where 0 is “not at all like me” and 40 is “very much like me.” Higher values represent a greater

acceptance of risk (m= 18, s.d.= 5) I correlate Risk Acceptance with Age (-0.31, p< .001), Sex (-0.13,

p< .001), and Education (0.11, p< .001) as an initial test of the validity and reliability of the current

measure. The correlations are statistically significant and in the expected direction. The older and

female populations are less likely to identify as risk accepting, while the higher educated are more

likely (Weber, Blais, & Betz 2002).

I use the Freedom House Index and Polity IV scores to code country freedoms. I include an

interaction between Risk Acceptance x Polity in the models. While Polity is a common measure of

institutional freedoms often used in the literature (Kapstein & Converse 2008), I also include an

interaction term between Freedom House x Risk Acceptance, which includes more detailed measures

of civil rights and liberties.4 Table A-2 in the Appendix includes a detailed listing of each country used

in the analysis.5

3 I use risk acceptance, risk seeking, and risk tolerance interchangeably. 4 See John Högström’s (2013) analysis about the differences between Freedom House and Polity IV scores. 5 I exclude Armenia, Burkina Faso, Libya, Pakistan, Singapore, and Tunisia because of inconclusive or contradictory measures between POLITY IV and Freedom House. Albania, Bangladesh, Bosnia, Czech Republic, Croatia, Dominican Republic, El Salvador, Guatemala, Israel, Italy, Latvia, Lithuania, Macedonia, Puerto Rico, Saudi Arabia, Slovakia, and Tanzania are excluded because there are no risk scores for these countries. Andorra, Hong Kong, Palestine, Serbia, Uganda, and Venezuela are

I also include theoretically important control variables in the models. These factors are social

structures, motivational attitudes, government support, political behaviors, and economic

satisfaction. The protest literature often finds that Education, Employment, Political Interest, Social

Trust, Party Member, and Labor Member have a positive effect on protesting, and that Age, Sex,

Religiosity, Ideology, and Government Confidence have a negative effect (Norris, Walgrave, & Van

Aelst 2005). Recent literature also finds a strong positive correlation between economic happiness

and protest behavior (Dalton, van Sickle, and Weldon 2010; Quaranta 2016). A detailed coding of all

control variables is available in the Appendix.

6 Findings

I use a two-level multinomial logit generalized structural equation model set to countries in the first

stage and individuals in the second with “never do” as the baseline (comparison) group in the

analyses that follow. I begin the analysis by considering the direct role of Risk Acceptance on Might

Demonstrate (b= 0.0013, p= .50; b= -0.0091, p= .48) in Table 1, Might Boycott (b= 0.0021, p< .001;

b= 0.0209, p< .001) in Table 2, and Might Petition (b= -0.0040, p< .10; b= -0.0040, p< .10) in Table 3

without interaction terms. Surprisingly, Risk Acceptance has little positive or statistically significant

direct effect on demonstrating and petitioning. Risk Acceptance does however, have a direct positive

and statistically significant effect on Might Boycott. This is somewhat inconsistent with previous

findings in the American literature that finds a direct relationship between high risk acceptance and

political participation (Kam 2012). One explanation for this finding may be that previous studies have

excluded because there is not a POLITY IV and/or Freedom House score. Belarus, China, Taiwan, Finland, Iran, Jordan, Kuwait, Qatar, and Uzbekistan are excluded due to a lack of observations.

not looked at the role of risk acceptance on more “costly” forms of political participation such as

protesting, or that variations in country-level risk acceptance behaviors may result in null direct

relationships.

I include an interaction term between risk acceptance, Freedom House, and Polity IV scores

on “Might Demonstrate” in the second and fourth columns of Table 1. The findings indicate a positive

and statistically significant effect for Risk Acceptance x Freedom House (b= 0.0012, p< .05) and Risk

Acceptance x Polity IV (b= 0.0010, p< .05) on “Might Demonstrate”, controlling for all other factors

in the model. The coefficient for Risk Acceptance is the effect of risk acceptance on Might

Demonstrate for nondemocracies (i.e., when Freedom House or Polity IV are equal to 0), while the

interaction tells us the differences in the effects of risk acceptance for non-democratic and

democratic countries.

Table 1 Two-Level Generalized Structural Equation Model Multinomial Logit of the Effect of Risk Acceptance on “Might Demonstrate” by System of Government 2005-2014 ________________________________________________________________________________

Freedom House Polity IV ________________________________________________________________________________ Risk Acceptance Risk Acceptance x Freedom House ---------- 0.0012* ---------- ---------- ---------- (0.0006) ---------- ---------- Risk Acceptance x Polity IV ---------- ---------- ---------- 0.0010* ---------- ---------- ---------- (0.0005) Freedom House 0.2645*** 0.2433*** ---------- ---------- (0.0308) (0.0325) ---------- ---------- Polity IV ---------- ---------- 0.1070*** 0.0879*** ---------- ---------- (0.0138) (0.0163) Risk Acceptance 0.0013 -0.0091 0.0014 -0.0158# (0.0020) (0.0060) (0.0020) (0.0083) Social Structure Age -0.0139*** -0.0138*** -0.0140*** -0.0139*** (0.0006) (0.0006) (0.0006) (0.0006) Sex -0.1034*** -0.1027*** -0.1029*** -0.1023*** (0.0179) (0.0180) (0.0179) (0.0180) Education 0.0804*** 0.0804*** 0.0803*** 0.0803*** (0.0046) (0.0046) (0.0046) (0.0185) Employment 0.1025*** 0.1014*** 0.1025*** 0.1018*** (0.0185) (0.0185) (0.0185) (0.0185) Religiosity 0.0087* 0.0093* 0.0087* 0.0093* (0.0040) (0.0040) (0.0040) (0.0040) Motivational Attitudes Ideology -0.0376*** -0.0375*** -0.0377*** -0.0377*** (0.0040) (0.0040) (0.0040) (0.0040) Political Interest 0.3493*** 0.3494*** 0.3487*** 0.3489*** (0.0102) (0.0102) (0.0102) (0.0102) Systems Support Government Confidence 0.0108 0.0105 0.0110 0.0105 (0.0105) (0.0105) (0.0105) (0.0105) Social Trust 0.0869*** 0.0866*** 0.0855*** 0.0851*** (0.0217) (0.0217) (0.0217) (0.0217) Political Behavior Party Member 0.1444*** 0.1449*** 0.1464*** 0.1463*** (0.0404) (0.0404) (0.0404) (0.0404) Labor Member 0.2197*** 0.2204*** 0.2209*** 0.2217*** (0.0399) (0.0399) (0.0399) (0.0399) Table Continued

________________________________________________________________________________

Freedom House Polity IV ________________________________________________________________________________ Economic Satisfaction Life Satisfaction -0.0446** -0.0145** -0.0141** -0.0141** (0.0046) (0.0046) (0.0046) (0.0046) Life Choice 0.0016 0.0017 0.0012 0.0014 (0.0040) (0.0040) (0.0040) (0.0041) GDP 8.77e-06 8.63e-06 7.42e-06 7.30e-06 (5.57e-06) (5.57e-06) (5.59e-06) (5.59e-06) Constant -1.1476*** -0.9528*** -1.6286*** -1.3061*** N 75,455 75,455 75,455 75,455

______________________________________________________________________________ Notes: Table entry is the multinomial regression coefficient set to the country and individual level of analysis. Dependent variable is scaled as 0 (Never Do); 1 (Might Do); 2 (Have Done) with baseline category “Never Do.” ***p<.001; **p<.01; *p<.05; #p<.10, two-tailed.

I graph the results from Table 1 in Figure 1 and Figure 2. Figure 1 presents the marginal effects

of Risk Acceptance and Freedom House on “Might Demonstrate.” We can clearly see that while

moving from the lowest freedom score (2) to the highest freedom score (12) increases the probability

of participation by a substantively and statistically significant margin, the differences between low

risk-accepting and high risk-accepting individuals are statistically insignificant. The risk accepting in

countries with the highest freedom score (12) report a willingness to demonstrate at ~40% compared

to only ~20% in countries with the lowest freedom score (2). In other words, the out-group effects

of risk acceptance are much stronger than the in-group effects. The low risk accepting are 19% more

likely to report a willingness that they “Might Demonstrate” compared to the most risk accepting in

countries with the highest freedom score (49.1% to 39.8%), although these results are statistically

insignificant. Figure 2 paints a similar picture. Figure 2 presents the marginal effects of Risk

Acceptance and Polity IV on “Might Demonstrate.” Once again, moving from the lowest freedom

score (2) to the highest freedom score (12) increases the probability of participation by a

substantively and statistically significant margin (16.6% to 32%), however the differences between

low risk-accepting and high risk-accepting individuals are statistically insignificant. According to the

Polity IV scores, the low risk accepting are 15% more likely to report a willingness that they “Might

Demonstrate” compared to the most risk accepting in countries with the highest freedom score

(41.7% to 35.4%), although these results are statistically insignificant. These results indicate that

while there is a statistically significant effect between risk acceptance and system of government,

there is little substantive difference between high and low risk acceptance scores.

Figure 1

Next, I include an interaction term between risk acceptance, Freedom House, and Polity IV

scores on “Might Boycott” in the second and fourth columns of Table 2. The findings indicate a

positive and statistically significant effect for Risk Acceptance x Freedom House (b= 0.0014, p< .05)

and Risk Acceptance x Polity IV (b= 0.0011, p< .05) on “Might Boycott”, controlling for all other factors

in the model. Again, the coefficient for Risk Acceptance is the effect of risk acceptance on Might

Boycott for nondemocracies (i.e., when Freedom House or Polity IV are equal to 0), while the

interaction tells us the differences in the effects of risk acceptance for non-democratic and

democratic countries.

Figure 2

I graph the results from Table 2 in Figure 3 and Figure 4. Figure 3 presents the marginal effects

of Risk Acceptance x Freedom House on “Might Boycott.” Moving from the lowest freedom score (2)

to the highest freedom score (12) increases the probability of participation by a substantively and

statistically significant margin. Further, the difference between participation rates for low and high

risk-accepting individuals is substantively and statistically significant (p< .05). The most risk accepting

are 52% more likely to report a willingness that they “Might Boycott” compared to the least risk

accepting in countries with a median (7) freedom score (20.6% to 31.4%). In countries with the lowest

freedom score (2), the most risk accepting are 41% more likely to report a willingness that they

“Might Boycott” compared to the least risk accepting, but the difference is statistically insignificant

(12.6% to 17.8%). In countries with the highest freedom score (12), the most risk accepting are 16%

more likely to report a willingness that they “Might Boycott” compared to the least risk accepting,

but this difference is also statistically insignificant (29% to 33.8%).

Table 2 Two-Level Generalized Structural Equation Model Multinomial Logit of the Effect of Risk Acceptance on “Might Boycott” by System of Government 2005-2014 ________________________________________________________________________________

________________________________________________________________________________

Freedom House Polity IV ________________________________________________________________________________ Motivational Attitudes Ideology -0.0400*** -0.0400*** -0.0400*** -0.0401*** (0.0041) (0.0041) (0.0041) (0.0041) Political Interest 0.3689*** 0.3691*** 0.3686*** 0.3687*** (0.0106) (0.0106) (0.0106) (0.0106) Systems Support Government Confidence -0.0561*** -0.0564*** -0.0558*** -0.0563*** (0.0110) (0.0110) (0.0110) (0.0110) Social Trust 0.1725*** 0.1725*** 0.1712*** 0.1711*** (0.0220) (0.0220) (0.0220) (0.0220) Political Behavior Party Member 0.0234 0.0231 0.0244 0.0239 (0.0393) (0.0393) (0.0393) (0.0393) Labor Member 0.1627*** 0.1635*** 0.1631*** 0.1640*** (0.0386) (0.0386) (0.0386) (0.0386) Economic Satisfaction Life Satisfaction -0.0187*** -0.0187*** -0.0185*** -0.0185*** (0.0048) (0.0048) (0.0048) (0.0048) Life Choice -0.0042 -0.0041 -0.0044 -0.0043 (0.0043) (0.0043) (0.0043) (0.0043) GDP -0.0001** -0.0001** -0.0001** -0.0001** (5.50e-06) (5.50e-06) (5.52e-06) (5.52e-06) Constant -1.2371*** -0.9993*** -1.4815*** -1.1295*** N 75,455 75,455 75,455 75,455

Figure 4 presents the marginal effects of Risk Acceptance x Polity IV on “Might Boycott.” While

the relationship is similar to the Freedom House scores (Figure 3), there are substantive differences

between the two measures. Figure 4 indicates that moving from the lowest (0) to the highest (40)

risk-acceptance score increases the marginal probability of reporting “Might Boycott” from 26.4% to

34.9% in countries with the highest Polity IV score (20). This represents a substantively and

statistically significant increase of 32% in the most democratic countries (p< .05). Figure 4 also

indicates that moving from the lowest (0) to the highest (40) risk-acceptance score increases the

marginal probability of reporting “Might Boycott” from 21% to 29.7% in countries with a median

Polity IV score (10). This represents a substantively and statistically significant increase of 41% in

median democratic countries (p< .05). Similar to Figure 3, there is no difference between reporting

“Might Boycott” in the least democratic countries (16% to 17.6%).

Figure 3

The overall results for demonstrating and boycotting paint an interesting picture. Democratic

risk-accepting citizens are much more likely to protest than their non-democratic counterparts, as

evidenced by the increasing “steps” of the marginal effect ladder in Figures 1-4. The slopes for the

risk accepting, however, indicate a more nuanced relationship between protesting and system of

government. The little to no substantive or statistical difference between low and high risk-accepting

individuals is unsurprising given the repressive nature of non-democratic regimes (Davenport 1995;

Henderson 1991; Regan & Henderson 2002). In other words, the risk accepting may be more likely

to protest their government, but they are not political kamikazes willing to risk life and limb for futile

endeavors in non-democratic countries that may use lethal force to quell such an uprising. A similar

relationship exists in highly democratic countries.

The overlap in confidence intervals for demonstrating in democracies also suggests that the

risk accepting are no more likely to participate compared to other members of society. This is

perhaps the most surprising result of all the findings. Given that the risk accepting are more likely to

participate in politics (Kam 2012), are more comfortable with uncertainty (Kahneman & Tversky

1979), and more likely to challenge the status quo (Kam & Simas 2012), theory would suggest that

we see an increase in demonstration participation between the high and low risk accepting in

democratic countries. One explanation may be the wording of the question in the World Values

Survey, “attending a peaceful/lawful demonstration” may not invoke enough uncertainty if one

participates in the action. Peaceful/lawful demonstrations may suggest that even the least risk-

accepting members of society are not concerned about potential negative consequences that may

arise from participation. In other words, attending a peaceful/lawful demonstration may not be a

“risky” proposition. Therefore, we must accept part of the null hypothesis for demonstrating.

Figure 4

The boycotting results provide the most compelling and extensive evidence that the risk

accepting are the most likely to protest. While there is not a statistically significant difference

between the low and high risk accepting in non-democratic countries, the risk accepting are much

more likely to boycott in countries with a median or high democracy score compared to other

members of society. The Freedom House scores indicate a 52% increase in the willingness to boycott

in median democratic countries. The Polity IV scores indicate a 41% increase in the willingness to

boycott in median democratic countries and a 32% increase in the most democratic countries. In

contrast to the demonstrating findings, the boycott findings indicate strong evidence to reject the

null hypothesis.

I also include an interaction term between risk acceptance, Freedom House, and Polity IV

scores on “Might Petition” in the second and fourth columns of Table 3. The findings indicate a

negative and statistically insignificant effect for Risk Acceptance x Freedom House (b= -0.0003) and a

positive and statistically insignificant effect for Risk Acceptance x Polity IV (b= 0.0004) on “Might

Petition”, controlling for all other factors in the model. This indicates risk acceptance has a null direct

effect on willingness to petition and by system of government. The null results for risk acceptance

on petitioning is somewhat surprising given Kam’s (2012) findings to the contrary in the United

States. However, these results may differ for a couple of reasons. First, Kam’s (2012) analysis focuses

solely on American petitioning behavior. Second, the current analysis also considers multiple points

in time, government structures, and economic satisfaction. Indeed, Dalton (2016) finds systematic

differences in participation rates between the United States and other Western democracies. These

findings also indicate that we must accept the null hypothesis for risk acceptance and petitioning.

Finally, I include an economic model of risk acceptance on the dependent variables. I include

these models to account for the possibility that the protesting effects are due to economic rather

than political factors (Dalton, van Sickle, and Weldon 2010; Quaranta 2016). Table A4 in the appendix

indicates that there is little to no evidence that economic factors motivate the risk accepting to

protest. The findings indicate that Risk Acceptance x Life Satisfaction and Risk Acceptance x Life

Choice are statistically insignificant predictors of demonstrating (b= 0.0002, p= .83; b= 0.0001, p=

.85), boycotting (b= 0.0003, p= .69; b= 0.0014, p= .06), and petitioning (b= -0.0003, p= .72; b= -0.0004,

p= .57). These results further strengthen the findings that there is an institutional, rather than

economical, effect of risk acceptance on protest behavior.

Table 3 Two-Level Generalized Structural Equation Model Multinomial Logit of the Effect of Risk Acceptance on “Might Petition” by System of Government 2005-2014 ________________________________________________________________________________

Freedom House Polity IV ________________________________________________________________________________ Risk Acceptance Risk Acceptance x Freedom House ---------- -0.0003 ---------- ---------- ---------- (0.0007) ---------- ---------- Risk Acceptance x Polity IV ---------- ---------- ---------- 0.0004 ---------- ---------- ---------- (0.0005) Freedom House 0.2015*** 0.2069*** ---------- ---------- (0.0299) (0.0321) ---------- ---------- Polity IV ---------- ---------- 0.0723*** 0.0644*** ---------- ---------- (0.0134) (0.0162) Risk Acceptance -0.0040# -0.0003 -0.0040# -0.0107 (0.0022) (0.0063) (0.0021) (0.0087) Social Structure Age -0.0105*** -0.0105*** -0.0105*** -0.0105*** (0.0007) (0.0007) (0.0007) (0.0007) Sex -0.1054*** -0.1057*** -0.1048*** -0.1047*** (0.0201) (0.0201) (0.0201) (0.0201) Education 0.0911*** 0.0911*** 0.0910*** 0.0910*** (0.0051) (0.0051) (0.0051) (0.0051) Employment 0.1019*** 0.1021*** 0.1022*** 0.1019*** (0.0206) (0.0206) (0.0206) (0.0206) Religiosity -0.0021 -0.0021 -0.0021 -0.0019 (0.0045) (0.0045) (0.0045) (0.0045) Motivational Attitudes Ideology -0.0134** -0.0134** -0.0135** -0.0135** (0.0043) (0.0043) (0.0043) (0.0043) Political Interest 0.3517*** 0.3517*** 0.3512*** 0.3512*** (0.0113) (0.0113) (0.0113) (0.0113) Systems Support Government Confidence 0.0007 0.0009 0.0012 0.0010 (0.0115) (0.0115) (0.0114) (0.0115) Social Trust 0.0516* 0.0515* 0.0495* 0.0492# (0.0253) (0.0253) (0.0253) (0.0253) Table Continued

________________________________________________________________________________

Freedom House Polity IV ________________________________________________________________________________ Political Behavior Party Member 0.1304** 0.1306** 0.1320** 0.1317** (0.0443) (0.0443) (0.0443) (0.0443) Labor Member 0.1865*** 0.1865*** 0.1872*** 0.1877*** (0.0446) (0.0446) (0.0446) (0.0446) Economic Satisfaction Life Satisfaction -0.0091# -0.0091# -0.0087# -0.0087# (0.0050) (0.0050) (0.0050) (0.0050) Life Choice 0.0065 0.0065 0.0062 0.0063 (0.0044) (0.0044) (0.0044) (0.0044) GDP -3.28-06 -3.39e-06 -4.40e-06 -4.57e-06 (7.45e-06) (7.45e-06) (7.51e-06) (7.51e-06) Constant -1.3732*** -1.4423*** -1.6293*** -1.5016*** N 75,000 75,000 75,000 75,000

7 Discussion and Conclusions

Countries vary in their type of government and their citizens’ attitudes towards risk acceptance.6

And, as the results in the present paper suggest, this variation in risk acceptance and type of

government matters for protest participation. I argue that this relationship is understood based on

a psychological predisposition to accept risk and the potential consequences that may stem from

protest participation, such as government concession or repression. I find that risk acceptance and

system of government predict protesting behavior. However, the substantive effects are limited to

political boycotts. The Freedom House scores indicate a 52% increase, while the Polity IV scores

indicate a 41% increase in the willingness to boycott in median democratic countries for the most

6 Georgia has the lowest mean risk acceptance score of 14.8 compared to Nigeria (21.6) and South Africa (21.5), representing more than one full standard deviation.

risk-accepting members of society. Further, the Polity IV scores indicate a 32% increase in the most

democratic countries. Additional analyses of economic satisfaction indicate that the risk accepting

are not motivated by economic factors. These findings are dependent upon some interesting caveats.

First, Polity IV scores provide a more robust relationship between risk acceptance and boycotting.

Second, risk acceptance has a statistically significant but substantively insignificant effect on

demonstrating. Finally, risk acceptance seems to have no effect on petitioning. While previous

studies find strong relationships between risk acceptance and political participation (Kam 2012),

candidate selection (Kam & Simas 2012; Tomz and van Houweling 2009), and political ambition

(Maestas et al. 2006), the results in the present paper present a more nuanced and limited effect of

risk acceptance on nontraditional political behaviors at the international level. Indeed, the findings

indicate that protest participation may be motivated by factors other than one’s psychological

predisposition to risk. These findings further indicate that, for the risk accepting, there are protest

participation weaknesses in democracies and nondemocracies. Curiously, while democratic citizens have

other avenues to participate in politics and challenge their government, citizens in nondemocracies may only

be able to influence their government through protests. This is especially true in electoral authoritarian

regimes that have symbolic elections (Schedler 2013). In electoral authoritarian regimes we would expect

individuals to protest because other avenues are either unavailable or ineffective. And yet, the differences

between the high and low risk accepting protest behaviors are null in nondemocracies. The risk accepting are

more likely to protest in democracies compared to other members of society, but the behavior is limited in

scope to boycotts. Either the low risk accepting in democracies abstain from demonstrating and petitioning

because they view the activity as futile, or the high risk accepting abstain because they view protesting as

exciting and hopelessly futile.

Ultimately, there is a systematic difference in providing alternative methods to express grievances

with political elites between democratic and non-democratic countries, and in the case of boycotting, the high

risk accepting are more likely than the low risk accepting to participate using the available methods. It should

come as little surprise, then, that non-democratic governments enforce repressive policies with

impunity given that the effectiveness of protesting is contingent upon citizen participation. Countries

with large risk-accepting populations may be more resilient in their pursuit of government

accountability. The findings suggest, theoretically, that countries with low risk-accepting populations

may provide democratic and non-democratic governments more flexibility and stability to enact

repressive policies because most individuals are not willing to risk the consequences of defiance.

These findings suggest a broader and interesting consequence of risk acceptance studies.

Indeed, while risk acceptance may influence a plethora of economic (Kahneman and Tversky 1979),

political (Kam 2012; Kam & Simas 2012; Tomz and van Houweling 2009), and social (Gardner &

Steinberg 2005; Lupton 1993) decisions, these studies are conducted in the relatively safe confines

of countries that are beholden to democratic principles. Once we consider more extreme costs to

participation, such as execution, torture, and imprisonment, there may be vast weaknesses in the

scholarly risk acceptance argument within broader society.

Appendix

Table A1 8-Question Scale of Risk Acceptance Questions _________________________________________________________________________________ Question 1: It is important to this person to think up new ideas and be creative; to do things one’s

own way.

Question 2: It is important to this person to be rich; to have a lot of money and expensive things.

Question 3: Living in secure surroundings is important to this person; to avoid anything that might

be dangerous.

Question 4: It is important to this person to have a good time; to “spoil” oneself.

Question 5: Being very successful is important to this person; to have people recognize one’s

achievements.

Question 6: Adventure and taking risks are important to this person; to have an exciting life.

Question 7: It is important to this person to always behave properly; to avoid doing anything

people would say is wrong

Question 8: Tradition is important to this person; to follow the customs handed down by one’s religion or family. ______________________________________________________________________________ Respondents can choose from “Very Much Like Me”; “Like Me”; “Somewhat Like Me”; “A Little Like Me”; “Not Like Me”; “Not at All Like Me”.

Table A2 List of Democratic and Non-Democratic Countries 2005-2014 _________________________________________________________________________________ COUNTRY Year(s) Democracy Nondemocracy N

Argentina 2006, 2013 X 1,992 Australia 2012 X 1,447 Azerbaijan 2011 X 1,002 Brazil 2006 X 1,493 Bulgaria 2005 X 942 Canada 2006 X 2,143 Chile 2006, 2011 X 1,883 Colombia 2012 X 1,506 Cyprus 2011 X 993 Ecuador 2013 X 1,201 Egypt 2013 X 4,549 Estonia 2011 X 1,509 Ethiopia 2007 X 1,481 France 2006 X 995 Georgia 2009, 2014 X 2,641 Germany 2006, 2013 X 4,043 Ghana 2007, 2012 X 3,065 Great Britain 2005 X 1,036 Hungary 2009 X 1,003 India 2006, 2014 X 3,146 Indonesia 2006 X 1,944 Iraq 2012 X 1,187 Japan 2010 X 2,201 Kazakhstan 2011 X 1,500 Kyrgyzstan 2011 X 1,497 Lebanon 2013 X 1,177 Malaysia 2006, 2012 X 2,501 Mali 2007 X 1,407 Mexico 2012 X 1,996 Moldova 2006 X 1,028 Morocco 2007, 2011 X 2,182 Netherlands 2006, 2012 X 2,859 New Zealand 2011 X 815 Nigeria 2011 X 1,759 Norway 2007 X 1,019 Peru 2012 X 1,158 Philippines 2012 X 1,199 Table A3 Cont’d

_________________________________________________________________________________ COUNTRY Year(s) Democracy Nondemocracy N

Poland 2012 X 950 Romania 2012 X 1,439 Rwanda 2012 X 2,936 Slovenia 2011 X 1,054 South Africa 2013 X 3,481 South Korea 2010 X 1,182 Spain 2007, 2011 X 2,357 Sweden 2006, 2011 X 2,197 Switzerland 2007 X 1,233 Thailand 2007, 2013 X 2,721 Trinidad and Tobago 2006, 2011 X 1,981 Turkey 2007, 2011 X 2,876 Ukraine 2006, 2011 X 2,467 United States 2006, 2011 X 3,408 Uruguay 2006, 2011 X 1,980 Yemen 2014 X 929 Zambia 2007 X 1,452 Zimbabwe 2012 X 1,500 Total 2005-2014 43 13 101,642 _________________________________________________________________________________ Notes: Country regimes are determined using the Freedom House Index and Polity IV scores. Democratic countries require a score of “7 or better in the Electoral Process subcategory and an overall political rights score of 20 or better” (p. 3) and a democracy classification from POLITY IV.

Table A3 Control Variable Coding

Freedom: A twelve-point scale reverse coded where 0 is low levels of rights and liberties and 12 is high levels of rights and liberties. Polity IV: A twenty-point scale reverse coded where 0 is low political freedoms and 20 is high levels of political freedoms. Bicameral: A dichotomous variable coded 1 if the country has a bicameral legislature and 0 otherwise. GDP: The per capita GDP for each country. Population Density: The population per square kilometer for each country. Political Interest: A four-point scale coded 0 for individuals with the lowest interest in politics to 3 for those with the highest level. Employment: A dichotomous variable where 1 is for individuals who are employed and 0 otherwise. Education: A eight-point scale where 1 is for those who lack an elementary education and 8 is for people who earned a university degree or higher. Age: Age in years. Sex: A dichotomous variable coded 1 for women and 0 for men. Year: A dichotomous variable for 2005-2014. Excluded year is 2008 and 2014 for each survey wave. _________________________________________________________________________________

Table A4 Two-Level Generalized Structural Equation Model Multinomial Logit of the Effect of Risk Acceptance on Might Protest by Economic Satisfaction 2005-2014 ____________________________________________________________________________________________________________

Demonstrate Boycott Petition ____________________________________________________________________________________________________________ Risk Acceptance Risk Acceptance x Life Satisfaction 0.0002 ---------- 0.0003 ---------- -0.0003 ---------- (0.0008) ---------- (0.0008) ---------- (0.0009) ---------- Risk Acceptance x Life Choice ---------- 0.0001 ---------- 0.0014# ---------- -0.0004 ---------- (0.0007) ---------- (0.0008) ---------- (0.0009) Life Satisfaction -0.0168 -0.0139** -0.0245 -0.0183*** -0.0028 -0.0085# (0.0150) (0.0046) (0.0162) (0.0048) (0.0165) (0.0050) Life Choice 0.0019 -0.0004 -0.0040 -0.0302* 0.0068 0.0147 (0.0040) (0.0133) (0.0134) (0.0144) (0.0044) (0.0147) Risk Acceptance 0.0006 0.0008 0.0189*** 0.0117* -0.0019 -0.0009 (0.0052) (0.0050) (0.0055) (0.0053) (0.0056) (0.0054) Social Structure Age -0.0139*** -0.0139*** -0.0104*** -0.0104*** -0.0104*** -0.0104*** (0.0006) (0.0006) (0.0006) (0.0006) (0.0007) (0.0007) Sex -0.1022*** -0.1022*** -0.1706*** -0.1706*** -0.1042*** -0.1041*** (0.0179) (0.0179) (0.0186) (0.0186) (0.0201) (0.0201) Education 0.0805*** 0.0805*** 0.0796*** 0.0795*** 0.0912*** 0.0912*** (0.0046) (0.0046) (0.0047) (0.0047) (0.0051) (0.0051) Employment 0.1041*** 0.1041*** 0.0926*** 0.0924*** 0.1032*** 0.1033*** (0.0185) (0.0185) (0.0192) (0.0192) (0.0206) (0.0206) Religiosity 0.0077# 0.0077# -0.0141*** -0.0139*** -0.0032 -0.0032 (0.0040) (0.0040) (0.0041) (0.0041) (0.0045) (0.0045) Motivational Attitudes Ideology -0.0379*** -0.0379*** -0.0403*** -0.0404*** -0.0137** -0.0136** (0.0040) (0.0040) (0.0041) (0.0041) (0.0043) (0.0043) Political Interest 0.3488*** 0.3488*** 0.3687*** 0.3688*** 0.3513*** 0.3512*** (0.0102) (0.0102) (0.0106) (0.0106) (0.0113) (0.0113) Table Continued

____________________________________________________________________________________________________________

Demonstrate Boycott Petition ____________________________________________________________________________________________________________ Systems Support Government Confidence 0.0122 0.0122 -0.0550*** -0.0552*** 0.0025 0.0025 (0.0105) (0.0105) (0.0110) (0.0110) (0.0115) (0.0115) Social Trust 0.0817*** 0.0817*** 0.1684*** 0.1690*** 0.0457# 0.0456# (0.0217) (0.0217) (0.0220) (0.0220) (0.0253) (0.0253) Political Behavior Party Member 0.1440*** 0.1442*** 0.0233 0.0237 0.1302** 0.1298** (0.0403) (0.0403) (0.0393) (0.0393) (0.0442) (0.0442) Labor Member 0.2162*** 0.2161*** 0.1604*** 0.1602*** 0.1829*** 0.1831*** (0.0398) (0.0398) (0.0385) (0.0385) (0.0446) (0.0446) Economic Satisfaction GDP 0.00001* 0.00001* -0.00001* -0.00001* 5.02e-06 4.98e-06 (5.54e-06) (5.54e-06) (5.47e-06) (5.47e-06) (7.36e-06) (7.36e-06) Constant -0.3736* -0.3767*** -0.7180*** -0.5872*** -0.8557*** -0.8720*** N 75,455 75,455 75,455 75,455 75,000 75,000

____________________________________________________________________________________________________________ Notes: Table entry is the multinomial regression coefficient set to the country and individual level of analysis. Dependent variable is scaled as 0 (Never Do); 1 (Might Do); 2 (Have Done) with baseline category “Never Do.” ***p<.001; **p<.01; *p<.05; #p<.10, two-tailed.

References

Berinsky, Adam J. and Jeffrey B. Lewis. 2007. An Estimate of Risk Aversion in the U.S.

Electorate. Quarterly Journal of Political Science 2: 139–54.

Blais, Andre. 2006. What Affects Voter Turnout? Annual Review of Political Science 9(1):

111-125.

Celestino, Mauricio Rivera & Kristian Skrede Gleditsch. 2013. “Fresh Carnations or all Thorn,

no Rose? Nonviolent Campaigns and Transitions in Autocracies.” Journal of Peace

Research 50(3): 385-400.

Dalton, Russell J. 2016. The Good Citizen: How a Younger Generation is Reshaping American Politics.

SAGE Publications Inc.

Davenport, Christian. 1995. Paths to State Repression. Lanham: Rowman & Littlefield Publishers,

Inc.

Davenport, Christian. 1999. "Human Rights and the Democratic Proposition." The Journal of Conflict Resolution 43:92–116.

DeNardo, J. 1985. Powers in Numbers: The Political Strategy of Protest and Rebellion. Princeton:

Princeton University Press.

Ehrlich, Sean and Cherie Maestas. 2010. Risk Orientation, Risk Exposure, and Policy Opinions:

The Case of Free Trade. Political Psychology 31(5):657-684.

Franklin, James C. 2009. Contentious Challenges and Government Responses in Latin America.

Political Research Quarterly 62(4): 700-714.

Freese, Jeremy. 2004. Risk Preferences and Gender Differences in Religiousness: Evidence

from the World Values Survey. Review of Religious Research 46(1): 88-91.

Gardner, Margo and Laurence Steinberg. 2005. “Peer Influence on Risk Taking, Risk Preference, and

Risky Decision Making in Adolescence and Adulthood: An Experimental Study.” Development Psychology 41(4): 625-635.

Gartner, Scott Sigmund and Patrick M. Regan. 1996. Threat and Repression: The Nonlinear Relationship between Government and Violence. Journal of Peace Research 33(3): 273- 287. Gurr, Ted Robert. (1970). Why Men Rebel. Princeton: Princeton University Press. Henderson, Conway. 1991. Conditions Affecting the Use of Political Repression. Journal of Conflict Resolution 35(1): 120-142.

Högström, John. 2013. “Does the Choice of Democracy Measure Matter? Comparisons between the

Two Leading Democracy Indices, Freedom House and Polity IV.” Government and

Opposition 48(2): 201-221.

Hoyle, Rick H., Michael T. Stephenson, Philip Palmgreen, Elizabeth Pugzles Lorch, and R. Lewis Donohew. Reliability and validity of a brief measure of sensation seeking. Personality and Individual Differences 32: 401-414. Inglehart, Ronald. (1977). The Silent Revolution: Changing Values and Political Styles among

Western Publics. Princeton: Princeton University Press.

Inglehart, Ronald. (1997). Modernization and Postmodernization: Cultural, Economic and

Political Change in 43 Societies. Princeton: Princeton University Press.

Kahneman, Daniel and Amos Tversky. 1979. Prospect Theory: An Analysis of Decision under

Risk. Econometrica 47(2): 263-292. Kam, Cindy D. (2012). Risk Attitudes and Political Participation. American Journal of

Political Science 56(4):817-836.

Kam, Cindy D., and Elizabeth N. Simas. (2012). Risk Attitudes, Candidate Characteristics, and

Vote Choice. Public Opinion Quarterly 76(4): 747-760.

Kam, Cindy D. and Elizabeth N. Simas. (2010). Risk Orientations and Policy Frames. The

Journal of Politics 72(2):381-396.

Kapstein, Ethan B. and Converse, Nathan. 2008. "Why Democracies Fail." Journal of Democracy, vol.

19(4): 57-68. Project MUSE, doi:10.1353/jod.0.0031.

Levy, Jack S. 2003. Applications of Prospect Theory to Political Science. Synthese 135(2):215-

241.

Lijphart, Arend. 2012. Patterns of Democracy: Government Forms and Performance in Thirty-Six

Countries. New Haven and London: Yale University Press.

Lupton, Deborah. 1993. “Risk as Moral Danger: The Social and Political Functions of Risk Discourse

in Public Health.” International Journal of Health Services 23(3): 425-435.

Moscovici, S. (1976). Social influence and social change. London: Academic Press.

Miller, Alan S. 2000. Going to Hell in Asia: The Relationship between Risk and Religion in a

Cross Cultural Setting. Review of Religious Research 42(1): 6–14.

Nemeth, C. J. (2003). Minority dissent and its ‘hidden’ benefits. New Review of Social Psychology, 2,

11–21.

Norris, Pippa, Stefaan Walgrave, and Peter Van Aelst. 2005. Who Demonstrates? Antistate Rebels,

Controversial Participants, or Everyone? Comparative Politics 37(2): 189-205.

Piven, Frances Fox & Richard A. Cloward. (1977). Poor People’s Movements: Why They

Succeed, How They Fail. New York: Vintage Books.

Primo, Jacobsmeier, and Milyo 2007 Rhine, S. L. 1996. An Analysis of the Impact of

Registration Factors on Turnout in 1992. Political Behavior 18:171-185.

Poe, Steven C., C. Neal Tate, and Linda Camp Keith. 1999. Repression of the Human Right to

Personal Integrity Revisited: A Global Cross-National Study Covering the Years 1976- 1993. International Studies Quarterly 43:291–313.

Przeworski, Adam, Michael E. Alvarez, Jose Antonio Cheibub, and Fernando Limongi. 2000.

Democracy and Development: Political Institutions and Well-Being in the World, 1950-1990. Cambridge: Cambridge University Press.

Quaranta, Mario. 2016. “Protesting in ‘hard times’: Evidence from a comparative analysis of

Europe, 2000-2014.” Current Sociology 64(5): 736-756. Regan, Patrick M. and Errol A. Henderson. 2002. Democracy, Threats and Political Repression in

Developing Countries: Are Democracies Internally Less Violent? Third World Quarterly 23:119–136.

Schedler, Andreas. 2013. The Politics of Uncertainty: Sustaining and Subverting Electoral

Authoritarianism. Oxford University Press.

Stephan, Maria J. & Erica Chenoweth. 2008. “Why Civil Resistance Works: The Strategic Logic

of Nonviolent Conflict.” International Security 33(1): 7-44.

Vanhuysse, Pieter. 2006. Divide and Pacify: Strategic Social Policies and Political Protest in Post-

Communist Democracies. Budapest and New York: Central European University Press. Weber, Elke U., Ann-Renee Blais, and Nancy E. Betz. 2002. A Domain-specific Risk-attitude Scale: Measuring Risk Perceptions and Risk Behaviors. Journal of Behavioral Decision Making 15: 263-290.

Zuckerman, Marvin. 1979. Sensation Seeking: Beyond the Optimal Level of Arousal. Hillsdale, NJ: John Wiley & Sons. Zuckerman, Marvin. 2007. Sensation Seeking and Risky Behavior. Washington, DC: American Psychological Association.