ArticlePDF Available


Mason Moseley1 y Kyu Chul Shin2
The contentious politics literature has long been divided on the
extent to which grievances –or “dissatisfaction caused by depriva-
tion” (Dalton et al., 2009)– drive citizen participation in protests.
Do grievances motivate citizens to take to the streets? To shed
light on how grievances affect protest, we focus on citizen evalu-
ations of public service provision in Latin America. Scant research
has examined the effect of poor public service delivery on con-
tentious participation in emerging democracies. We highlight two
mechanisms associated with public service evaluations that facili-
tate mobilization: 1) firsthand experience with poor governance
and 2) clear attribution of responsibility for poor service provi-
sion. To test our argument, we utilize data from the 2012 and 2014
AmericasBarometer national surveys of Brazil, and then generalize
to Latin America in multilevel models of protest drawing from 18
countries. The results are consistent: where firsthand experience
with state incompetence fuels declining system support and spe-
cific attribution of blame for underperformance, as in the case of
public service evaluations in Latin America, grievances fuel par-
ticipation in protest.
Keywords: protest, public services, grievances, contentious poli-
tics, social movements
As rates of protest participation have risen across the world
(e.g., Dalton et al., 2009; Beaulieu, 2014; Moseley, 2018),
1 West Virginia University.
2 West Virginia University.
the contentious politics literature has grappled with the rel-
ative importance of grievances in mobilizing collective ac-
tors. On the one hand, many accounts of protest events em-
phasize the economic motivations of claimants –e.g., youth
unemployment in the Arab Spring (Campante and Chor,
2012) or state retrenchment in Latin America (Silva, 2009)
during the neoliberal era. Yet, at the same time, empirical
studies have found little evidence of grievance-motivated
contention, highlighting instead the role of political oppor-
tunities and organizational resources as the key determi-
nants of successful mobilization (e.g., Dalton et al., 2009;
Machado et al., 2011; Arce, 2014; Boulding, 2014). More-
over, several studies have even attributed swelling rates of
contention to economic progress (e.g., Moseley, 2018; Mu-
rillo and Mangonnet, 2016), thus offering accounts that are
diametrically opposed to grievance-based interpretations of
Against the backdrop of these apparent contradictions
in the existing literature, we ask the following questions: To
what extent do grievances motivate citizens to take to the
streets? Further, do evaluations of public service delivery,
in particular, help explain the recent explosion of mass pro-
tests across Latin America? Relatively little research has ex-
amined the role of poor public service delivery in spurring
contentious participation.3 In focusing on public service pro-
tests in Latin America, we highlight two characteristics of
this particular grievance that we argue facilitate mobiliza-
tion: 1) firsthand experience with poor governance, which
undermines faith in the political system, and 2) clear attri-
bution of responsibility for poor service provision, which
supplies citizens with targets for protest. Throughout much
of Latin America, economic progress under recent center-
left governments has resulted in expanded access to public
services, yet in many cases their quality remains lackluster
3 One partial exception would be work on the case of South Africa (e.g., De Juan
and Wegner, 2017).
(e.g., Levy and Schady, 2013). We argue that citizens con-
nect public service quality to elected officials’ performance
and, through their firsthand experience with those services,
are more motivated to seek redress than they would be if
not due to that direct experience.
To test our argument, we utilize data from the 2012 and
2014 AmericasBarometer surveys, conducted by the Latin
American Public Opinion Project (Lapop). The AmericasBa-
rometer surveys include not only a special battery of ques-
tions to gauge satisfaction with the quality of public health
services, roads, and schools, but also contain an item on
participation in recent protest demonstrations. We begin
with a case study of Brazil, where protests erupting in 2013
have inspired numerous grievance-based explanations. In
the analysis of survey data from Brazil, we find support for
our central expectation that public services evaluations are
strongly related to protest participation. Then, we general-
ize our argument to 18 Latin American democracies. Finally,
we offer a test of the mechanisms we argue connect public
service grievances and protest –namely, the role firsthand
experience with poor governance plays in fueling declining
trust in the political system and increased attribution of re-
sponsibility for poor governance. The results are consistent:
in the case of public service evaluations, where individu-
als can credibly assess blame for poor governance based on
firsthand experience, grievances can fuel citizen participa-
tion in protest.
Grievances and Protest: The Existing Literature
As long as protest has existed, so too have scholars en-
deavored to understand why some individuals go to great
lengths to give voice to their claims, while others choose
to abstain. Yet the study of protest behavior became mod-
ernized and scientific following World War II (Wilkinson,
2009). One of the most important modern breakthroughs
in explaining protest behavior emerged in the 1960s and
stressed the importance of solving collective action prob-
lems. While previous scholarly work had mostly assumed
that shared interest was enough to gather motivated ac-
tors together, Olson (1965) countered that individuals of-
ten have an incentive to free ride on the efforts of others
because they can enjoy the benefits of coordinated action
without contributing. In the absence of selective incentives,
it is challenging to mobilize a group of individuals not just
because of the costs of group formation, but the desire to
free ride.
Eventually, two schools emerged that attempted to solve
the collective action problem, forming a modern theoretical
dichotomy in the protest literature:1) grievance-based the-
ories and 2) structural theories. The relationship between
grievances, defined here as “dissatisfaction caused by de-
privation” (Dalton et al., 2009; 6) and protest was most fa-
mously advanced by Ted Gurr in his classic work Why Men
Rebel (1970). Gurr argues that in contexts characterized by
high levels of poverty and low standards of living, protest
and other forms of contentious activity are more likely (Gurr,
1970). Leaning on research from psychology, he claims that
where the gap between citizens’ expectations and the real-
ity of their circumstances becomes too wide (i.e., “relative
deprivation”), they lash out via more contentious modes of
political behavior.
For much of the 1970s, grievance-based approaches
predominated (van Stekelenburg and Klandermans, 2013).
The logic makes intuitive sense: individuals living in con-
ditions that fail to meet their expectations should be more
likely to engage in contentious behavior because they are
frustrated with their current situation. Yet soon after the
emergence of grievance-based arguments, a competing
narrative emerged. Structural theories, in contrast to Gurr’s
argument, note that protests require individuals to have
the necessary organizational resources and political know-
how to mobilize contention, and these skills are often most
highly developed in the affluent (Dalton et al., 2009). Re-
source mobilization theory thus implies that in many cas-
es where grievances are present protests fail to materialize,
and, on the other side of the coin, relatively minor griev-
ances can balloon into massive protest movements when
harnessed by skilled “political entrepreneurs” (McCarthy
and Zald, 1977). According to this perspective, the pres-
ence of educated citizens and motivated elites who are
sympathetic to a specific cause is more important for the
success of a movement than poor living conditions. To sup-
port this claim, McAdam (1982) famously noted that Afri-
can-American protests during the Civil Rights Movement
only started to gain momentum after sympathetic Northern
elites and black activists, fueled by an emerging African-
American middle class and the growth of historically black
universities, joined forces in the 1960s. In their cross-na-
tional study of the determinants of protest using data from
the World Values Survey, Dalton et al. (2009) conclude,
“without the resources and skills to become politically en-
gaged… grievances are typically not translated into politi-
cal action” (22).
The shift from grievance-based arguments to those fo-
cused on resources is also reflected in the literature on Latin
America. While in the past scholars have argued that pov-
erty and joblessness, particularly following the implemen-
tation of neoliberal reforms in the 1990s (Walton and Ra-
gan, 1990; Silva, 2009), were a driving force in mobilizing
protestors, much has changed in the region. In the past de-
cade, Latin America has witnessed a rise in living standards
with an increasing number of individuals joining the mid-
dle class. Moreover, in the 2000s many Latin American na-
tions elected leftist and center-left politicians to the pres-
idency in a wave known as the “pink tide” (Graham and
Smith, 2012). Promising to direct more funding towards
public goods provision, these leaders increasingly invest-
ed in programs such as health care and education, incorpo-
rating previous political “outsiders” in unprecedented ways
(Garay, 2016). Yet, even in this era of relative economic
prosperity and democratic stability, protest rates have actu-
ally risen across the region (Mangonnet and Murillo, 2016;
Moseley, 2018), presenting an important challenge to griev-
ance-based theory.
At the same time, anecdotal evidence suggests it might
be time to revisit the grievance thesis in Latin America. Re-
cent upheaval in Brazil has widely been attributed to ram-
pant corruption, high crime rates, and a sagging economy.
The same goes for Venezuela, where a severe economic re-
cession, triggered in the eyes of many by irresponsible fis-
cal policies and declining oil prices, has coincided with un-
precedented rates of street-based participation. In Chile,
a country characterized by relatively low rates of conten-
tion since democratization, student-led protests for educa-
tional reform have grabbed headlines since 2011. In short,
journalistic accounts of recent protests across Latin Amer-
ica suggest some role for grievances –however, the ques-
tion remains of just how much grievances matter, and un-
der what conditions they propel citizens to the streets.
How Public Service Evaluations Fuel
Protest Participation
Despite an extensive literature documenting the relatively
weak relationship between grievances and collective ac-
tion, we argue that perceptions of public service quality mo-
tivate protestors. This is based on two primary mechanisms:
1) poor public service provision offers citizens firsthand ex-
perience with poor governance, which erodes support for
the political system, and 2) provides them with ammunition
to blame high profile policymakers for substandard service
quality. We address each step of this process in turn.
Declining System Support through Firsthand Experience
with Poor Governance
Scholars in political science and sociology have long puz-
zled over the correlation between personal experience with
grievances and protest participation. Most theoretical work
has revolved around individuals’ ability to connect their
own personal circumstances to larger societal problems.
Studies have found that where individuals have person-
al experience with motivating claims, they are more moti-
vated to take to the streets, so long as they connect those
personal experiences to larger group concerns (e.g., Cros-
by, 1976) –a scenario that Foster and Matheson (1999) dub
“double relative deprivation,” in that it affects both person-
al and group interests. For certain grievances, like discrimi-
nation, the linkage between personal experience and wid-
er systemic injustices is not always apparent (Crosby et al.,
1986). However, in the case of public service grievances,
firsthand experience with poor governance provides clear
evidence that highlights the state’s general inability to sat-
isfy its basic obligations to citizens.
Where personal grievances are clearly connected to
poor governance, they can affect larger perceptions of the
political system, and drive participation in politics. For ex-
ample, Seligson (2006) argues and finds that experience
with corruption has a stronger impact on regime legitimacy
than perceptions of corruption. In her article on crime vic-
timization and political participation, Bateson (2012) finds a
nearly universal positive effect for experience with crime on
the likelihood that individuals take part in elections, local
community activities, and protest –in many cases, victims
“seek assistance from elected officials or lobby for policy
changes that are narrowly related to the crimes they have
suffered” (571). Gingerich (2009) finds that experience with
corruption pushes individuals to protest, implying a proac-
tive response fueled by firsthand knowledge of public offi-
cials’ involvement in political graft. Finally, in their study of
South Africa, De Juan and Wegner (2017) uncover evidence
that perceptions of public service quality are tightly corre-
lated with general trust in government, and thus fuel pro-
test. Given the tight correlation between support for the po-
litical system and protest participation (Moseley, 2018), we
view the potential role that public service grievances play
in diminishing such support as critical to understanding the
connection between public service evaluations and protest
We argue that firsthand knowledge of poor gover-
nance can erode faith in core political institutions, and al-
so provide the psychological jolt necessary to spur con-
tention. That is, where mere perceptions of government
performance can seem distant, actually suffering the con-
sequences of such inadequacies carries a stronger impact
on the likelihood that aggrieved individuals will take ac-
tion against the government. In some ways, this argument
seems at odds with findings from the economic voting liter-
ature, which has generally uncovered a powerful associa-
tion between vote choice and sociotropic evaluations of the
country’s economy than pocketbook considerations (e. g.,
Markus, 1988) –in other words, a stronger effect for percep-
tions than firsthand experience. The difference, in our view,
is that overcoming the significant costs associated with
protesting requires extra motivation on the part of citizens,
making it a fundamentally distinct act from voting. Where
individuals have firsthand knowledge of the state’s incom-
petence in providing quality healthcare services, schools,
and roads, we argue they will be more positively inclined
to take direct action against the government, overcoming
traditional barriers to protest.
In the case of public service provision, a child’s bad ex-
perience with resources at a local school or treatment at
the public health clinic should thus carry more weight for a
parent’s decision to protest than reading about such inad-
equacies in the paper. The more tangible the grievance, the
more likely we argue it produces a meaningful behavioral
response. Yet as others have noted (Foster and Matheson,
1995; Walker and Mann, 1987), not all personal grievances
translate into political action –in our view, that direct con-
nection only heightens the probability that individuals will
take action. One of the principal contributions of this study,
then, is to highlight the role that firsthand knowledge of
poor service quality can have in diminishing support for the
political system at large, and fueling potential anti-govern-
ment action.
Figure 1.
Predicted Eects of Public Service Evaluations
on System Support
Support for the Political System
95% CI Fitted values
0 20 40 60 80 100
Perceptions of Public Service Quality
In terms of observable implications, we should therefore
find that public service evaluations are positively associat-
ed with support for the political system. This expectation is
borne out in Figure 1.4 Building on previous research (Mose-
ley, 2018), we know that low levels of support for the politi-
cal system are associated with an increased willingness to
confront public officials for their ineffectiveness. Therefore,
the declines in system support associated with negative
evaluations of public service quality serve as a key mecha-
4 Figures 1 and 2 are graphs of linear predictions from estimated OLS regres-
sion models of system support and approval of anti-government protests on
public service evaluations. “System support” comes from Booth and Seligson’s
(2009) measure of support for key regime institutions. See Appendix for ques-
tion wording.
nism in explaining the relationship between public service
grievances and protest. In our view, not all grievances are
experienced in tangible ways by a broad enough swath of
citizens to motivate protest. For example, mere perceptions
of insecurity or poor economic conditions –widely consid-
ered to have motivated recent mobilization in Brazil (Simões,
2013)–without direct experience with those grievances
might not create a strong enough trigger to actually incite
action against the government. Our intention in this paper,
then, is not only to demonstrate that public service evalua-
tions are correlated with protest participation, but that one
key mechanism by which this relationship exists lies in the
effect of experience with poor governance on support for
the political system.
The Downside of Credit-Claiming:
Blaming Politicians for Poor Public Services
There is bountiful evidence that firsthand experience with
poor governance can provide the additional motivation in-
dividuals need to translate grievances into action. Yet that
increased “protest potential” requires a target for collective
action. In the case of public service grievances, we argue
the chain of responsibility is readily apparent to aggrieved
To this point, much of the literature on blame attribution
has focused on performance voting, proposing conditions
under which voters will punish politicians for poor economic
performance, in particular. A number of studies have brought
empirical evidence to bear on the question of blame attribu-
tion in different institutional contexts. Powell and Whitton
(1993) were the first to argue that economic voting is con-
tingent on the degree to which “clarity of responsibility” for
past economic performance is readily accessible to voters.
Anderson (1995) finds that economic voting is suppressed
in countries led by coalition governments, as citizens find
it more difficult to assign responsibility for economic under-
performance to one specific party. In their comprehensive
treatment of the relationship between institutional settings
and economic voting, Duch and Stevenson (2008) discov-
er that where there is a high degree of state control over
economic policy, voters are more likely to punish incum-
bents for perceived mismanagement (see also Carlin and
Singh, 2015). Likewise, scholars have found that in federal
systems, voters face a more difficult task in terms of blame
attribution, balancing the relative culpability of provincial
and national-level politicians (Atkeson and Partin, 1995).
Yet in the contentious politics literature, clarity of re-
sponsibility had not been thoroughly examined until Jave-
line’s (2003) study of protest behavior in Russia in the
late1990s. By this point, grievance-based arguments had
lost much of their cachet, and were often considered sec-
ondary to structural arguments. For Javeline, specificity in
terms of blame attribution “reduces the costs of informa-
tion, organization, and opportunity” (109), and thus makes
the mobilization of contention more feasible. In her analysis
of the wage arrears crisis in Russia, individuals who were
able to make specific attributions of blame were five times
more likely to protest than those who offered nebulous cri-
tiques. Further, she finds that political entrepreneurs can
most efficiently mobilize individuals who offer non-specific
critiques of public figures, given the already-high likelihood
that citizens who offer specific attribution of blame will pro-
test. Thus, “some grievances may compel the use of more
resources than others to inform or convince members about
blame successfully” (Javeline, 2003; 119).
There is some reason to believe that grievances have
played a role in the normalization of protest in the region
(Moseley, 2018), in part fueled by increasing clarity of re-
sponsibility. While in the recent past scholars have argued
that poverty and joblessness, particularly following the
implementation of neoliberal reforms in the 1990s (Silva,
2009), were a driving force in mobilizing protestors, much
has changed in the region. Latin America has witnessed re-
cently a rise in living standards with increasing numbers
of individuals leaving lower socioeconomic brackets and
joining the middle class, which has been fueled in part by
an extension of the welfare state (Garay, 2016). One of the
most important innovations has been the conditional cash
transfer program, which presidents across the region have
utilized not only to alleviate extreme poverty, but to boost
their support at election time (Layton et al., 2017).
Indeed, many Latin American politicians have attempt-
ed to bolster public service provision in an explicit effort
to woo voters. During the 1990s, Brazilian President Fer-
nando Henrique Cardoso launched a new program, Cartão
Cidadão, designed to facilitate transferring social welfare
benefits. This card was introduced months before an elec-
tion, and represented an attempt to claim credit for trans-
fers (Garay, 2016). His successor, Lula da Silva, launched
the famous Bolsa Familia program in 2004, and reaped the
electoral rewards when he was reelected in 2006. In Ar-
gentina, Néstor Kirchner launched policies negotiated with
social movement leaders in direct response to their de-
mands for improved pensions and health care services (Ga-
ray, 2007). The Argentine government also enacted univer-
sal child allowances during a time in which poverty rates
happened to be lower than in previous years, with at least
the partial objective of taking credit for benefits that would
placate social movements and stop protests (Garay, 2016).
In all of these cases, politicians endeavored to highlight
their role in providing citizens with new and/or improved
public services.
In direct contrast to Gurr’s argument regarding pover-
ty and contentious behavior, the significant gains made by
large swaths of the population have not mitigated, but rath-
er coincided with an increase in the number of those who
have opted to take to the streets in places such as Argen-
tina, Brazil, and Chile (Moseley, 2018; Mangonnet and Mu-
rillo, 2016). In our view, this can be traced in part to the
extension of the Latin American safety net, which in turn
made more salient the connection between politicians (es-
pecially presidents) and public services. When many ser-
vices were provided by the private sector, or presidents ne-
glected to draw the connection between policy inputs and
outputs, citizens found it more difficult to apportion blame
for perceived underperformance. But now, those connec-
tions have been made clearer.
Figure 2.
Predicted Eects of Public Service Evaluations
on Presidential Approval
Presidential Approval
95% CI Fitted values
0 20 40 60 80 100
Perceptions of Public Service Quality
If public service grievances highlight issues related to poor
governance, and citizens blame prominent politicians for
said underperformance, then this could be a second path-
way through which public service evaluations affect pro-
test participation. Indeed, there appears to be a strong cor-
relation between perceptions of public service quality and
presidential approval (Figure 2), lending some credence
to the notion that citizens blame prominent politicians for
perceived state incompetence. The implication here is that
grievances associated with a clear chain of responsibility
–e.g., the quality of public services or corruption (see Ta-
vits, 2007)– should be characterized by stronger predictive
power in explaining protest participation than grievances
that are less easily traced to their origins –e.g., pocketbook
economic evaluations (see Kinder and Kiewet, 1981; also,
Gomez and Wilson, 2001). We should also expect that the
former class of grievances should be more strongly corre-
lated with evaluations of politicians than the latter group.5
The extent to which public service grievances reflect poor-
ly on elected officials, then, would seem to be an important
mechanism by which evaluations of service quality trans-
lated into higher rates of protest participation.
A Grievance-Based Theory of Public Service Protests
Since Gurr (1970), much of the theory-building within the
protest literature has shifted away from grievance-based
arguments, focusing instead on access to organizational
tools. There is no doubt that resources such as education,
income, and social capital matter a great deal in determin-
ing who engages in protest. However, the shift away from
grievance-based arguments has potentially obfuscated the
importance of key variables. We argue that certain griev-
ances have actually played an important role in mobilizing
Latin American demonstrators in the 2010s. Specifically, the
connection between the state’s inability to provide satis-
factory public services, such as health care, transportation,
and education, and contentious behavior among citizens, is
one that we seek to explore further here.
We argue that in the case of public service evaluations,
the conditions are ripe for translating grievances into ac-
tion. Evaluations of public services are particularly apt moti-
vators of protest due to the combination of direct experience
with poor governance, and the increasingly clear connec-
5 This probably depends on other factors, including the salience of the issue at that
moment in time (Epstein and Segal, 2000), or individuals’ awareness about pol-
itics (Zaller, 1992; Gomez and Wilson, 2001). In the case of grievances related to
public service provision, we would thus expect such evaluations to be strongly
correlated with presidential approval, and even more so among the politically
interested during times of heightened grievance salience.
tion between politicians and service quality. The vast ma-
jority of Latin Americans believe that the state should take
an active role in the educational and health sectors to re-
duce inequality (Lapop, 2012-2014), which places most Lat-
in American countries far above countries like the U.S. and
even Canada in terms of preferences for state intervention.
With an increasing number of Latin American countries of-
fering near-universal health coverage (World Bank, 2013)
and access to public education (World Bank, 2017), not to
mention citizens’ automatic interaction with roads and
other aspects of state infrastructure, the majority of Latin
Americans have direct experience with these shortcomings
in terms of public service delivery.
While Latin Americans have borne witness to increased
access to public services like healthcare and education,
the quality of those services has often lagged behind ex-
pectations. Due in part to a lack of tax revenue vis-à-vis
OECD countries (Zovatto, 2015; Melguizo, 2017), Latin Amer-
ican countries often struggle to adequately fund education,
healthcare, and infrastructure. As of 2011, millions of Latin
American children were still not enrolled in school, and a
preponderance of primary and secondary public schools are
chronically underfunded (Unesco, 2013). Over half of citi-
zens in most Latin American countries offer negative eval-
uations of the services available in their countries (Lapop,
2012-2014). Bribery at the point of delivery has also prov-
en to be an intractable problem in many Latin American
countries, with as many as one-third of those who accessed
public services reporting paying a bribe to do so (Lapop,
2012-2014; Transparency International, 2017). Given the
importance of public services in reducing inequality and
fighting poverty, the current shortcomings of the public sec-
tor in Latin America could undermine future social and eco-
nomic progress (Oxfam, 2017). Further, local observers have
pointed to poor public service provision as a key motivator
in recent episodes of contention from transportation in Bra-
zil (Alves, 2013) to education in Chile (Long, 2011).
The general expectation, then, is that when individ-
uals believe the state offers high quality public services,
they should be less likely to engage in contentious behav-
ior on average. When a government is perceived as inef-
fective in its duties, contentious behavior is more likely to
occur. On the issue of public service delivery, citizens can
apportion blame and have firsthand knowledge of state in-
competence. Particularly within the context of Latin Ameri-
ca, politicians have endeavored to connect their fate to ex-
panded state services. Further, virtually all Latin American
citizens have direct experience with some facet of public
service provision, unlike perceptions of crime, corruption,
or national economic performance, with which many citi-
zens have not interacted firsthand. We argue this firsthand
knowledge can erode support for political system, and pro-
vide citizens with targets for their frustration.
One key implication of our argument, then, is that public
service evaluations should not be the only grievances asso-
ciated with protest participation. Experience with corrup-
tion, for example, would seem to combine a relatively di-
rect chain of responsibility with firsthand experience with
poor governance (see Gingerich, 2009). The same could be
true of crime victimization. Yet many of the individual and
country level measures used to capture grievances in pre-
vious studies, like sociotropic and personal economic con-
siderations or GDP growth (see Dalton et al., 2009), seem to
require a more difficult calculus on the part of prospective
protestors to attribute blame, or might be too distant to de-
press support for key regime institutions, and thus motivate
participation. Our expectation is that public service evalua-
tions will be more powerfully associated with protest par-
ticipation due to the reasons outlined above.
To our knowledge, no study has yet explored how pub-
lic service evaluations are associated with protest partici-
pation in Latin America, or across countries. However, a re-
cent article on South Africa finds a positive and significant
relationship between access to basic public services (e.g.,
water, electricity, and plumbing) and the incidence of vio-
lent and non-violent protests (De Juan and Wegner, 2017).
The authors identify the phenomenon connecting the two
as “horizontal inequality,” a form of relative deprivation in
which individuals compare public service inadequacies in
their own neighborhoods with more effective delivery in
neighboring areas. In the case of public services like health,
education, and transit infrastructure, it is unclear that rela-
tive deprivation would be the mechanism at work. Evalua-
tions of these public services are unrelated to social class,
leading us to believe that the role poor perceived delivery
plays in mobilizing protestors has more to do with the sig-
nals it sends to citizens regarding state incompetence and
responsibility than its effect on aggrieved individuals’ ob-
jective well-being vis-à-vis other citizens. In our view,
negative public service evaluations represent a grievance
that lowers barriers to participate in protests –we intend to
provide an empirical test of that perspective in the follow-
ing section.
Data and Measurement
To test our argument, we utilize data from the 2012 and
2014 AmericasBarometer national surveys of Brazil, and
then generalize those results to AmericasBarometer data
from 18 Latin American countries. The AmericasBarome-
ter surveys are conducted biennially by the Latin American
Public Opinion Project (Lapop), housed at Vanderbilt Uni-
versity. Since 2004, the AmericasBarometer has measured
democratic attitudes and behaviors using national probabil-
ity samples of voting-age adults in countries throughout the
Western Hemisphere.6
We choose the Brazilian case for a number of rea-
sons. First, recent protests in Brazil have inspired numer-
6 For more information on the Lapop, its sources of funding, and the sampling
methodology and question wording employed, please visit www.vanderbilt.
edu/lapop/. See Appendix for descriptive statistics.
ous grievance-based explanations, ranging from corruption
(Rose-Ackerman, 2017) to economic recession (Carvalho,
2017) to poor public service provision (Watts, 2013). Bra-
zil thus offers an excellent test case to adjudicate be-
tween rival grievance-based explanations. Second, Brazil
went from being a relatively low protest country in 2012
to a highly contentious case in 2014 (Moseley and Layton,
2014). Drawing on two national surveys carried out in two
very distinct contexts thus helps us generalize to other Latin
American countries. Finally, as the region’s largest democ-
racy and economy, and currently at a crossroads follow-
ing Dilma Rousseff’s impeachment and the rise of far-right
presidential candidate Jair Bolsonaro, we argue the Brazil-
ian case is particularly vital to understanding the conten-
tious consequences of grievances in Latin America.
Our dependent variable is whether or not individu-
als report participating in a street march or demonstra-
tion during the prior 12 months. This provides a relatively
straightforward measure of the common street-based ac-
tivism that has been on the rise in Latin America in recent
years (Moseley, 2018). Our key independent variable is an
index drawn from three specific survey items. In 2012, the
AmericasBarometer began to include three specific ques-
tions aimed to gauge citizen evaluations of the quality of
public services in their community. These included ques-
tions about the quality of roads, the quality of public edu-
cation, and the quality of public health services. We com-
bine responses to these three questions in a single index
we call the Public Services Index, but also run a regional
model that includes each variable separately that we in-
clude in the Appendix Table A4.
In addition to measuring perceptions of public servic-
es, we include several variables to estimate the effects of
a variety of potential protest-motivating grievances. These
include corruption and crime victimization, perceptions of
the government’s effort to combat corruption, perception
of crime in one’s neighborhood, and sociotropic and pock-
etbook economic evaluations (all are scaled 0-100). In the
regional analysis, in which we utilize multilevel modeling
techniques, we also include second level economic vari-
ables like inequality (GINI), economic growth, and unem-
ployment, given the potential relevance of these trends in
motivating protest behavior. Each logistic regression model
also includes controls for skin color, sex, age, wealth, level
of education, civic activism, and the size of city where the
individual lives. Altogether, these variables form a compre-
hensive list of reasons why individuals might take to the
In the second section of the analysis, we also report
results from a Sobel-Goodman mediation test, which is
designed to offer evidence for the mechanisms we ar-
gue connect public service grievances and protest partici-
pation. We use the AmericasBarometer measure for sys-
tem support, which we believe will capture the extent to
which experience with poor governance causes citizens
to lose faith in key regime institutions, including trust in
congress, the courts, and pride in the system (Booth and
Seligson, 2009; see Appendix). To measure whether pub-
lic service grievances translate into more negative evalu-
ations of incumbent politicians, as predicted by our theo-
ry, we use a simple measure of presidential approval. Our
expectation is that while the relationship between pub-
lic service evaluations and protest is significant, part of
that effect flows through the two mechanisms we describe
We therefore adopt a two-stage approach to testing our
hypotheses. First, we evaluate which grievances are most
strongly associated with protest in the Brazilian case, uti-
lizing logistic regression analysis of complex survey data.
We then provide a general sense of the relationship be-
tween public service evaluations and protest participation
across Latin America, estimating regional models of pro-
test which include fixed effects for country and survey year,
and second level variables for inequality (GINI), economic
growth, and unemployment.7 Then, we test for the mech-
anisms we argue make public service grievances a potent
source of protest activity –declining support for the political
system and more negative evaluations of incumbent presi-
dents. These complementary approaches allow us to both
offer a generalizable account of the relationship between
public service grievances and protest, while also increasing
internal validity by focusing on a country where we have a
clear understanding of the nature of contention at that mo-
ment and highlighting the mechanisms through which ser-
vice evaluations and protest are related.
In Table 1, we present results of two estimated logistic re-
gression models of protest participation in Brazil, pooling
data from the 2012 and 2014 country surveys while includ-
ing a dummy for survey year. Model 1 includes only con-
trol variables, whereas Model 2 includes a set of relevant
grievances. Both models reveal a significant negative rela-
tionship between public service evaluations and reported
protest participation. In Model 2, only two grievances are
significantly related to protest –public service evaluations
and crime victimization. All other grievances are uncorre-
lated with protest participation, though corruption victim-
ization nears statistical significance (p<.12). Further, it ap-
pears that wealth actually has a positive and statistically
significant effect on protest participation, along with com-
munity activism and education. These results thus produce
little evidence for a theory of relative deprivation, given
that protestors seem to come from more privileged back-
grounds in the case of Brazil.
7 We supplement standard logistic regression models of participation with an
instrumental variables approach, which helps account for issues related to en-
dogeneity, and present those findings in the Appendix Table A2.
Table 1.
Estimated Logistic Regression Models of Protest in Brazil
DV: Protest Participation (0 or 1)
Variables Model 1 Model 2
Perception of Public Service Quality -0.012*** -0.009**
(0.004) (0.005)
Skin Color -0.009 -0.008
(0.038) (0.039)
Female -0.192 -0.130
(0.160) (0.164)
Age -0.031*** -0.031***
(0.007) (0.007)
Wealth Quintile 0.132** 0.127*
(0.065) (0.066)
Education 0.131*** 0.128***
(0.026) (0.027)
Community Activism 0.018*** 0.017***
(0.004) (0.004)
Urban -0.491 -0.433
(0.360) (0.365)
Size of Place -0.045 -0.015
(0.082) (0.085)
Corruption Victim 0.003
Crime Victim 0.005***
Perception of Crime 0.000
Perception of Efforts to Combat Corruption -0.001
Sociotropic Economic Evaluation -0.000
Pocketbook Economic Evaluation 0.001
2014 0.574*** 0.605***
(0.165) (0.177)
Constant -2.696*** -3.206***
(0.653) (0.703)
Observations 2,974 2,853
Standard errors in parentheses.
*** p<0.01, ** p<0.05, * p<0.1
On substantive effects, again it appears that public ser-
vice evaluations outpace other grievances in terms of pro-
test potential –a decrease from the most positive opinion of
public service quality to the lowest evaluation more than
doubles the likelihood that individuals have participated in
a street march or demonstration during the previous year
(Figure 3). These effects fail to achieve the substantive im-
portance of education or community participation, which
can increase the odds of participation by as much as 500
percent, and thus offer support for the resource mobiliza-
tion framework. Still, the observed effects for public service
evaluations register as consequential in terms of under-
standing recent protests in Brazil.
Figure 3.
Predicted Probabilities of Protest Participation in Brazil
0 20 40 60 80 100
Evaluation of Public Services - Brazil 2012-2014
95% CI Fitted values
Put simply, a less focused investigation of how grievanc-
es are correlated with protest in Brazil might lead to the
conclusion that they do not particularly matter in under-
standing contentious behavior. Sociotropic and pocketbook
evaluations fail to exert significant effects on the probabil-
ity that individuals protest, nor do protestors seem to come
from marginalized populations. However, when individuals
have firsthand experience with state incompetence and can
reasonably attribute blame –as we argue is the case with
public service evaluations– it is clear that grievances are
correlated with mobilization.
Table 2 reports results from two logistic regression models
of protest participation across 18 countries in Latin Ameri-
ca. Model 3 includes only control variables at the individual
and country levels, while Model 4 includes the full battery
of grievances from Model 2, along with second-level eco-
nomic variables that could reveal evidence of protest spik-
ing during hard economic times. Results from both models
lend strong support to the notion that public service evalua-
tions are associated with the decision to protest.
Table 2.
Estimated Logistic Regression Models of Protest
in Latin America
DV: Protest Participation (0 or 1)
Variables Model 3 Model 4
Perception of Public Service Quality -0.007*** -0.005***
(0.001) (0.001)
Skin Color 0.036*** 0.035***
(0.010) (0.010)
Female -0.344*** -0.288***
(0.032) (0.034)
Age -0.007*** -0.005***
(0.001) (0.001)
Wealth Quintile 0.024* 0. 014
(0.013) (0.013)
Education 0.055*** 0.050***
(0.004) (0.005)
Community Activism 0.016*** 0.014***
(0.001) (0.001)
Urban -0.003 0.033
(0.060) (0.062)
Size of Place -0.062*** -0.034*
(0.017) (0.018)
Corruption Victim 0.005***
Crime Victim 0.005***
Perception of Crime -0.000
Perception of Efforts to Combat
Corruption 0.000
Sociotropic Economic Evaluation 0.000
Pocketbook Economic Evaluation -0.001
2014 -0.033 -0.033
(0.077) (0.074)
GINI 3.638 4.19 4
(7.841) (7.590)
GDP Growth -0.044* -0.044**
(0.023) (0.022)
Unemployment -0.099 -0.103
(0.100) (0.097)
Constant -4 .147 -5.014
(3.871) (3.746)
Observations 56,354 53,364
Number of Country Years 36 36
Standard errors in parentheses.
*** p<0.01, ** p<0.05, * p<0.1
According to the results of the two models, public service
evaluations have a significant negative impact on the likeli-
hood that individuals take part in protests. In other words,
an increase in the perceived quality of roads, schools, and
public health services decreases the probability that one
will take to the streets. Other grievances are also significant
predictors of protest –namely, corruption and crime victim-
ization. On the other hand, perceptions of crime, corruption,
and the current economic situation (either personal or so-
ciotropic) fail to achieve statistical significance.
Figure 3 plots standardized coefficients to better com-
pare the relative import of each variable in Model 4. Clear-
ly, having direct experience with crime or corruption is sig-
nificantly associated with protest participation, in keeping
with findings from Gingerich (2009). Among all of the vari-
ables in the model, public service evaluations seem to have
one of the strongest negative effects on protest behavior.
To put this in perspective, moving from the most negative
to most positive evaluations of public services decreases
the probability of protesting by nearly 70 percent. Finally,
it should be noted that education and community activism
continue to exert the strongest effects on protest participa-
tion, in keeping with the resource mobilization approach to
understanding contention and recent findings from the pro-
test literature (Figure 3).
Figure 4.
Predictors of Protest Participation in Latin America
(Standardized coecients)
-0.20 -0.10 0.00 0.10 0.20 0.30 0.40
Size of Place
Community Activism
Years of Education
Wealth (quintile)
Skin Color
Pocketbook Economic Evaluation
Sociotropic Economic Evaluation
Perception of Insecutity
Govermment Fight Corruption
Crime Victim
Corruption Victim
Public Service Evaluation
95% Confidence Interval
(Design-Effect Based)
Source: AmericasBarometer by Lapop.
F = 79.800
N = 53364
We fail to uncover much support for the notion that coun-
try-level grievances exert significant influence on individ-
ual level protest behavior, yet we must caution that this
analysis draws on only 18 country cases over two rounds of
the AmericasBarometer. There is some initial evidence that
economic downturns have a slight impact on the likelihood
that individuals will protest (a 5% decrease in growth rate
results in a .02 shift in the probability of protesting), yet we
find little effect for either inequality or unemployment. A
country-level analysis of the effect of economic factors on
rates of activism would therefore find little evidence for
grievance-based protest in Latin America. Yet we find com-
pelling evidence that grievances can correlate with protest
participation, as is the case with perceptions of public ser-
vice quality. In the following section, we seek to shed light
on the mechanisms by which public service grievances
translate into protest participation.
Testing for Causal Mechanisms using Sobel-Goodman
Mediation Tests
Our argument rests on the notion that public service griev-
ances activate protestors through firsthand experience with
poor governance, which makes citizens more critical of the
political system, and supplies them with targets for claim
making. In this section, we carry out a Sobel-Goodman me-
diation test to evaluate the extent to which 1) system sup-
port and 2) presidential approval serve as mediators in the
relationship between public service grievances and protest
The Sobel-Goodman approach tests whether a mediator
carries at least part of the effect of an independent variable
on the dependent variable of interest (Sobel, 1982). It con-
stitutes an effective way of testing for the presence of an
important causal mechanism –in this case, we aim to ascer-
tain if the effect of public service evaluations flows to the
probability of protesting through declining system support,
and more negative evaluations of incumbent politicians.
Our argument is that firsthand experience with poor gov-
ernance via substandard public service delivery decreases
support for key regime institutions and elevates blame di-
rected at incumbent politicians (in this case, the president).8
While we have already reported graphs indicating the
strong relationship between public service evaluations and
both system support and presidential approval, this statisti-
cal approach represents a more direct test of the notion that
these two variables serve as important mechanisms con-
necting public service grievances and protest participation.
Table 3.
Results of Sobel-Goodman Mediation Tests
Brazil Latin America
Support Presidential
Approval System
Support Presidential
Test Coeff. -0.005
(0.006) -0.009**
(0.005) -0.009***
(0.001) -0.009***
Proportion of Total
Effect Mediated 0.08 0 .15 0 .16 0 .16
Standard errors in parentheses.
*** p<0.01, ** p<0.05, * p<0.1
Table 3 reports results from the Sobel-Goodman mediation
tests, which control for the same battery of individual le-
vel variables included in the models presented in Tables 1
and 2. They indicate that, except in the case of system sup-
port in Brazil, our expectations are borne out –individuals’
firsthand experience with poor governance translates into
an increased likelihood of protesting through declining sys-
tem support (in Latin America) and more negative evalua-
tions of the incumbent president (Brazil and Latin America).
Significant results indicate an important reduction in the
8 We lack a measure across Latin American countries that asks respondents to
rate any other single elected official other than the president, but given what
we know about recent credit-claiming for service improvement by presidents,
it seems like the most appropriate test of our argument.
effect of public service evaluations on protest when the me-
diator is included in the model –in other words, in the mo-
dels of protest in Latin America, more than 0.30 of the to-
tal effect exerted by perceptions of public service quality on
the likelihood of protest is the result of their impact on sys-
tem support and presidential approval, which in turn makes
individuals more likely to protest. The absence of a medita-
tion effect for system support in Brazil indicates that public
service evaluations affect protest through other pathways,
and runs counter to results from South Africa (De Juan and
Wegner, 2017).
Overall, the results provide support for the two mecha-
nisms we identify in our argument as connectors between
public service grievances and protest participation. While
public service evaluations clearly exert an independent ef-
fect on the likelihood that individuals protest, they also de-
press system support and presidential approval, which in
turn are positively associated with the probability of pro-
testing. Public service evaluations thus represent a griev-
ance that enables citizens to connect their own personal
experiences with poor governance, and then provides indi-
viduals with a target for their discontent. When grievances
are characterized by these important protest-producing fac-
tors, they can indeed matter in understanding recent pro-
test participation in Latin America.
The increasing salience of protest in many parts of Lat-
in America in the 2010s calls for scholarly attention to the
sources and consequences of popular contention. While
previous work has demonstrated the importance of indi-
vidual-level variables like community activism and educa-
tion as predictors of protest, this paper has sought to shed
light on the importance of grievances –in particular, dis-
satisfaction with the quality of public services–in mobiliz-
ing citizens. We find that while the organizational resources
stressed in the extant literature very clearly matter for un-
derstanding protest participation, grievances too can push
citizens to the streets under certain conditions.
The importance of public service evaluations as a pre-
dictor of protest in Latin America is indicative of a number
of phenomena, chief of which is the importance of individ-
uals increasingly making connections between their own
living conditions and government performance. Where citi-
zens perceive blame for poor public services and have di-
rect knowledge of poor governance, they are more prone
to seeking redress through contentious modes of behavior.
Latin Americans are widely supportive of democratic gover-
nance and a more interventionist state, and where the state
fails to deliver on those expectations, citizens across the re-
gion have shown a willingness to hold public officials ac-
countable through non-institutional methods.
Recent work has indicated that the surge in protest
participation across Latin America is not a sign of creep-
ing extremism, but rather a conventional response to poor
government performance amid high levels of citizen en-
gagement in politics (Moseley, 2018). The findings pre-
sented here complement that perspective, as we shed light
on one particularly important sphere in which Latin Amer-
ican regime performance leaves much to be desired, and
has thus played an important part in sparking contentious
participation. While not all grievances translate into action,
claims regarding public service provision satisfy the key
criteria that seem to make protest more likely. In this new
era of democratic governance and citizen engagement in
politics in Latin America, policymakers can no longer expect
citizen complacency with respect to poor governance.
Alexander, Peter (2010). “Rebellion of the Poor: South Africa’s Ser-
vice Delivery Protests –a Preliminary Analysis.” Review of Afri-
can Political Economy 37(123): 25-40.
Alves, Helena (2013). “Breaking Down Brazil’s Protests,” Aljazeera,
June 21, 2013,
Anderson, Christopher (1995). Blaming the Government: Citizens
and the Economy in Five
European Democracies. North Castle, NY: ME Sharpe.
Arce, Moisés (2014). Resource Extraction and Protest in Peru. Pitts-
burgh, PA: University of Pittsburgh Press.
Arce, Moisés and Jorge Mangonnet (2013). “Competitiveness, Par-
tisanship, and Subnational Protest in Argentina.” Comparative
Political Studies 46(8): 895-919.
Atkeson, Lonna Rae and Randall W. Partin (1995). “Economic and
Referendum Voting: A Comparison of Gubernatorial and Sena-
torial Elections.” American Political Science Review 89(1): 99-
Bateson, Regina (2012). “Crime Victimization and Political Partic-
ipation.” American Political Science Review 106(3): 570-587.
Beaulieu, Emily (2014). Electoral Protest and Democracy in the De-
veloping World. New York, NY: Cambridge University Press.
Blattman, Christopher (2009). “From Violence to Voting: War and
Political Participation in Uganda.” American Political Science
Review 103(2): 231-247.
Boulding, Carew (2014). NGOs, Political Protest, and Civil Society.
New York, NY: Cambridge University Press.
Campante, Filipe R. and Davin Chor (2012). “Why was the Arab
World Poised for Revolution? Schooling, Economic Opportuni-
ties, and the Arab Spring.” The Journal of Economic Perspec-
tives 26(2) (2012): 167-187.
Carlin, Ryan E. and Shane P. Singh (2015). “Executive Power and Eco-
nomic Accountability.” The Journal of Politics 77(4): 1031-1044.
Carvalho, Laura (2017). “Facing Unemployment, Austerity and
Scandal, Brazil Struggles to Keep it Together.” The Conversa-
tion, January 27, 2017,
Crosby, Faye (1976). “A Model of Egoistical Relative Deprivation.”
Psychological Review 83(2): 85.
Dalton, Russell J.; Alix van Sickle and Steven Weldon (2009). “The
Individual-Institutional Nexus of Protest.” British Journal of Po-
litical Science 40(1): 51-73.
Duch, Raymond M. and Randolph T. Stevenson (2008). The Eco-
nomic Vote: How Political and Economic Institutions Condition
Election Results. New York, NY: Cambridge University Press.
Dunning, Thad (2012). Natural Experiments in the Social Sciences:
A Design-Based Approach. New York, NY: Cambridge Univer-
sity Press.
Epstein, Lee and Jeffrey A. Segal (2000). “Measuring Issue Sa-
lience.” American Journal of Political Science 44(1): 66-83.
Falleti, Tulia G. and Thea N. Riofrancos (2018). “Endogenous Par-
ticipation: Strengthening Prior Consultation in Extractive Econ-
omies.” World Politics. 70 (1): 86-121.
Foster, Mindi D. and Kimberly Matheson (1999). “Perceiving and
Responding to the Personal/Group Discrimination Discrepancy.”
Personality and Social Psychology Bulletin 25(10): 1319-1329.
Garay, Candelaria (2007). “Social Policy and Collective Action: Un-
employed Workers, Community Associations, and Protest in Ar-
gentina.” Politics & Society 35(2): 301-328.
Garay, Candelaria (2016). Social Policy Expansion in Latin Ameri-
ca. New York, NY: Cambridge University Press.
Gomez, Brad T. and J. Matthew Wilson (2001). “Political Sophisti-
cation and Economic Voting in the American Electorate: A The-
ory of Heterogeneous Attribution.” American Journal of Political
Science 45(4): 899-914.
Graham, Richard and Peter H. Smith (eds.) (2012). New Approach-
es to Latin American History. Austin, TX: University of Texas
Gurr, Ted Robert (1970). Why Men Rebel. Princeton, NJ: Princeton
University Press.
Javeline, Debra (2003). “The Role of Blame in Collective Action:
Evidence from Russia.”American Political Science Review 97(1):
107- 121.
Kinder, Donald R. and D. Roderick Kiewiet (1981). “Sociotropic
Politics: The American Case.” British Journal of Political Sci-
ence 11(2): 129-161.
Latin American Public Opinion Project (2014). “AmericasBarometer
Layton, Matthew L.; Maureen M. Donaghy and Lúcio R. Rennó
(2017). “Does Welfare Provision Promote Democratic State Le-
gitimacy? Evidence from Brazil’s Bolsa Família Program.” Latin
American Politics and Society 59(4): 99-120.
Long, Gideon (2011). “Chile Student Protests Point to Deep Dis-
content,” BBC,
Luna, Juan Pablo and Hillel D. Soifer (2015). “Surveying State Ca-
pacity.” AmericasBarometer Insights 119.
Machado, Fabiana; Carlos Scartascini and Mariano Tommasi
(2011). “Political Institutions and Street Protests in Latin Ameri-
ca.” Journal of Conflict Resolution 55(3): 340-365.
Mangonnet, Jorge and María Victoria Murillo (2016). “Protests of
Abundance: Commodity Rents and Rural Lockouts in Argenti-
na,” presented at the Annual Conference of the American Politi-
cal Science Association, Philadelphia, September 2016.
McAdam, Douglas (1982). Political Process and the Development of
Black Insurgency. Chicago: University of Chicago Press.
McCarthy, John D. and Mayer N. Zald (1977). “Resource Mobiliza-
tion and Social Movements: A Partial Theory.” American Jour-
nal of Sociology 82(6): 1212-1241.
Melguizo, Angel; Sebastián Nieto-Parra; José Ramón Perea and
Jaime Ariel Perez (2017). “No Sympathy for the Devil! Policy Pri-
orities to Overcome the Middle-Income Trap in Latin America.”
Working Paper Nº 340. Research area: Latin American Econom-
ic Outlook.
Moseley, Mason W. (2015). “Contentious Engagement: Under-
standing Protest Participation in Latin American Democracies.”
Journal of Politics in Latin America 7(3): 3-48.
Moseley, Mason W. (2018). Protest State: The Rise of Everyday
Contention in Latin America. New York, NY: Oxford University
Moseley, Mason and Matthew L. Layton (2013). “Prosperity and
Protest in Brazil: The Wave of the Future for Latin America?”
Insights Series Nº I0893. Vanderbilt University: Latin American
Public Opinion Project (Lapop).
Murillo, María Victoria and Jorge Mangonnet (2016). “Protests of
Abundance: Commodity Rents and Rural Lockouts in Argenti-
na,” presented at the Annual Conference of the American Politi-
cal Science Association, Philadelphia, September 2016.
Olson, Mancur (1965). The Logic of Collective Action. Cambridge,
MA: Harvard University Press.
Oxfam (2017). “Latin America Remains the Most Unequal Region
in the World,” Oxfam, December 18, 2017, http://blogs.oxfam.
Powell Jr.; G. Bingham and Guy D. Whitten (1993). “A Cross-Nation-
al Analysis of Economic Voting: Taking Account of the Political
Context.” American Journal of Political Science 37(2): 391-414.
Rose-Ackerman, Susan (2017). “Political Corruption and Reform in
Democracies: Theoretical Perspectives,” In Junich Kawata (ed.),
Comparing Political Corruption and Clientelism. New York, NY:
Routledge, 65-82.
Seligson, Mitchell A. (2006). “The Measurement and Impact of
Corruption Victimization: Survey Evidence from Latin America.”
World Development 34(2): 381-404.
Silva, Eduardo (2009). Challenging Neoliberalism in Latin Ameri-
ca. New York, NY: Cambridge University Press.
Simões, Rogério, “Broken Promises and Corruption Fuel Brazil Pro-
tests,” CNN, June 21, 2013
Sovey, Allison J. and Donald P. Green (2011). “Instrumental Vari-
ables Estimation in Political Science: A Readers’ Guide.” Ameri-
can Journal of Political Science 55(1): 188-200.
Tavits, Margit (2007). “Clarity of Responsibility and Corrup-
tion.” American Journal of Political Science 51(1): 218-229.
Transparency International (2017). “About 1 in 3 People Using Pub-
lic Services in Latin America and Caribbean Paid a Bribe in Past
Year.”Transparency International, https://www.transparency.
Unesco (2013). “The State of Education in Latin America and the
Caribbean: Towards a Quality Education for All – 2015.” Unesco, MULTIMEDIA/FIELD/
Van Stekelenburg, Jacquelien and Bert Klandermans (2013). “The So-
cial Psychology of Protest.” Current Sociology 61(5-6): 886-905.
Walker, Lain and Leon Mann (1987). “Unemployment, Relative De-
privation, and Social Protest.” Personality and Social Psycholo-
gy Bulletin 13(2): 275-283.
Walton, John and Charles Ragin (1990). “Global and National
Sources of Political Protest: Third World Responses to the Debt
Crisis.” American Sociological Review 55(6): 876-890.
Watts, Jonathan (2013). “Brazil Erupts in Protest: More than a Mil-
lion on the Streets.” The Guardian, June 21, 2003, https://www.
Wilkinson, Steven I. (2009). “Riots.” Annual Review of Political Sci-
ence 12: 329-343.
World Bank (2013). “Universal Healthcare on the Rise in Lat-
in America.” The World Bank,
World Bank (2017). “Higher Education Expanding in Latin Ameri-
ca and the Caribbean, but Falling Short of Potential.” The World
Yashar, Deborah J. (2005). Contesting Citizenship in Latin Ameri-
ca: The Rise of Indigenous Movements and the Postliberal Chal-
lenge. New York, NY: Cambridge University Press.
Zaller, John (1992). The Nature and Origins of Mass Opinion. New
York, NY: Cambridge University Press.
Zovatto, Daniel (2015). “Dinero y política en América Latina.” Re-
vista de Derecho Electoral 2(26): 5-26.
Table A1.
Question Wording and Descriptive Statistics
(Regional Statistics)
Variable Question Wording or Explanation N Mean Standard
Deviation Min Max
Dependent Variables
“In the last 12 months, have you par-
ticipated in a demonstration or protest
0-No, 100-Yes
57,829 8.06 27.22 0 100
Independent Variables
A weighted index of three questions
asking respondents about the quality
of roads, schools, and health services.
0-Lowest, 100-Highest.
58,102 51.02 19 .10 0100
A weighted index of five questions
that measure support for the existing
system. The five ask respondents “To
what extent do you think the courts
of justice guarantee a fair trial?,” “To
what extent do you respect the politi-
cal institutions of (country)?,” “To what
extent do you think that citizens’ basic
rights are well protected by the politi-
cal system of (country)?,” “To what ex-
tent do you feel proud of living under
the political system of (country)?,” and
“To what extent do you think that one
should support the political system of
0-Lowest, 100-Highest
57,166 51. 8 0 23.56 0 100
“Speaking in general of the cur-
rent administration, how would you
rate the job performance of President
0-Lowest, 100-Highest.
57,413 56.81 24.37 0 10 0
with Health
“And thinking about this city/area
where you live, are you very satis-
fied, satisfied, dissatisfied, or very dis-
satisfied with the condition of public
health services?”
Originally 1-Very Satisfied, 4-Very
Dissatisfied. Very Satisfied Convert-
ed to 100 point index, 0-Lowest, 100
55,870 47.99 25.62 0 100
Dummy variable asking individuals if
they were victims of corruption.
0-No, 100-Yes. 58,174 20.51 40.38 0 100
Crime Victim
“Has any other person living in your
household been a victim of any type
of crime in the past 12 months? That
is, has any other person living in your
household been a victim of robbery,
burglary, assault, fraud, blackmail, ex-
tortion, violent threats or any other
type of crime in the past 12 months?”
0-No, 100-Yes.
58,088 32.17 46.71 0 100
Perception of
“Speaking of the neighborhood where
you live and thinking of the possibility
of being assaulted or robbed, do you
feel very safe, somewhat safe, some-
what unsafe or very unsafe?”
Originally 1-Very Safe, 4-Very Unsafe.
Converted to 100 point index, 0-Low-
est, 100 Highest.
57,942 43.52 29.95 0 100
Efforts to
Fight Crime
“To what extent would you say the
current administration combats (fights)
government corruption?”
Originally 1-Not at All, 7-A Lot. Con-
verted to 100 point index, 0-Lowest,
100 Highest.
56,382 42.42 30.30 0 100
“Do you think that the country’s cur-
rent economic situation is better than,
the same as or worse than it was 12
months ago?”
0-Worse, 50-Same, 100-Better.
57,452 38.14 36.00 0 100
“Do you think that your economic sit-
uation is better than, the same as, or
worse than it was 12 months ago?”
0-Worse, 50-Same, 100-Better.
57,809 45.33 35.04 0 100
Skin Color
“[When the interview is complete,
WITHOUT asking, please use the color
chart and circle the number that most
closely corresponds to the color of the
face of the respondent]”
1-Lightest, 11-Darkest
58,072 4.49 1.69 1 11
Female Sex of individual. 1 if male, 2 if female. 58,230 1.51 0.50 1 2
Age Age of individual.
Theoretically can be anywhere be-
tween 0 into the 100s. 57,958 40.33 16 .10 16 99
A weighted index that measures
wealth based on the possession of
certain household goods such as tele-
visions, refrigerators, convention-
al and cellular telephones, vehicles,
washing machines, microwave ovens,
indoor plumbing, indoor bathrooms
and computers.
1-Lowest, 5-Highest
57,681 2.95 1.42 1 5
Education “How many years of schooling have
you completed?”
0-Lowest, 18+- Highest. 57,933 9.31 4.52 0 18
Urban 1 if interview took place in urban
setting; 2 if rural. 58,231 1.29 0.45 1 2
Size of Place 1 if National Capital (Metropolitan Ar-
ea; 2 if Large City; 3 if Medium City; 4
if Small City; 5 if Rural Area 58,231 3.07 1.51 15
2014 Control for the year 2014. - - - - -
Public Health
“Have you used any public health ser-
vices in the last twelve months?”
0-No, 100-Yes. 31,445 7.43 26.22 0 100
Approval Same as Above Dependent Variable. - - - - -
Ideology Ideology placement that goes from
1-10 with 1 being most liberal and 10
most conservative. 47,858 5.49 2.68 1 10
A weighted index of community par-
ticipation in religious associations,
parents’ associations at schools, and
community organizations.
0-Lowest, 100-Highest.
58,207 27.37 23.30 0 10 0
GINI GINI coefficient for each country year. 58,231 0.48 0.04 0.41 0.57
Growth Economic growth for each country
year. 58,231 3.74 2.44 -4.0 10.2
ment Unemployment rate for each country
year. 58,231 6 . 17 2.84 2.3 14.7
In spite of this compelling initial evidence that grievances
surrounding public service provision are associated with
swelling rates of contention, the possibility remains that
this relationship is spurious for two reasons: 1) participation
in protests actually increases awareness about public servi-
ce provision, and thus fuels lower evaluations, and 2) eva-
luations of public service provision are shaped by evalua-
tions of political actors and thus are not truly objective. To
account for these limitations in our first two models, we use
an instrumental approach that endeavors to extract what is
exogenous from public service evaluations and determine if
it indeed shapes protest participation.
The logic of instrumental variables regression goes as
follows: when one suspects a key independent variable is
endogenously related to the dependent variable of interest,
it must identify an instrument that satisfies two conditions.
First, the instrument must be correlated with the endoge-
nous regressor –in this case, public service evaluations. Se-
cond, the instrument must be uncorrelated with the error
term –in other words, it satisfies the exclusion restriction
(Dunning, 2012). In our case, the exclusion restriction re-
quires that the instrument only affects protest participation
through perceptions of public service quality. The instru-
ment we have chosen comes from a survey item included
in the 2014 AmericasBarometer designed to measure state
capacity (Luna and Soifer, 2015). It asks how long it would
take for the police to arrive at your house if you called them
(Lapop, 2014). Unsurprisingly, this variable is highly related
to perceptions of public service quality (r = -0.21***), and
wholly unrelated to protest participation (r = -0.01).
Table A2.
Public Service Evaluations and Protest in Latin America -
IV Regression
Variables (1)
Second stage (DV = Protest Participation)
Public Service Index -0.086**
Female -1.613***
Age -0.065***
Wealth (quintile) 0.848***
Skin Color 0.410***
Constant 13.201***
First stage (DV = Public Service Index)
Police Response Time -2.694***
Female 0.378*
Age 0.042***
Wealth (quintile) -0.436***
Skin Color 0.250***
Constant 57.192***
Observations 27 ,19 0
R-squared 0.007
Robust standard errors in parentheses.
*** p<0.01, ** p<0.05, * p<0.1
Table A2 reports results from an instrumental variables re-
gression model (2SLS). The first stage models perceptions
of public service provision includes our measure for state
capacity, while the second stage models participation in a
protest. Given the nature of the exclusion, we only include
other covariates that are completely exogenous to pro-
test participation –i.e., age, gender, wealth, and skin col-
or (Sovey and Green, 2011). The results indicate that after
purging our results of the potential bias associated with an
endogenous regressor, the results presented about hold.
This suggests that the causal arrow flows primarily from
public service evaluations to protest rather than the oth-
er way around. It also rules out the possibility that public
service evaluations are actually proxies for presidential ap-
proval as we argue that our state capacity measure is large-
ly independent from political affiliations.
Table A3.
Public Service Evaluations and Presidential Approval in Brazil -
Separate 2012 and 2014 Models
DV: Presidential
Approval in 2012
DV: Presidential
Approval in 2014
Variables (1) (2)
Public Service Evaluations 0.138*** 0.075
(0.037) (0.050)
Interest in Politics 0.094*** -0.110**
(0.036) (0.047)
Public Service Evaluations x
Interest in Politics -0.001 0.003***
(0.001) (0.001)
Corruption Victim -0.039** -0.016
(0.016) (0.018)
Crime Victim -0.006 -0.009
(0.012) (0.014)
Perception of Crime -0.020 -0.010
(0.017) (0.019)
Government Efforts to Fight Crime 0.141*** 0.197***
(0.017) (0.022)
Sociotropic Economic Situation 0.092*** 0.116***
(0.017) (0.017)
Pocketbook Economic Situation 0.036** 0.068***
(0.015) (0.018)
Skin Color -0 .118 0.046
(0.214) (0.305)
Female 1.037 2.878**
(0.999) (1.215 )
Age 0.064* -0.019
(0.039) (0.042)
Wealth Quintile -0.740* -1.845***
(0.389) (0.496)
Education 0.12 8 -0.528***
(0 .161) (0.190)
Urban 2.817 3.060
(1.896) (2.483)
Size 0.003 0 .17 9
(0.429) (0.810)
Constant 60.441*** 64.464***
(4.545) (5.650)
Observations 1,409 1,438
R-squared 0.13 8 0.237
Standard errors in parentheses.
*** p<0.01, ** p<0.05, * p<0.1
Table A4.
Public Service Evaluations and Protest in Latin America
DV: Participated
in a Protest
(0 or 1)
Variables (1)
Health -0.002**
Roads -0.001
Schools -0.002**
Corruption Victim 0.005***
Crime Victim 0.005***
Government Efforts to Fight Crime -0.000
Perception of Crime 0.000
Sociotropic Economic Situation 0.0002
Pocketbook Economic Situation -0.0005
Skin Color 0.030***
Female -0.298***
Age -0.005***
Wealth Quintile 0.012
Education 0.051***
Community Participation 0.014***
Urban 0.017
Size of Place -0.027
GINI 3.653
Economic Growth -0.045**
Unemployment -0 .131
(0 .101)
Constant -4.619
Observations 48,420
Number of Country Years 36
Standard errors in parentheses.
*** p<0.01, ** p<0.05, * p<0.1
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
Conditional cash transfer programs may boost the electoral fortunes of incumbents among beneficiary groups, but do they also influence recipient attitudes toward state legitimacy? This article examines the relationship between Brazil's Bolsa Família program and recipients’ sense of the Brazilian state's political legitimacy, from 2007 to 2014. Using AmericasBarometer data and propensity score matching, this study provides evidence that targeted cash benefits correlate with citizens’ views of the state, but that this relationship is limited to increasing trust in core state institutions, local government, and incumbent political actors. Diffuse dimensions of regime legitimacy, including recipients’ sense of political community, support for regime principles, and retrospective perceptions of national economic performance, are largely unaffected by the receipt of targeted benefits. Over time, the evidence also suggests that the impact of program receipt on these measures of support remains largely unchanged.
Full-text available
Why has protest participation seemingly exploded across much of Latin America in recent years? How do individual- and country-level characteristics interact to explain the rise of contentious politics in countries like Argentina, Brazil, and Venezuela? I contend that the recent wave of protests in Latin America is the result of trends in community engagement and institutional development across the region’s young democracies. Specifically, I argue that low-quality institutions in democratic regimes push an increasingly large number of civically active Latin Americans toward more radical modes of political participation, as governments’ abilities to deliver on citizens’ expectations fail to match the capacity for mobilization of active democrats. Drawing on cross-national surveys of Latin America, I test this argument, finding that an interactive relationship between community engagement and ineffective political institutions helps explain the recent spike in protest activity in certain cases and the vast differences in protest participation observed throughout the region.
In the midst of an unprecedented era of democratic governments and economic prosperity, why are a record number of Latin American citizens choosing to participate in protests? This book argues that increasingly engaged citizenries, forged by economic progress and technological advances throughout the region, combined with dysfunctional political institutions have fueled more contentious modes of participation in Latin America, as citizens’ demands for government responsiveness have overwhelmed many regimes’ institutional capacity to provide it. Where weak institutions and active citizenries collide, countries can morph into “protest states,” where contentious participation becomes so common as to render it a conventional characteristic of everyday political life. Drawing on cross-national surveys from Latin America and a case study of Argentina, which includes a rich dataset of protest events and dozens of interviews with political elites and citizen activists, Moseley tests this explanation against other leading theories in the contentious politics literature. Rather than emphasizing how worsening economic conditions and mounting grievances fuel protest, this book builds the case that it is actually the improvement of economic conditions amid low-quality political institutions that lies at the root of surging contention in the region. In presenting and systematically defending this novel approach, Protest State offers a comprehensive multilevel, mixed-methods study of one of the most intriguing puzzles in Latin American politics today.
Throughout the twentieth century, much of the population in Latin America lacked access to social protection. Since the 1990s, however, social policy for millions of outsiders - rural, informal, and unemployed workers and dependents - has been expanded dramatically. Social Policy Expansion in Latin America shows that the critical factors driving expansion are electoral competition for the vote of outsiders and social mobilization for policy change. The balance of partisan power and the involvement of social movements in policy design explain cross-national variation in policy models, in terms of benefit levels, coverage, and civil society participation in implementation. The book draws on in-depth case studies of policy making in Argentina, Brazil, Chile, and Mexico over several administrations and across three policy areas: Health care, pensions, and income support. Secondary case studies illustrate how the theory applies to other developing countries.
This book proposes a selection model for explaining cross-national variation in economic voting: Rational voters condition the economic vote on whether incumbents are responsible for economic outcomes, because this is the optimal way to identify and elect competent economic managers under conditions of uncertainty. This model explores how political and economic institutions alter the quality of the signal that the previous economy provides about the competence of candidates. The rational economic voter is also attentive to strategic cues regarding the responsibility of parties for economic outcomes and their electoral competitiveness. Theoretical propositions are derived, linking variation in economic and political institutions to variability in economic voting. The authors demonstrate that there is economic voting, and that it varies significantly across political contexts. The data consist of 165 election studies conducted in 19 different countries over a 20-year time period.
In this groundbreaking study, Moisés Arce exposes a long–standing climate of popular contention in Peru. Looking beneath the surface to the subnational, regional, and local level as inception points, he rigorously dissects the political conditions that set the stage for protest. Focusing on natural resource extraction and its key role in the political economy of Peru and other developing countries, Arce reveals a wide disparity in the incidence, forms, and consequences of collective action.
This book shows that the third wave of democracy has been accompanied by a worldwide wave of opposition-initiated, election-related protests. Such electoral protests result from a failure on the part of incumbent and opposition elites in the developing world to negotiate acceptable terms of electoral conduct, and their consequences for democracy depend on the context in which they occur. Where election boycotts receive international support, they increase the probability of democratic reform, but where support is primarily domestic, there is a higher probability of authoritarian backsliding. Based on an extensive new data set covering nearly thirty years of electoral protest and election-related reform in the developing world, this book explores the causes of different types of electoral protest and their consequences for democracy. Statistical analysis and case studies provide readers with a complete picture of the dynamics surrounding developing world elections, protest, and democratization.
This unique book is the first comprehensive guide to the discovery, analysis, and evaluation of natural experiments – an increasingly popular methodology in the social sciences. Thad Dunning provides an introduction to key issues in causal inference, including model specification, and emphasizes the importance of strong research design over complex statistical analysis. Surveying many examples of standard natural experiments, regression-discontinuity designs, and instrumental-variables designs, Dunning highlights both the strengths and potential weaknesses of these methods, aiding researchers in better harnessing the promise of natural experiments while avoiding the pitfalls. Dunning also demonstrates the contribution of qualitative methods to natural experiments and proposes new ways to integrate qualitative and quantitative techniques. Chapters complete with exercises, and appendices covering specialized topics such as cluster-randomized natural experiments, make this an ideal teaching tool as well as a valuable book for professional researchers.
At the turn of the twentieth century, a concatenation of diverse social movements arose unexpectedly in Latin America, culminating in massive anti–free market demonstrations. These events ushered in governments in Argentina, Bolivia, Ecuador, and Venezuela that advocated socialization and planning, challenging the consensus over neoliberal hegemony and the weakness of movements to oppose it. Eduardo Silva offers the first comprehensive comparative account of these extraordinary events, arguing that the shift was influenced by favorable political associational space, a reformist orientation to demands, economic crisis, and mechanisms that facilitated horizontal linkages among a wide variety of social movement organizations. His analysis applies Karl Polanyi’s theory of the double movement of market society to these events, predicting the dawning of an era more supportive of government intervention in the economy and society.