1

Income Inequality, Citizen Polarization, and Political Protest

John D. Griffin University of Colorado Boulder

John.Griffin@colorado.edu

Chad Kiewiet de Jonge Centro de Investigación y Docencia Económicas

chad.kiewiet@cide.edu

Abstract

Over the past several decades, two important trends have emerged across the world: a dramatic rise in economic inequality within nations and an increase in the frequency of mass protest. Although it is tempting to link these two trends, prior studies of the relationship between economic inequality and political unrest have proven contradictory and thereby inconclusive. We argue that existing studies centered on deprivation theory do not adequately account for the political mechanisms by which inequality does or does not lead to protest. Building on Gurr’s theory of deprivation, we suggest that research needs to examine not only perceptions of deprivation born of inequality, but also, more importantly, polarization in such perceptions. Using data from 12 separate cross national survey projects, we build an original, comparable measure of political polarization and examine the effect of income inequality on polarization and the effect of polarization on political protest in 93 countries over the period 1973 to 2010. These analyses yield three novel conclusions. First, we find that countries with higher levels of income inequality tend to have more polarized citizens. Moreover, countries with polarized citizens are more likely to endure nonviolent protest. Finally, protests are most likely in countries where average citizen satisfaction is high but this satisfaction is also polarized, a finding somewhat contrary to Gurr’s original theoretical framework.

2

Income Inequality, Citizen Polarization, and Political Protest

I. Introduction

The last five years have seen major political protests in nearly every corner of the globe

including Tunisia, Egypt, Libya, Yemen, Syria, Venezuela, the United States, the United

Kingdom, Greece, Spain, Turkey, South Africa, Argentina, Ukraine, and Thailand. Indeed,

taking a longer view the incidence of political protest appears to be very much on the rise (e.g.

Norris 2002, 197): “the systematic cross-national survey evidence that is available confirms that

a significant long-term rise in protest politics has indeed occurred. . . . [T]he proportion of

citizens engaged in protest politics has risen, and risen dramatically, during the late twentieth

century.” We could also point to the emergence of anti-protest laws in Canada, the United

States, and Spain (among others) as an indicator of this trend.1 There is substantial evidence that

these political protests have real consequences for governments and citizens alike: “protests can

build political movements that ultimately affect policymaking, and . . . they do so by influencing

political views” (Madestam, Shoag, Veuger, and Yanagizawa-Drott 2013, 1633; see also Collins

and Margo 2007; Lohmann 1993, 1994; Skocpol and Williamson 2011; Finkel and Muller 1998).

Another trend over the last several decades that has attracted considerable attention has

been a widespread and significant increase in income inequality within countries (e.g. Piketty

2013). The Gini coefficient, a common measure of income inequality that ranges from 0 (when

all citizens have identical incomes) to 1 (when all income goes to only one person), stood at an

average of 0.29 in OECD countries in the mid-1980s. By the late 2000s, however, the mean had

increased by almost 10% to 0.316. The coefficient rose in 17 of the 22 OECD countries for

1 http://www.theguardian.com/world/2013/nov/21/spain-government-strict-anti-protest-laws

3

which long-term data series are available (OECD 2011).2 Stated another way, since 1980 the

share of wealth owned by the richest 1% has expanded in all but two of the 26 nations tracked by

researchers in the World Top Incomes Database (Puzzhangera 2014).3 This has led to

considerable discussion about the consequences of growing inequality. According to one

commentary, “A consensus has formed around the idea that income inequality is the single most

important challenge facing the world” (Global Post and Ford Foundation).4

It is tempting to link these two trends. However, substantial previous research on the

relationship between inequality and protest has proven contradictory, as studies have theorized

and found positive, negative, curvilinear, and null relationships between inequality and protest.

In this paper, we argue that existing analyses overlook two important components of grievance

theory on which the theories generally are based: (1) the translation of objective sources of

grievance (e.g. inequality) to subjective perceptions of relative deprivation, and (2) societal

polarization in such perceptions. We argue that this second component—polarization in

deprivation—is key for understanding the rise of protest across the world.

Using data from 12 separate cross national survey projects, we develop new measures of

grievance and polarization in such grievances that are comparable across countries. We then

examine the effect of income inequality on political grievance and polarization, and the effect of

political grievance and polarization on political protest in 115 countries from 1973 to 2010. Our

analyses yield three novel conclusions. First, we find that countries with higher levels of income

inequality also tend to have more polarized citizens. Moreover, countries with polarized citizens

are more likely to experience nonviolent protest, despite the fact that that there is no direct

2 http://www.oecd.org/els/soc/49499779.pdf. 3 http://topincomes.g-mond.parisschoolofeconomics.eu/; http://www.latimes.com/business/la-fi-mo-oxfam-world-economic-forum-income-inequality-20140120-story.html. 4 http://www.globalpost.com/special-reports/global-income-inequality-great-divide-globalpost.

4

relationship between inequality and protest. Finally, protests are most likely in countries where

average citizen grievances are actually rather low, but this satisfaction is also polarized, a finding

that is contrary to key predictions in Gurr’s original formulation.

II. Background and Expectations

According to grievance theory, (e.g. Gurr 1970), mass protest results from a sense of

relative deprivation, including dissatisfaction with poverty, economic conditions, or other

grievances related to living conditions. Scholars with this perspective have sought to test the

relationship between various potential sources of such deprivation and protest, and the

relationship between economic inequality and protest has proven to be one of the most studied

implications of the grievance theory research agenda (Lichbach 1988, 1989; Sigelman and

Simpson 1977; Boix 2003; Russett 1964; Acemoglu and Robinson 2009; Nagel 1974). The

results of such analyses are all over the map, with studies suggesting a positive, negative,

convex, concave, and null relationship (see Lichbach 1989 for a review).

For example, numerous studies argue for a linear relationship between a country’s level

of income inequality and political protest activity (e.g., Lichbach 1988; Sigelman and Simpson

1977; Boix 2003). Similarly, Russett finds that in societies with more equal land distribution

democracy is more stable (Russett 1964). According to one survey piece, “there is little surprise

that inequality is also strongly associated with political instability. While the sources of political

conflict vary from country to country, conflict generally originates from severe social grievances,

including class conflict and the perception of inequality among ethnic, religious or other groups.

. . . [W]e find that unequal societies, in general, are much more prone to political instability, or,

5

in other words, to be destabilized or overthrown by unconstitutional or forceful means, which

includes politically-motivated violence and terrorism” (UNICEF Report).

Others argue that protest is least likely in societies with either very little or a great deal of

income inequality, but for different reasons. In societies with little inequality there is little basis

for grievance. In societies with high levels of inequality, meanwhile, the political opportunity

structure may not facilitate political protest (Acemoglu and Robinson 2009). To explain the

same pattern, Nagel contends that in societies with very high levels of income inequality the poor

are not as sensitive to inequality because in this context the poor are unlikely to compare their

economic position to that of the wealthy (Nagel 1974).

We argue that these studies suffer from two major shortcomings. Their first problem is

that there is a disconnect between the objective (potential) source of grievance (economic

inequality) and actual subjective appreciation of relative deprivation, a key distinction made by

Gurr (1970). In other words, to the extent that economic inequality is positively related to

protest, its effects are dependent on individuals subjectively viewing such inequality as causing a

discrepancy between their “value expectations and their value capabilities” (Gurr 1970, 12). An

objective source of grievance is thus not enough to spur protest; it is the realization of a gulf

between life outcomes and the results to which individuals believe they are entitled that is also

required.

For two reasons, citizens’ subjective evaluations of inequality may not be a carbon copy

of objective conditions. First, citizens may not hold accurate perceptions of the degree of

income inequality (Delli Karpini and Keeter 1996). For instance, in January 2014 the Pew

Research Center for People and the Press conducted a survey of Americans in which respondents

were asked if “the gap between the rich and everybody else in the U.S. had increased, decreased,

6

or stayed about the same.” Given the widespread attention to growing income inequality in the

U.S., it is rather remarkable that 35% of Americans either thought inequality had remained

unchanged (25%), had decreased (8%), or did not know (2%).5 Similarly, several studies

focusing on the U.S. show that Americans both underestimate (Norton and Ariely 2011) and

overestimate (Chambers et al. 2013) the actual level of income inequality.

Second, even if citizens hold accurate perceptions about the level of income inequality,

their expectations about whether inequality is a legitimate basis of grievance for them will vary.

For instance, if citizens in a country tend to believe that individuals are chiefly responsible for

their own economic circumstances, rather than the government being responsible, then even in

the face of substantial inequality grievances will be less likely to emerge (Verba 1985). For both

of these reasons, income inequality as an objective condition will only be imperfectly related to

income-based political grievance.

While a plethora of studies seeking to tie economic inequality and other objective

grievance sources to protest have essentially ignored this subjective psychological component,

more recent survey-based work has sought to test relative deprivation predictions using survey

items that actually tap into perceived grievances (e.g., Dalton, Van Sickle, and Weldon 2009;

Booth and Seligson 2009, ch. 5). Although the shift toward examining subjective grievances

rather than (or in addition to) potential objective sources of deprivation moves analyses closer to

the original logic undergirding Gurr’s grievance theory, results have proven mixed, particularly

in studies focusing on long-standing democracies (e.g., Dalton, Van Sickle, and Weldon 2009).

Those who put great stock in the need for collective action have taken this lack of strong or

consistent relationships between expressed dissatisfaction with economic, political, or life

5 http://www.people-press.org/files/legacy-questionnaires/4-28-14%20Inequality%20Topline.pdf.

7

outcomes and protest as support for their research agenda, which suggests that protest potential is

essentially unrelated to actual grievances (Lichbach 1996).

The second problem with prior treatments of the grievance/protest relationship is that

they overly focuses on “deprivation” without sufficiently taking into account the “relative”

component of Gurr’s theory. Gurr argues that mobilization will actually be greatest not only

when a particular individual or group is dissatisfied with their lot, but also when other reference

groups in society are not as badly deprived. In other words, it is the inequality in perceived

outcomes that is the largest driver of protest. Neither objective economic inequality nor

subjective grievance—the main focuses of previous work—is sufficient; rather, what is more

important is the polarization of subjective perceptions of deprivation. Protest from this

perspective is thus more likely when some groups in society are particularly aggrieved while

others are quite satisfied with their lives. Thus we argue that the “relative” component of Gurr’s

relative deprivation grievance theory, and in particular the relationship between grievances

arising from economic inequality and protest, has largely been ignored in empirical analyses.

Another way of describing the second hole in the empirical literature that we identify is

to say that studies have overwhelming focused on mean levels of expressed grievances rather

than the dispersion of such perceptions. For example, at the cross national level, studies have

examined the relationship of various subjective measures of perceived grievances with the

democratic political system, expressed in the language of life satisfaction, satisfaction with the

performance of the political system or current government, and sociotropic and egotropic

evaluations of the economy and protest (Anderson and Mendes 2006; Dalton, Van Sickle, and

Weldon 2009; Booth and Seligson 2009). These studies find that a greater sense of grievance

decreases citizens’ interest in politics and their willingness to participate in conventional forms

8

of political activism while at the same time increasing the likelihood of more unconventional

protest activities (Booth and Seligson 2009; Dalton 2004, ch. 8; Norris 2011, ch. 11). High levels

of grievance are likely to reduce good governance and the application of the rule of law, as under

these conditions citizens are less likely to comply voluntarily with laws, including tax

noncompliance and fraudulent application for government benefits (Dalton 2004, 166; Norris

2011, ch. 11; Tyler 2006). Virtually all these studies share a focus on the mean or median

constituent. In so doing, they generally do not attend to the disagreement in evaluations that

exists within countries, or whether citizens are politically polarized.

Admittedly, some scholars have begun to shift the focus from comparing the experience

of the average citizen across democratic systems to comparing the experience of citizens within a

single political system (Bartels 2008; Gilens 2005; Jacobs and Page 2005; Adams and Ezrow

2009). These studies seek to disaggregate the public, asking whether within countries some

citizens better evaluate the political system than others, or whether judgments are polarized. To

date, though, these studies of citizen polarization have not been truly comparative across

countries. Most comparative studies only include a handful of cases or are single-country case

studies (Adams and Ezrow 2009; Ezrow 2010; see Powell 2007 for review). This is not

surprising, because it is very difficult to compare polarization in multiple countries in an

encompassing manner. And yet, to the extent that Gurr’s (1970) original predictions about the

importance of relative differences in grievances are correct, we also must examine the

polarization in people’s evaluations of their life situation and/or the political system.

Finally, little thought has been given in prior studies to the relationship between the

average level of grievance in a country and the dispersion of the country’s grievances. Figures

1(a) and 1(b) illustrate four different ideal type theoretical distributions relating mean levels of

9

grievances with polarization or dispersion of such grievances. In both figures, 푥̅ denotes the

average level of grievances in a society, with the average level of perceived deprivation lower for

the two distributions in Figure 1(a) compared to the two distributions in Figure 1(b). Examining

Figure 1(a), in countries with what might be termed inclusive distributions, there is a low average

level of grievance and only small differences in such evaluations across groups in society. In

countries with majoritarian distributions, there can be an equally low levels of grievance on

average (푥̅), but much more pronounced differences in evaluations across groups in society.6 In

these countries, most citizens feel well served by the political system but a smaller minority does

not.

Turning to Figure 1(b), which shows distributions with equally high average levels of

grievance, in countries with what could be termed oligarchic distributions, there is a high

average level of grievance and significant differences across groups in society. In these

countries, the vast majority of citizens are dissatisfied with the functioning of the political system

but a small minority feels well served by it. Finally, in countries with exclusionary distributions,

grievance is equally high on average and only small differences in such evaluations occur across

groups in society.

Most tests of grievance theory have assumed that high average levels of grievance are the

key driver of protest, while ignoring the differences in distributions of such grievance. That is,

studies have generally assumed that we would see higher levels of protest in polities with

distributions in Figure 1(b) and lower levels of protest in societies with distributions pictured in

Figure 1(a). What we contend, however, is that protest should actually be highest in distributions

with significant polarization, that is, it should be highest in countries with what we term

6 In this case, we might say that in inclusive distributions, the modal level of satisfaction is lower than it might otherwise be in perhaps due to efforts integrate more citizens into the “coalition of supporters.”

10

FIGURE 1(A): TWO DISTRIBUTIONS

WHEN GRIEVANCE PERCEPTIONS ARE LOW ON AVERAGE

FIGURE 1(B): TWO DISTRIBUTIONS WHEN GRIEVANCE PERCEPTIONS ARE HIGH ON AVERAGE

Perceived Deprivation / Grievance

Inclusive System Majoritarian System

Perceived Deprivation / Grievance

Exclusive System Oligarchic System

푥̅

11

majoritarian and oligarchic distributions of grievances, and lower in countries in which there are

high levels of accord, whether that be low levels of shared grievance (inclusive) or high levels

thereof (exclusive). According to Gurr’s theory, protest should prove particularly high in

countries with oligarchic distributions, where there are both high levels of grievances and large

gulfs in satisfaction between different groups.

III. Measuring Grievances and Polarization Cross-Nationally

In order to test our contention that accounting for both mean levels of grievances and

polarization in such grievances are important for understanding their potential role as mediators

between objective economic inequality and protest, , we constructed new, cross-nationally

comparable measures of the constructs based on a common survey question that has been

included on cross-national surveys across the globe over the last several decades. More

specifically, we utilize the mean and variance, respectively, in citizens’ satisfaction with the way

that democracy works in their country as our key measures of perceived deprivation and

polarization in such views.. While the ubiquity of the item is one of its principal advantages, we

argue that it also serves as a good summary measure of grievances with the political system (see

also Fuchs 1993, 1999; Fuchs, Guidorossi, and Svenson 1995; Harmel and Robertson 1986;

Klingemann 1999; Lockerbie 1993; McDonough, Barnes, and Lopez Pina 1986; Morlini and

Tarchi 1996; Toka 1995; Weil 1989; Widmaier 1988, 1990). As noted by Anderson and Guillroy

(1997, 70), the "satisfaction with the way democracy is working item is indicative of “system

support at a low level of generalization,” meaning that the item captures political evaluations that

lie between diffuse support for the political regime and specific support for the current

government (Norris 2011; Easton 1965). This more medium-term evaluation reflects a person's

perceptions about the ability of the political system to offer the opportunity for his/her views to

12

be implemented, or the level of more deep-rooted grievance. It exhibits the degree to which

strong governmental performance is realized (or not) over a period of time, as well as citizen

agreement with policy outputs when their favored party controls government as well as when it

does not. Thus, when citizens are not only aggrieved with current government policies but also

doubt that future democratic governments will also address their grievances, we would expect

higher levels of protest. Our further argument is that the variance in such grievances is perhaps

even more important than its mean.

We operationalize grievance using this measure of satisfaction with democracy. While

the exact question wording varies slightly across surveys, the standard satisfaction item asks

respondents to evaluate the functioning of democracy in their country using a four point scale.

We compiled surveys from12 cross-national survey projects that include this four-point item,

resulting in data for 93 countries between 1973 and 2010, or a total of 1,029 country-years of

data.7 For each country-year, our measure of grievance is simply the average of the inverted

satisfaction with democracy response scale, rescaled to a range of 0 to 100.8 We restricted our

analysis to countries that were at least semi-democratic (i.e., greater than or equal to “0” on the

Polity combined autocracy-democracy index).

Our measure of polarization is based on the aggregate level standard deviation in

responses concerning citizen grievance. However, since the response scale for the satisfaction

with democracy item is relatively short, the observed variance may be truncated, especially in

cases in which the means approach the scale minima and maxima. For example, if on a four-

7 The sources of data include the Afrobarometer, Asianbarometer, Candidate Countries Eurobarometer, Central and Eastern Eurobarometer, Comparative Study of Electoral Systems (CSES), Eurobarometer, European Social Survey, AmericasBarometer, LatinBarometer, New Europe Barometer, Post-Communist Publics Study, and the World Values Study. 8 For certain country-years, multiple surveys were available. To create our grievance and polarization measures, we averaged across the survey means and standard deviations, respectively within each country-year.

13

point scale the country average is 3.5 (a high average performance evaluation), it is likely that

some respondents would have chosen a higher value (say a response corresponding to 5 or 6) if it

were available to them.9 In other words, there is a structural dependency between the mean and

standard deviation. To account for this problem, we follow the adjustment proposed by Dehly

and Kohler (2011), which divides the observed standard deviation by the maximum possible

standard deviation at the corresponding mean value.10 As a consequence, our measure of

polarization is in effect the percent of the maximum standard deviation at the respective mean

value.11

Figure 2 plots polarization versus grievance on the y and x axes, respectively. Country-

year values are indicated with small grey labels while country averages across all survey years

are plotted with larger black labels. The level of grievance in the sample ranges from a low of 41

in Denmark in 2006 to a high of 87.4 in Paraguay in 2002, with a median value of 63.9 in

Mexico in 2012, and country average ranging from a low of 46.5 in Ghana to a high of 80.4 in

Ukraine. On the other hand, polarization ranges from a low of 36.4 in the case of South Korea in

2001 and a high of 84.9 for Lesotho in 2000, with country averages ranging from 41.1 in South

Korea to 76.8 in Lesotho.

9 More generally, the variance of a mean is partially dependent on the relative placement of the mean for items with definite endpoints. As a result, variance estimates for countries demonstrating relatively high or low means may understate the true amount of disagreement among citizens about the quality of representation in the country. For a detailed discussion in terms of measures of inequality of happiness, see Kalmijn and Veehoven (2005) and Delhey and Kohler (2011). 10 The maximum standard deviation at a given mean value is calculated as follows: max(휎) =

( )∗( )∗ , where u and l represent the upper and lower bounds of the satisfaction with democracy

scale, µ is the mean satisfaction with democracy, and N is the number of respondents. 11 We also rescaled the polarization item to a 0-100 scale.

14

FIGURE 2: SCATTERPLOT OF GRIEVANCE AND POLARIZATION, BY COUNTRY-YEAR AND COUNTRY AVERAGES

There is evidence of a slight positive correlation between the mean level of grievance and

polarization (r=0.25), although the figure makes clear that there are countries that approach the

four ideal type distributions described in Figure 1. Figure 3 plots the distribution of responses to

the satisfaction with democracy item for four country-years that best approximate these ideal

types. Norway in 1997 perhaps best approaches the inclusive ideal type distribution, since in that

year Norwegians overwhelming were satisfied with democratic performance, with fewer than 10

percent of respondents expressing dissatisfaction. While the average level of grievance is

essentially equal for both Norway 1997 and Ghana 2008, there is significantly greater

disagreement in the latter case, with over half of all respondents saying they were “very

AL92

AL93

AL94AL95

AL96

AL02AL05

AL08

AR95AR96

AR97

AR98

AR99AR00

AR01

AR02

AR03

AR04AR05

AR06 AR07

AR08

AR09

AR10

AM92

AM93

AM94

AM95

AM08

AU96

AU04

AU07

AT95 AT97

AT98AT99

AT00AT01

AT03

AT04AT05AT06AT07

AT08AT09

AT10

AT11

BD02

BY92BY93

BY94

BY95BE73

BE76

BE77 BE78

BE79

BE80

BE81

BE82BE83

BE84BE85

BE86BE87BE88

BE89BE90BE91

BE92BE93BE94

BE95

BE97BE98

BE99BE00

BE01

BE03

BE04

BE05BE06BE07 BE09

BE10

BE11

BJ05BJ08

BO96BO97

BO98 BO00

BO01

BO02

BO03

BO04

BO05

BO06

BO07BO08

BO09

BO10

BW99 BW03

BW05BW08

BR95

BR96

BR97

BR98

BR00

BR01

BR02

BR03BR04

BR05

BR06BR07

BR08

BR09

BR10

BG92

BG93

BG94

BG95 BG96BG97

BG99

BG01BG02BG03 BG04

BG05

BG06BG07

BG08

BG09

BG10BG11

BF08

KH08

CA97

CA00

CA04

CA06

CA08

CA10

CV02

CV05

CV08

CL95

CL96

CL97CL98

CL00 CL01

CL02

CL03

CL04

CL05

CL06

CL07

CL08

CL09

CL10

CO96

CO97

CO98CO00

CO01CO02

CO03

CO04

CO05

CO06

CO07

CO08

CO09

CO10

CR96

CR97

CR98

CR00

CR01

CR02

CR03

CR04

CR05

CR06

CR07

CR08

CR09

CR10

HR99

HR05HR06

HR07

HR08

HR09

HR10

CY02

CY03

CY04CY05

CY06

CY07

CY08

CY09

CY10

CY11

CZ93

CZ94

CZ95

CZ96

CZ97

CZ99

CZ01

CZ02

CZ03CZ04

CZ05

CZ06CZ07CZ08

CZ09 CZ10

CZ11

DK73

DK76DK77

DK78DK79

DK80

DK81

DK82

DK83

DK84DK85

DK86

DK87

DK88

DK89

DK90

DK91DK92

DK93DK94

DK95DK97

DK98 DK99

DK00DK01

DK03

DK04DK05DK06

DK07DK08

DK09DK10DK11

DO04DO05

DO06

DO07

DO08

DO09

DO10EC96

EC97

EC98

EC00

EC01

EC02EC03EC04

EC05

EC06

EC07

EC08

EC09

EC10

SV96

SV97SV98SV00

SV01

SV02SV03

SV04SV05SV06

SV07

SV08

SV09

SV10

EE92

EE93EE94

EE95

EE96EE97

EE99

EE01

EE02EE03

EE04EE05EE06

EE07

EE08EE09

EE10EE11

FI93

FI94

FI95

FI97FI98

FI99FI00

FI01

FI03

FI04

FI05

FI06

FI07

FI09

FI10

FI11

FR73FR76FR77FR78 FR79 FR80

FR81

FR82

FR83FR84FR85

FR86 FR87

FR88FR89

FR90

FR91

FR92FR93

FR94 FR95FR97

FR98

FR99FR00FR01FR02

FR03 FR04

FR05

FR06FR07

FR08FR09

FR10

FR11

GE92

GE94

GE95

GE96

GE08

DE91DE92

DE93DE94

DE95

DE97DE98DE99

DE00 DE01

DE02

DE03 DE04

DE05

DE06DE07

DE08

DE09DE10

DE11

GH99

GH02

GH05

GH08

GR80

GR81

GR82

GR83GR84

GR85

GR86

GR87

GR88GR89

GR90

GR91

GR92GR93

GR94

GR95

GR97

GR98

GR99

GR00

GR01

GR03

GR04GR05GR06GR07 GR08

GR09GR10 GR11

GT96

GT97

GT98 GT00

GT01

GT02

GT03

GT04

GT05GT06

GT07

GT08

GT09

GT10

GY06

GY10

HT06HT08

HT10

HN96HN97

HN98

HN00 HN01

HN02

HN03

HN04

HN05

HN06

HN07

HN08

HN09

HN10

HU92HU93

HU94

HU95HU96

HU97HU98

HU99HU01

HU02

HU03

HU04

HU05HU06 HU07

HU08

HU09

HU10

HU11

IN01

ID01

ID06

IR00

IE73

IE76IE77

IE78

IE79IE80

IE81

IE82IE83IE84IE85IE86IE87IE88IE89IE90

IE91

IE92IE93

IE94

IE95IE97

IE98

IE99IE00IE01

IE02

IE03

IE04IE05

IE06

IE07 IE08

IE09

IE10IE11

IL96 IL03

IL06IT73

IT76IT77IT78

IT79

IT80

IT81IT82

IT83IT84IT85

IT86IT87

IT88IT89 IT90IT91 IT92IT93

IT94

IT95IT97

IT98

IT99

IT00

IT01

IT03

IT04IT05

IT06

IT07

IT09

IT10

IT11

JM06

JM08JM10

JP96

JP00

JP03JP04JP07

KE03

KE05

KE08

KR00 KR01

KR03

KR04

KR06

KR08

KG05LV92

LV93 LV94LV95

LV96

LV97

LV99

LV01

LV02LV03

LV04

LV05

LV06LV07

LV08

LV09

LV10

LV11

LS00

LS03

LS05

LS08

LR08

LT92

LT93LT94

LT95

LT96

LT97

LT99

LT01

LT02LT03

LT04

LT05LT06

LT07

LT08

LT09LT10

LT11

MK93

MK94

MK95

MK96

MK01

MK07

MK08

MK09 MK10

MG05 MG08

MW99

MW03 MW05

MW08

MY06

ML01ML02

ML05

ML08

MX95

MX96

MX97

MX98

MX00

MX01

MX02

MX03

MX04

MX05MX06

MX07

MX08

MX09

MX10MD92

MD96

MD02

MD08MN03

MN06

MZ02

MZ05

MZ08

NA99

NA03

NA05

NA08

NL73

NL76

NL77NL78NL79

NL80

NL81

NL82NL83NL84

NL85

NL86NL87NL88

NL89NL90NL91NL92

NL93NL94

NL95

NL97NL98

NL99

NL00

NL01

NL02

NL03NL04

NL05

NL06

NL07NL08

NL09

NL10

NL11

NZ96

NZ02

NZ08

NI96NI97

NI98 NI00

NI01

NI02

NI03

NI04

NI05

NI06

NI07

NI08

NI09

NI10NG00

NG03 NG05

NG08

NO90NO91NO92

NO93NO94 NO95NO97

NO01

NO05

NO08

NO09

PA96

PA97

PA98PA00

PA01

PA02

PA03PA04

PA05PA06

PA07PA08

PA09

PA10

PY95

PY96

PY97

PY98

PY00

PY01

PY02PY03

PY04

PY05PY06

PY07

PY08

PY09

PY10

PE95

PE96

PE97

PE98

PE00PE01 PE02 PE03

PE04

PE05

PE06PE07

PE08

PE09

PE10

PE11

PH01

PH02

PH04

PH05PH10

PL92

PL93PL94

PL95 PL96

PL97PL99

PL01

PL02 PL03

PL04PL05

PL06

PL07

PL08

PL09

PL10PL11

PT85

PT86

PT87PT88

PT89PT90

PT91 PT92

PT93PT94

PT95

PT97

PT98PT99 PT00

PT01

PT02

PT03

PT04

PT05

PT06

PT07

PT08PT09

PT10 PT11

RO92

RO93

RO94

RO95

RO96

RO97

RO99RO01

RO02

RO03RO04

RO05RO06RO07

RO08

RO09

RO10

RO11

RU92RU93

RU94RU95RU96

RU99

RU00

RU04RU08

SN02

SN05

SN08

SK93SK94

SK95

SK96SK97

SK99

SK01SK02

SK03

SK04

SK05

SK06

SK07

SK08

SK09

SK10

SK11

SI92

SI93

SI94SI95

SI96

SI97

SI99SI01

SI02

SI03SI04SI05

SI06SI07

SI08

SI09SI10

SI11

ZA00

ZA01

ZA02ZA06

ZA08

ZA09ES85ES86 ES87

ES88

ES89

ES90

ES91

ES92

ES93 ES94

ES95

ES96

ES97

ES98

ES99ES00

ES01ES03

ES04

ES05ES06

ES07

ES08

ES09

ES10

ES11

SR10

SE95SE97

SE98

SE99

SE00SE01

SE02

SE03

SE04

SE05SE06

SE07

SE09SE10

SE11

CH99CH03

CH07CH08

TW96

TW01

TW04

TW06

TW08

TH01

TH02TT10

TR01

TR02TR03

TR04

TR05

TR06TR07

TR09

TR10

TR11

UA92

UA93

UA94

UA95

UA96

UA98

UA99

UA00UA04 UA08

GB73

GB76GB77

GB78

GB79

GB80

GB81

GB82

GB83GB84

GB85GB86GB87

GB88

GB89GB90

GB91GB92

GB93GB94

GB95GB97

GB98GB99

GB00

GB01

GB03

GB04

GB05

GB06GB07

GB09GB10GB11

US96

US99

US04

US06

US08US10

UY95 UY96UY97

UY98

UY00UY01

UY02

UY03

UY04

UY05

UY06

UY07

UY08

UY09

UY10

VE95

VE96VE97

VE98

VE00 VE01

VE02

VE03

VE04VE05

VE06

VE07

VE08ZM99 ZM03

ZM05

ZM09ZW09

ALB

ARG

ARM

AUSAUT

BGD

BLRBEL

BEN

BOL

BWA

BRA

BGR

BFA

KHM

CAN

CPV

CHL

COL

CRI

HRVCYP

CZEDNK

DOM

ECU

SLV

EST

FIN

FRA

GEO

DEU

GHA

GRC

GTM

GUY

HTI

HND

HUNIND

IDN

IRN

IRL

ISR

ITA

JAMJPN

KEN

KOR

KGZ

LVA

LSO

LBR

LTU

MKD MDG

MWI

MYS

MLI

MEX

MDAMNG

MOZ

NAM

NLD

NZL

NIC

NGA

NOR

PAN

PRY

PER

PHL

POLPRT

ROURUS

SEN

SVK

SVN

ZAF

ESP

SUR

SWE

CHE

TWNTHA

TTO

TUR

UKR

GBR

USA

URY

VEN

ZMB

ZWE

3545

5565

7585

Pola

rizat

ion

40 50 60 70 80 90Grievances

15

satisfied” in the latter, while over 15% of respondents expressed dissatisfaction. As a

consequence, Ghana in 2008 perhaps best represents the majoritarian distribution. In contrast,

while average levels of grievance are equally low in the cases of Georgia in 1994 and Moldova

in 1996, there is much greater polarization in opinions in the former than in the later. A sizeable

minority of Georgians (20%) are actually satisfied or very satisfied with democratic

performance, while 55 percent are very unsatisfied, indicating an oligarchic distribution. On the

other hand, there is significantly more accord among Moldovans in 1996, as over two thirds said

they were unsatisfied, and less than 7 percent expressed satisfaction.

FIGURE 3: EMPIRICAL EXAMPLES OF IDEAL TYPE DISTRIBUTIONS

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

Very Unsatisfied Unsatisfied Satisfied Very Satisfied

Perc

enta

ge R

espo

nse

Norway 1997 Ghana 2008 Georgia 1994 Moldova 1996

16

IV. ADDITIONAL DATA AND METHODS

Beyond the polarization and grievance measures described above, the main explanatory

variable in the analysis is economic inequality, which we hypothesize to predict polarization and

grievance, but to have a negligible impact on protest once we control for the latter two variables.

Our measure of inequality is the gini coefficient for income net taxes and transfers included in

the Standardized World Income Inequality Database (Solt 2009), which compiles distributional

data from a variety of sources. To take into account the uncertainty of the estimates, the

inequality dataset provides 100 imputed estimates for each country year. As a result, all models

were run using multiple imputation techniques based on the 100 imputations for each country

year provided in the inequality the SWIID database. Since both linear and curvilinear

relationships have been hypothesized (see above), all models include both the (mean centered)

gini coefficient and its quadratic.

Our measure of protest is based on the number of anti-government protests and strikes

recorded in a given year by the Cross National Time Series (CNTS) Data Archive (Banks 2011).

The event counts are based on coding of mentions in news sources, primarily the New York

Times. Details of the operationalization and measurement of these measures are available from

CNTS.

In terms of control variables, the literature on protest provides a plethora of potential

variables, including those associated with grievance theory as well as political opportunity

structure theories.12 Grievance theory models include the natural log of GDP per capita, growth

12 Many argue that changing economic circumstances in a nation are a key determinant of citizen protest, including inflation and the growth rate, which lead to a sense of grievance (Gurr 1970; Hendrix and Haggard 2014). Dissatisfaction with the provision of collective goods has also been tied to protest activity (Finkel and Muller 1998). Others stress instead the level of economic resources the average citizen possesses, where resources are thought to facilitate mobilization (e.g. McCarthy and Zald 1977),

17

in GDP per capita, and the natural log of inflation (inflation <0 set to 0). Political opportunity

structure models include dummy variables indicating if an executive or legislative election took

place, the level of democracy (Polity 2), regime durability, a measure of repression/political

terror, and the level of urbanicity. Additional time invariant variables included in the robustness

checks include the natural log of population, a dummy variable indicating presidential or

parliamentary systems, and ethnic fractionalization. The descriptive statistics of all of the

variables are available in Appendix A.

We present two sets of models. The first set of models examines the relationship

between economic inequality as an explanatory variable predicting two outcome variables –

citizen grievance (country mean) and citizen polarization (adjusted country standard

deviations).13 The second set of models predicts the incidence of nonviolent political protest as a

function of income inequality, grievance, and polarization. The first set of models include linear

models with fixed effects as well as lagged dependent and independent variables. 14 Since the

protest variable is a count variable, we report negative binomial regressions for the protest

models.

or where resources increase an individual’s willingness to pay the costs of protest activity (Cicchetti, Freeman, and Haveman 1971). Rather than this focus on the level and change in a country’s economic conditions, the “political opportunity structure” approach to explaining protest activity emphasizes that institutional structures and political processes influence the level of protest activity by establishing a level of political “openness” (Dalton et al. 2009; Vrablikova 2014). Consistent with this general orientation, studies have found that party identification is a predictor of protest activity (Finkel and Opp 1991). In contrast, others argue that protest activity emerges precisely when governments attempt to stunt conventional channels of participation (Kitschelt 1986). 13 Both the grievance and polarization variables are mean centered since in the protest models we include interactions between these two variables. 14 Hausman tests suggest that in all cases fixed effects models are preferred to random effects specifications.

18

V. RESULTS

Table 1 presents the results of the models predicting grievance and polarization. Income

inequality is significantly related to both grievance and polarization, with a strong positive effect

of inequality on polarization and a curvilinear effect of inequality on grievance. Since the

inequality variables are mean centered, the significant coefficient for inequality squared in the

grievance model indicates that the positive association with grievance accelerates as inequality

increases. The lagged dependent variables, poor economic growth, high levels of inflation,

political terror/repression, and regime durability are also predictive of high levels of grievance

and polarization.

TABLE 1: PREDICTING GRIEVANCE AND POLARIZATION

Grievance Polarization Grievance (t-1) 0.374** (0.037) Polarization (t-1) 0.210** (0.039) GINI (t-1) 0.247** (0.082) 0.154* (0.07) GINI^2 (t-1) 0.011* (0.005) 0.000 (0.005) GDP/cap (t-1) -1.484 (1.567) 0.428 (1.388) Growth (t-1) -0.174** (0.04) -0.119** (0.036) Ln Inflation (t-1) -0.429* (0.206) -0.581** (0.189) Legislative Election (t-1) 0.350 (0.34) -0.203 (0.308) Executive Election (t-1) -0.487 (0.496) -0.365 (0.446) Polity (t-1) 0.427+ (0.22) -0.037 (0.189) Regime Durability (t-1) -0.155** (0.05) -0.159** (0.045) Political Terror (t-1) 1.071** (0.333) 0.214 (0.299) Urbanicity (t-1) 0.149+ (0.09) 0.049 (0.08) Constant 2.872 (15.436) -1.725 (13.729) N 755 755 + denotes p<0.10; * p<0.05; **p<0.01. Both models include fixed effects. Standard errors in parentheses. Significance stars are based on two-tailed tests.

Table 2 reports models predicting protest, with Model 1 including income inequality and

the control variables but not grievance and polarization. Against most existing theories, yet

19

consistent with others, the model suggests that inequality is negatively and significantly related

to protest. The addition of the grievance and polarization variables in Model 2 suggests that what

is most important about the grievance/protest relationship is not the mean level of grievance but

rather the dispersion of such grievances. The coefficients reveal that grievance is unrelated to

protest, but polarization of grievance is positively related to protest (p<0.05, one tailed test). As

polarization increases, the likelihood of protest also increases.

TABLE 2: PREDICTING POLITICAL PROTEST Model 1 Model 2 Model 3 Protest 0.075+ (0.046) 0.084+ (0.046) 0.086+ (0.046) Grievance 0.009 (0.018) 0.024 (0.020) Polarization 0.037+ (0.023) 0.046* (0.024) Grievance*Polarization -0.004+ (0.002) GINI -0.117** (0.044) -0.123** (0.045) -0.130** (0.044) GINI^2 0.003 (0.003) 0.002 (0.003) 0.003 (0.003) GDP/cap 0.768 (0.966) 1.018 (0.981) 1.721+ (0.888) Growth -0.054* (0.022) -0.046* (0.022) -0.062** (0.024) Ln Inflation 0.051 (0.112) 0.070 (0.112) 0.147 (0.108) Legislative Election 0.116 (0.197) 0.095 (0.197) 0.259 (0.195) Executive Election -0.124 (0.284) -0.094 (0.286) -0.284 (0.287) Polity 0.004 (0.082) -0.002 (0.085) -0.005 (0.066) Regime Durability -0.015 (0.024) -0.012 (0.024) -0.007 (0.021) Political Terror -0.245 (0.170) -0.257 (0.171) -0.235 (0.172) Urbanicity -0.129* (0.057) -0.123* (0.057) -0.120* (0.055) Constant -0.023 (8.076) -2.686 (8.222) -4.556 (8.242) N 897 897 897 + p<0.10 * p<0.05 **p<0.01. Cell entries report the results of fixed effect, negative binomial egressions. Standard errors in parentheses. Significance stars are based on two-tailed tests.

20

Gurr’s (1970) theory also suggests that we would expect an interactive relationship

between grievance and polarization on the one hand and protest on the other. According to

grievance theory, we would expect the highest level of protest when grievances are most

widespread (high average grievance) and there are large grievance differences across groups

(high polarization). That is, adopting our language above we would expect to observe the most

protests in oligarchic contexts. Model 3 tests this hypothesis and points to a different conclusion.

That is, the model predicts that protest is actually most frequent in majoritarian contexts, or

when average grievance is fairly low, but there is significant polarization in such opinions. That

is, protest is most likely when many members of society are well served by the political system

but a significant minority very clearly does not feel well served.

To illustrate this relationship more clearly, Figure 4 plots the predicted probability of at

least one protest event across three different levels of grievance and a range of polarization

values. In countries with high average levels of grievance, protest is essentially unrelated to

polarization, with the predicted probability of at least one protest never more than 20% and

actually declining modestly as polarization increases. On the other hand, while the probability of

protest is quite low among polities with low to moderate levels of grievance and low levels of

polarization (inclusive distributions), the predicted probability of protest increases significantly

among more majoritarian distributions. At high values of polarization (80) and satisfaction (40),

the model predicts that there is a 78 percent probability of at least one protest. This relationship

perhaps suggests that the existence of aggrieved minorities who have doubts about the efficacy

of more conventional participation channels are most likely to protest.

21

FIGURE 4: PREDICTED PROBABILITY OF AT LEAST ONE PROTEST, BY LEVELS OF POLARIZATION AND GRIEVANCE

VI. Conclusion

To recapitulate our findings, we have observed that countries with higher levels of

income inequality also tend to have more polarized citizens. Moreover, countries with polarized

citizens are more likely to experience nonviolent protest. Finally, protests are especially (indeed

most) likely in countries where average citizen satisfaction is high but satisfaction is also

polarized. These findings are new, and they prompt a number of additional puzzles.

First, the analysis suggests that while inequality leads to both greater levels of grievance

and polarization, and that polarization and grievance are related to protest, the negative

relationship between inequality and protest suggests that it is unclear how perceived grievances

mediate the inequality/protest relationship. Indeed, the negative and significant coefficient

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

30 40 50 60 70 80 90

Pred

icte

d Pr

obab

ility

of 1

or m

ore

prot

ests

Polarization

Low Grievances Moderate Grievances High Grievances

22

estimates for the effect of inequality on protest are hardly changed across the models, suggesting

an effect independent of perceived grievances and variation in such grievances.

A possible explanation for this puzzling result could reflect the interaction between

grievances and polarization. In contrast to Gurr’s predictions, we found that protest is most likely

when most citizens are relatively satisfied, but when there are aggrieved minorities, that is, what

we term majoritarian systems. However, high levels of inequality are more likely to lead to what

we termed oligarchic systems in which the majority of citizens are dissatisfied with the political

system while a minority is satisfied. In direct contrast to Gurr, our findings suggest that such

systems are actually the least likely to see protest. This finding is consistent with other works

that suggest a demobilizing effect of inequality on participation (e.g., Solt 2006). Our

contribution is to illuminate a mechanism by which inequality leads to this result.

An important final implication of our finding that polarization leads to protest is that it

underscores the importance of political knowledge among citizens. Only when citizens are

aware of how well the democratic process works for them relative to how well it works for their

fellow citizens do they take to the streets.

23

VII. References

Acemoglu, Daron, Simon Johnson, James A. Robinson, and Pierre Yared. 2009. "Reevaluating

the Modernization Hypothesis." Journal of Monetary Economics 56(8): 1043-58.

Adams, James, and Lawrence Ezrow. 2009. "Who do European Parties Represent? How Western

European Parties Represent the Policy Preferences of Opinion Leaders." The Journal of

Politics 71(1): 206-23.

Anderson, Christopher J., and Christine A. Guillory. 1997. "Political Institutions and Satisfaction

with Democracy: A Cross-National Analysis of Consensus and Majoritarian Systems."

American Political Science Review, 91(1): 66-81.

Anderson, Christopher J., and Silvia M. Mendes. 2006. "Learning to Lose: Election Outcomes,

Democratic Experience and Political Protest Potential." British Journal of Political

Science 36(1): 91-111.

Banks, Arthur S. 2011. "Cross-National Time-Series Data Archive, 1815-[2011].".

Bartels, Larry M. 2008. "Unequal democracy." New York: Russell Sage Foundation.

Boix, Carles. 2003. “Democracy and Redistribution”. Cambridge, Cambridge University Press,

2003.

Booth, John A., and Mitchell A. Seligson. 2009. “The Legitimacy Puzzle in Latin America:

Political Support and Democracy in Eight Nations.” Cambridge University Press, 2009.

Chambers, John R., Lawton K. Swan, and Martin Heesacker. 2014. "Better Off Than We Know

Distorted Perceptions of Incomes and Income Inequality in America." Psychological

science 25(2): 613-18.

24

Cicchetti, Charles J., A. Myrick Freeman, Robert H. Haveman, and Jack L. Knetsch. 1971. "On

the Economics of Mass Demonstrations: a Case Study of the November 1969 March on

Washington." The American Economic Review, 61(4): 719-24.

Collins, William J., and Robert A. Margo. 2007. "The Economic Aftermath of the 1960s Riots in

American Cities: Evidence from Property Values." The Journal of Economic History

67(4): 849-83.

Delhey, Jan, and Ulrich Kohler. 2011. "Is Happiness Inequality Immune to Income Inequality?

New Evidence Through Instrument-Effect-Corrected Standard Deviations." Social

Science Research 40(3): 742-56.

Delli Carpini, Michael X., and Scott Keeter. 1996. "What Americans Don’t know About Politics

and Why it Matters." New Haven, Yale University Press.

Dalton, R., Van Sickle, A., & Weldon, S. 2009. “The Individual-Institutional Nexus of Protest

Behaviour.” British Journal of Political Science 40, 51 –73

Easton, David. 1965. “A Framework for Political Analysis. Vol. 25.” Englewood Cliffs, NJ,

Prentice-Hall.

Eriksson, K., & Simpson, B. (2012). What do Americans know about inequality? It

depends on how you ask them. Judgment and Decision Making, 7, 741-745.

Finkel, Steven E., and Edward N. Muller. 1998. "Rational Choice and the Dynamics of

Collective Political Action: Evaluating Alternative Models with Panel Data." American

Political Science Review, 92-(1): 37-49.

Finkel, Steven E., and Karl-Dieter Opp. 1991 "Party Identification and Participation in

Collective Political Action." The Journal of Politics 53(2): 339-71.

25

Gilens, Martin. 2005. "Inequality and Democratic Responsiveness." Public Opinion Quarterly

69(5): 778-96.

Gurr, Ted. 1970. "Why men rebel." Princeton, Paradigm Publishers.

Harmel, Robert, and John D. Robertson. 1986. "Government Stability and Regime Support: A

cross-National Analysis." The Journal of Politics 48(4): 1029-40.

Jacobs, Lawrence R., and Benjamin I. Page. 2005. "Who influences US foreign policy?."

American Political Science Review 99(1): 107-123.

Kitschelt, Herbert P. 1986. "Political Opportunity Structures and Political Protest: Anti-Nuclear

Movements in Four Democracies." British Journal of Political Science 16(1): 57-85.

Klingemann, Hans-Dieter, and Bernhard Wessels. 1999. "Political Consequences of Germany's

Mexed-Member System: Personalization at the Grass-Roots?."

Lichbach, Mark Irving, ed. 1996. “The Cooperator's Dilemma.” University of Michigan Press,

1996.

Lockerbie, Brad. 1993. "Economic Dissatisfaction and Political Alienation in Western Europe."

European Journal of Political Research 23(3): 281-93.

Lohmann, Susanne. 1993. "A Signaling Model of Informative and Manipulative Political

Action." American Political Science Review 87(2): 319-33.

Lichbach, Mark Irving. 1989. "An Evaluation of “Does Economic Inequality Breed Political

Conflict?” World Politics 41(4): 431-70.

Lohmann, Susanne. 1994. "The Dynamics of Informational Cascades: The Monday

Demonstrations in Leipzig, East Germany, 1989–91." World politics 47(1): 42-101.

26

Madestam, Andreas, Daniel Shoag, Stan Veuger, and David Yanagizawa-Drott. 2013 "Do

Political Protests Matter? Evidence from the Tea Party Movement*." The Quarterly

Journal of Economics: 21(2):1-59

McCarthy, John D., and Mayer N. Zald. 1977. "Resource Mobilization and Social Movements: A

partial Theory." American Journal of Sociology, 86(2): 1212-41.

McDonough, Peter, Samuel H. Barnes, and Antonio López Pina. 1986. "The Growth of

Democratic Legitimacy in Spain." The American Political Science Review 80(3): 735-60.

Nagel, Thomas. 1974. "What is it like to be a Bat?" The Philosophical Review 83(4): 435-450.

Norris, Pippa. 2011 “Democratic Deficit: Critical Citizens Revisited.” Cambridge University

Press, 2011.

Norton, Michael I., and Dan Ariely. 2011. "Building a Better America—One Wealth Quintile at

a Time." Perspectives on Psychological Science 6(1): 9-12.

Russett, Bruce M. 1964. "Inequality and Instability: The Relation of Land Tenure to Politics."

World Politics 16(3): 442-54.

Sigelman, Lee, and Miles Simpson. 1977. "A Cross-National Test of the Linkage Between

Economic Inequality and Political Violence." Journal of Conflict Resolution, 21(1): 105-

28.

Skocpol, Theda, and Vanessa Williamson. 2011. "The Tea Party and the Remaking of American

Conservatism." New York, The Oxford University Press.

Solt, Frederick. 2009. "Standardizing the World Income Inequality Database."Social Science

Quarterly 90(2): 231-42.

Toka, Gabor. 1995. "Political Support in East-Central Europe." Citizens and the State: 354-82.

Tyler, Tom R. 2006. “Why People Obey the Law”. Princeton University Press, 2006.

27

Verba, Sidney. 1985. “Equality in America: The View From the Top”. Harvard University Press,

1985.

Widmaier, Ulrich. 1990. "Political Stability in OECD Nations." International Political Science

Review 11(2): 219-42.

28

VIII. Appendix A: Summary Statistics

N Mean St. Dev. Min Max Grievance 1042 64.93581 8.526368 40.96777 87.40012 Polarization 1042 53.85091 6.258634 36.36218 84.86587 Gini (average imputation) 1007 35.26988 9.63852 19.3986 69.10886 Protest 1015 0.4157635 1.213236 0 22 Ln GDP/cap 1008 9.354813 0.930272 5.703587 10.85082 Growth 1030 2.776054 4.790745 -44.9 18.28661 Ln Inflation 1030 1.727786 1.322055 0 8.706424 Legislative Election 1033 0.2904163 0.4541744 0 1 Executive Election 1034 0.1276596 0.3338719 0 1 Polity2 1042 8.692898 1.792215 0 10 Regime Durability 1042 30.59021 31.78594 0 201 Political Terror 1020 1.903922 0.9839658 1 5 Urbanicity 1037 67.58535 15.04106 13.444 97.4854