Dear colleagues,
Thank you for taking the time to read this manuscript. I wrote it as a chapter to be included in an
edited volume looking into how protest waves are realigning democratic politics around the
world. The editors explicitly asked me to take up the issue of how these protests are influencing
democratic reform. Consequently, I have spent a lot of time in the chapter defining what
constitutes reform and discussing how to measure it. My goal in this endeavor has not been to
formulate a new theory, but to answer what amounts to a more empirical question. I welcome
any comments or suggestions that you may have.
José Alemán
1
CHAPTER 9: PROTEST AND INSTITUTIONAL CHANGE, In Protest and Democracy,
forthcoming, Lynne Rienner Publishers, 2014.
Scholars interested in social movements have for decades debated democracy’s effect on
dissent – that is, whether democracy increases or dampens protest. Comparatively speaking, less
has been written about how political protest affects democracy (Giugni 1998, 1999). Once a
democratic regime is installed, the expectation is that violent challenges will decline while
nonviolent protest flourishes (Schatzman 2005).1 While social protest is linked with measures of
good governance (Welzel et al. 2005, 140), little is known about what precise dimensions of
democracy social protest affects. This chapter takes up the issue of how recent protest waves are
contributing to institutional reforms in democratic countries.
The need to study how protest affects democratic politics stems not only from our limited
knowledge in this area, but also from the reality that protest is a familiar manifestation of
political behavior in modern democratic settings (Dalton et al. 2010; Johnston 2011). Some time
ago, students of contentious politics began to speak of protest as so common and expected a form
of political participation in democracies that they coined the term “modular” to refer to its use by
multiple constituencies with diverse objectives (Tarrow 1998, 37-40). Little, however, is known
about the consequences of these forms of political engagement. While protestors may not have to
fear for their personal integrity (political imprisonment, torture, disappearance, or extra judicial
killing) in most democracies today, many new democracies do not live up to their full potential
in the areas of representativeness, accountability, equality and participation (Foweraker and
Krznaric 2000, 2003, 314-315; Freedom House 2012; Lijphart 1993, 149). Instead, some are far
from attaining the ideal of a political system that, in perfecting these attributes, creates the
conditions for broad and equal citizenship among its subjects (Tilly 2007).2
2
The current wave of democratization gives us then the opportunity to examine how
protest may be realigning democratic politics around the world. More specifically, the chapter
looks at how social protest contributes to institutional reforms in both new and established
democracies, leaving for another chapter the question of changes in government policies
stemming from protest activity. It is by now well established that social movements can bring
about three types of changes in their immediate and surrounding environment: elevating the
status of challengers vis-à-vis third parties and authorities; catalyzing changes in policy; and
bringing about transformations in society, the political system, and/or prevailing beliefs
(Kitschelt 1986; Giugni 1998, 1999). Social movement scholars have particularly focused on
policy changes as a measurable outcome of political activity since these changes can be easily
measured and explained (Giugni 1999, xxii). Our data and empirical approach do not allow us to
say much about changes in particular government policies. Instead, the analysis in this chapter
pays more attention to systemic changes, both because they matter in their own right, and
because they can also alter power relations between various social movements and their
surroundings.
The chapter evaluates changes in the protection of civil and workers’ rights, changes that
have made democratic regimes more inclusive (or participatory), and increases in the ability of
other branches of government to constrain chief executives. The most robust findings to emerge
in regards to the effect of social protest on these dimensions of democratic governance is that
anti-government demonstrations help constrain the power of the executive, particularly in more
developed democracies. The results also indicate that riots have adverse effects on human rights,
particularly in more developed countries. Their effect on human rights, however, is positive in
3
the presence of judicial independence. A few words are in order regarding the choice of these
changes as indicators of democratic reform.
Markers of democratic reform
The most expansive definitions of liberal democracy sketch a vision of society where
certain spheres of private activity are beyond the reach of the state. Individuals in these
democracies are not only granted equal treatment before the law, but are also protected from the
uneven and arbitrary exercise of state power and from abuse by third parties, majorities or
unelected veto groups. These conceptualizations of democracy are very difficult to operationalize
however, and may reflect unique Western values and experiences. Consequently, I do not make
an attempt to hold democracies to such high evaluative standards.
For the purposes of this chapter, I will take the existence of many imperfect democracies
– particularly in the developing world – as a reality and argue that, although a political regime
cannot be fully democratic unless its citizens enjoy a great deal of autonomy, legal protection,
collective representation, meaningful participation, and political accountability, changes that
result in a more accountable and participatory political system can be considered indicators of
democratic reform. I make the assumption that guaranteeing and protecting certain human rights
helps democracies become more equitable and representative, but the analysis I present is more
definitive with respect to changes that make democracies more participatory and accountable.3
Political scientists tend to focus on representation and participation, the two processes that serve
to minimally distinguish democratic regimes from their alternatives, but how governments create
political accountability is equally important. I have selected four measures of reform for
4
consideration: the protection of civil and workers’ rights, higher electoral participation, and more
constrained executives.
Since Robert Dahl’s (1971) treatise on democracy, scholars have debated whether human
rights should be understood as part and parcel of the definition of political regimes, or as factors
that facilitate the functioning of these regimes, but are not constitutive of their basic properties.
While some call for separating “elements of repression that are necessary to the definition of
democracy” from “those that democracy is likely to influence once it has been established”
(Davenport 2007, 101), others do not regard countries that tolerate civil rights violations as
democratic. Absent these basic rights, the argument goes, any formal guarantees democracy
makes are rendered meaningless because citizens are unable to express their preferences and
have them weighted equally by their representatives or those who claim to compete for such a
role (Munck 2009, 126). Some go as far as to note that countries that are constitutional
democracies but violate civil rights be more appropriately considered as examples of an alltogether different regime category – competitive authoritarianism (Levitsky and Way 2002).
The question then is whether we should lump regime characteristics – the ability of
politicians to run for office and compete for citizen support in the presence of free and fair
elections and institutions that empower citizens – with the state’s ability to engage in coercive
activity that restricts or abrogates basic civil rights such as the freedoms of expression,
association and assembly (Davenport 2007, 101). In deciding between these two competing
positions, I note that democracy’s effect on human rights is conditional, that is, it depends on the
level of democracy a country has attained and the existence of an environment conducive to
domestic and international tranquility (Davenport 2007, 9). Only the most stable and
consolidated democracies continuously uphold the fundamental rights of all their citizens (Bueno
5
de Mesquita et al. 2005). As such, the word “reform” to refer to the protection of these rights
may be somewhat of a misnomer because it implies that once a country has achieved the highest
possible human rights score, it cannot revert to a more repressive situation. However,
democracies cum regimes do not automatically protect civil rights; countries have to work
continuously to uphold these rights, a characteristic of all the reforms considered here.
Dahl (1971, 3) himself defined “polyarchy”4 as a political system that provides eight
institutional guarantees: the freedom to form and join organizations, freedom of expression, the
right to vote, the right to compete for political support; and the existence of broad eligibility for
public office, alternative sources of information, free and fair elections, and institutions for
linking government policies to votes and other expressions of citizens’ preferences. As we can
see, five of these guarantees invoke the words “freedom”, “free” or “right”.
Historically, freedom from state coercive behavior went hand in hand with increased
representation and participation (Diamond 1999; Zakaria 2004). In last few decades, however,
many new democracies demonstrate severe deficiencies in their ability to guarantee and protect
various human rights (e.g., Foweraker and Krznaric 2002). In some countries, the government is
technically chosen through popular contests, but civil liberties may not be fully guaranteed or
extensively protected (Freedom House, 2012). The phenomenon led some to refer to these
countries as “illiberal democracies” (Collier and Levitsky 1997, 440; Diamond 1999, 4).5 As a
result, after ascertaining that various rights and liberties are empirically related to many of the
indicators political scientists use to measure democracy, I assess the effects of protest on one
component of this rights measure, workers’ rights, as well as on the full measure of civil rights.
Collective labor rights are considered distinct from more basic human rights such as civil
rights and political freedoms (Mosley and Uno 2007, 924) even though democracies should
6
uniformly guarantee them. Of particular interest to us then is the extent to which some countries
also provide other employment-related protections to workers.6 This leads me to consider
workers rights as instances of reform and to examine them separately from other rights. Workers
rights refer to the freedom of association at the workplace and the right to bargain collectively
with employers, as well as “the prohibition on the use of any form of forced or compulsory labor;
a minimum age for the employment of children; and acceptable conditions of work with respect
to minimum wages, hours of work, and occupational safety and health” (Cingranelli and
Richards 2010).7
Dahl also claimed that there is a logical or definitional correspondence between the eight
institutional guarantees he enumerated and one or both of his political regime dimensions –
contestation and inclusiveness (Coppedge et al. 2008, 633). Contestation refers to the existence
of parties and politicians from which citizens can choose the one(s) most closely aligned with
their preferences. Inclusiveness refers to the ability of as many citizens as possible to participate
in the selection of their leaders and policies.
Dahl’s framework is known as minimalist, that is, as including as few elements as
necessary to make it possible to use analytically. In this conception, democracy is simply a set of
procedures that are certain (known to all) for arriving at outcomes that are uncertain and
collectively binding (Przeworski 1991). Dahl’s framework, however, begs asking to what extent
regime definitions should include characteristics of the political system that transcend the
process of forming a government. As Munck (2009, 124-5) and others have persuasively argued,
empowering the demos to have a periodic say in the making of state policies does not guarantee
these policies will be congruent with the wishes of a majority or even a subset of the people; the
government should also, thorough its actions, remain accountable to the people, particularly in
7
the middle of the electoral cycle when the people are not able to use the ballot box to pass
judgment on their elected representatives.8
The third set of reforms to be evaluated concerns then not human rights, but changes in
political institutions or their operation. Western democratic theory has a strong normative
preference for polities where the executive has parity with or is subordinate to a legislature,
ruling party, and/or other accountability groups.9 The presence of strong presidents that do not
seem very accountable to other branches of government, particularly in Latin America, has led
some to contrast these systems, which O’Donnell (1994) referred to as “delegative democracies”,
with truly representative ones.10 I see this phenomenon, however, as more symptomatic of
diminished accountability than attenuated representation.11 Consequently, I ask to what extent
checks and balances are empirically related to Dahl’s regime dimensions and examine the
possibility that political protest may push democracies in the direction of more political
accountability. I also examine the effect political protest is having on the level of inclusiveness
(or rather the quality of participation) in the world’s democracies. The next session describes the
measures of political protest used in the analysis and their expected effects.
Predictors of democratic reform
The analysis relies on yearly counts of three indicators of mass political protest – antigovernment demonstrations, general strikes, and riots, all derived from Banks (2007) CrossNational Time-Series Data Archive (CNTSDA). Anti-government demonstrations are peaceful
rallies of at least 100 participants; general strikes are coordinated campaigns of disruption
involving at least 1,000 workers and more than one employer; and riots are violent
8
demonstrations by more than 100 citizens.12 For various reasons, these indicators are suitable for
the analysis I present.
First, previous research (Schatzman 2005, 298) has distinguished theoretically and
empirically between political challenges that aim to overthrow the ruling regime, and those that
attempt “to affect policy and leadership, usually through nonviolent means.” The two types of
challenge are usually referred to as “rebellion” and “protest”, with anti-government
demonstrations, general strikes, and riots constituting the protest dimension. Secondly, not only
is the CNTSDA one of the most widely used sources of political data; 13 these three indicators are
also common in studies of social protest, regime change, and political conflict (e.g., Schatzman
2005).
Thirdly, in a world where “the effects of social movements are often indirect, unintended,
and sometimes even in contradiction to their goals” (Giugni 1998, 386), the indicators are
proximate enough to the outcomes we want to assess – democratic reforms, to enable us to draw
valid causal inferences. Demonstrations and general strikes, for example, refer to actions that
target the national/central government explicitly, but exclude rallies whose primary targets are
third parties. At the same time, they are not too proximate to render the analysis tautological. The
measures, that is, do not reduce dissidents’ goals to their participation in one of these events, as
in the case of other contentious events coded by Banks such as guerrilla warfare and revolutions,
allowing us to move analytically from a focus on immediate outcomes to the medium- and longterm consequences of social movement behavior (Giugni 1998).
Finally, the fact that these events are high profile, and that consistency in coding has
been maintained across time and space, means that bias resulting from underreporting of events
not deemed newsworthy by the media is not likely to be very large. Since all the countries
9
analyzed are democracies, the reported event counts should also be fairly accurate. Nevertheless,
this does not mean media reports are completely free of bias and as such I take certain
precautions (which I detail in the empirical section) to ensure that the results are robust to these
potential problems.
Although the analysis stops short of suggesting possible processes or mechanisms that
link political protest to reform14, it does take advantage of historically contingent political
opportunity structures that may condition the effects of protest on society (Giugni 1998, xxi).
Defined as “consistent-but not necessarily formal or permanent dimensions of the political
environment that provide incentives for people to undertake collective action by affecting their
expectations for success or failure” (Tarrow, 1994, 85), the concept of political opportunities is
designed to capture variables that mediate the relationship between direct action on the one hand
and the political system, economy, or society on the other (Koopmans 1999, 105). The chapter
evaluates three variables that may condition the effects of protest on institutional reform: the
level of development, the partisan alignment across and within different branches of government,
and judicial independence. A few paragraphs are in order about the choice of each of these
variables.
Students of contentious politics have identified the level of development as a crucial
mediator of political protest (Dalton et al. 2010). Countries that are more developed have
populations that are more urbanized, educated and self-expressive, all variables that have been
found to lead to a greater incidence of mass protests. Large urban concentrations, for example,
make dissidents easier to recruit and organize, thereby facilitating the formation of opposition
movements (Tilly 1978, 82). Citizens of more developed and urbanized environments are in turn
more educated and autonomous. In so doing, they not only possess the capacity to engage in
10
more frequent and visible forms of collective action, but also the identities that place a premium
on presenting demands to the government and third parties for consideration.
It is also widely accepted that the most democratic countries are the most economically
developed (Aleman and Yang 2011), although the effect of development on democracy appears
to vary by dimension of democracy examined and region of world (Foweraker and Landmand
2004). The measure of development used, derived from the World Development Indicators, is
per capita GDP in current international dollars, measured at purchasing power parity (PPP).15
A second political opportunity structure considered here is the partisan alignment within
and across different branches of government (policy disagreement). Dissidents may find it
easier/more difficult to mobilize when their ideological allies/opponents hold positions in the
various branches of government. Their success may also depend critically on how cohesive
ideologically and institutionally the government is (Johnston 2011, 39-40). The measure of
cohesion I use, the Political Constraints (POLCON) index, is based on a spatial model of
congruence in party control within and across branches of government (Henisz and Zelner 2010).
The index, which ranges from 0 to 1 (0 being the most cohesive, 1 the least), can be used to
indicate how far particular branches of government deviate from an existing policy regime.16
The third and final political opportunity structure I examine is the presence of an
independent judiciary, which some see as central to the concept of the rule of law (Carothers,
2006). An independent judiciary is linked to the existence of liberal democracy because it is
alleged to protect not only individuals from the uneven and arbitrary exercise of state power, but
also minorities from the whims of majorities or unelected veto groups such as the military and
big business. Yet, as Helmke and Rosenbluth (2009) have argued, it is precisely in those
countries where the rule of law and an independent judiciary seem to be most needed where they
11
are harder to find. For data on this institution I rely on Cingranelli and Richards (2010), who
provide a measure of judicial independence that seems to be inspired by de jure considerations of
what a fully independent judiciary would look like while distinguishing between de facto partial
and no judicial independence (Ríos-Figueroa and Staton 2011, 10).
Although an independent judiciary seems to be part of the definition of a constrained
executive (Marshall et al. 2011, 24), the pairwise correlation between these two variables for the
sample of countries I analyze is actually not large (r=0.45; p<0.000). As Helmke and Rosenbluth
(2009, 358) make clear, many poor and newer democracies are characterized by a great deal of
judicial instability even after consolidating their democratic regimes along other dimensions. The
notion of judicial independence also appears to be logically connected to the fragmentation of
power expected of highly competitive political systems and its attendant manifestation, policy
disagreement among and within the different branches of government. While the correlation
between judicial independence and the policy disagreement indicator in the sample analyzed is
statistically significant, it is not very high (r=0.25; p<0.000). Consequently, I make use of both
as mediators of the effect of political protest on institutional reforms. Before proceeding with a
discussion of the estimation strategy and results, I provide a brief description of the control
variables used in the regression models.
Contestation is included as a control in all models except in the model for executive
constraints (which, as a factor analysis indicates, is a component of contestation). A great deal of
variation in political regimes is due to this dimension (Coppedge 2012). Consequently,
contestation is expected to have a positive association with worker’s rights, human rights, and
inclusiveness.
12
Durability controls for the possibility that older democracies score higher in all empirical
measures of reform. Specifically, democratic political regimes engage in fewer violations of
physical integrity rights, particularly the longer a country manages to avoid these violations
(Carey 2008). Durability is a measure of the length of time since the last regime transition
derived from the Polity IV project (Marshall et al. 2011).
Ethnic fractionalization has long been seen as detrimental to democracy. Societal
pluralism can cause different ethnic groups to struggle for control of the state (Alesina et al.
2003, 173), which can result in lower inclusiveness, less constrained executives, and less respect
for human rights. The measure used is an index based on ethno-politically relevant groups from
the Ethnic Power Relations dataset (Weidmann et al. 2009).
Market inequality. Inequality in pre-tax, pre-transfer income has long been considered
detrimental to democratic stability (Houle 2009). Great gaps in the income different classes
receive may create grievances that make the wealthy less prone to tolerate the less well-off in the
political arena. As a result, market inequality could make democratic reforms more difficult to
propose and implement. The measure of inequality used is the pre-tax, pre-transfer Gini
coefficient from the Standardized World Income Inequality Database (Solt 2012).
Oil Rents. Countries that depend on natural resource wealth have long been considered
inauspicious for democratic rule (Ross 2012). The unusually large rents oil generates reduce the
need to tax the population, thereby depriving it of collective representation. Oil rents should also
reduce accountability since the recipients of these rents tend to be state owned enterprises (with
the exception of oil companies in the United States, which are in private hands). I use an
indicator of oil rents from the World Development Indicators database that measures the
13
difference between the value of crude oil production at world prices and total costs of production
(as a percentage of GDP).
Globalization. There is no scholarly consensus on the effect of trade and capital flows on
democracy (Coppedge 2012, 301-2). More openness to trade is known to improve labor
conditions in particular (Flanagan 2006), but less is known about its effects on other aspects of
democratic governance. I use a measure of economic globalization developed by Axel Dreher
(2006) that weighs financial and trade flows as a percentage of GDP.17
The following section introduces the methodology used.
Empirical approach
The approach I follow is to first derive a series of indicators of political regimes that can
be used as a baseline in the subsequent analysis of democratic reform. Following Dahl (1971), I
conceive of regimes as exhibiting a certain number of dimensions that are empirically verifiable
and that make some countries minimally democratic. Some of these dimensions can be used
subsequently to explain particular reforms if we suspect that there is a relationship between the
two. Alternatively, individual components of these dimensions can be identified as indicators of
particular reforms and evaluated further. As the reader can surmise, my approach is heavily
inductive as I seek to build on the theoretical and empirical work of scholars that have thought
about how best to conceptualize and measure political regimes.
In a widely cited study, Coppedge et al. (2008) confirmed the existence of Dahl’s (1971)
two regime dimensions – contestation and inclusiveness – for the post World War II period.
Using exploratory factor analysis, the authors examined the most commonly used indicators of
political regimes and found that for most years, they indeed clustered in two clearly distinct yet
14
correlated empirical dimensions which they identified as “contestation” and “inclusiveness”. The
most recent time period they examine, 1981-2000, overlaps to some extent with the time period
in my analysis (1994-2007), but one of the variables they factor-analyzed, Bollen’s extent of the
suffrage indexing the percentage of the population twenty years or older that has the right to vote
in national elections, did not extend past the year 2000. Nevertheless, I created a similar measure
by calculating the voting age population as a percentage of the overall population, and was
indeed able to reproduce the results in Coppedge et al. (2008) using all fourteen variables they
included.18 Similar to Coppedge et al. (2008), who were able to explain 71% of the variance in
their data, the overall variance explained by my analysis was 69%.
I then made a few adjustments to their analysis. First, I dropped one of the variables they
used, the ordinal indicator of civil liberties from Freedom House, and added Freedom’s House’s
indicator of freedom of the press. My rationale for doing this is that the civil liberties measure is
a comprehensive inventory not only of basic freedoms such as the freedoms of expression,
association, and assembly, but also of institutions designed to uphold these freedoms such as an
independent judiciary and the rule of law. Freedom House also includes in this measure property
and economic rights that are best examined as outcomes rather than included de jure in a regime
definition. The adjustments, however, should not change the results too much since the civil
liberties indicator includes information on freedom of expression, association, and assembly that
is already coded separately.19
Next, I added to this list a number of variables tapping into other rights and freedoms
that, if upheld and protected, should render democracies more representative, participatory,
accountable, and egalitarian. These include the rights to physical integrity aggregated as an
index, worker’s rights, the right to freedom of religious worship, women’s social rights, and
15
women’s economic rights, all derived from the CIRI Human Rights Data Project (Cingranelli and
Richards 2010). Although these rights are all enumerated in the Freedom House ordinal measure
of civil liberties, it is important to analyze them separately since Dahl himself did not draw links
between particular rights and his two regime dimensions. The overall variance explained was
still approximately 68%. Table 1 presents the rotated factor loadings and unique variances for the
nineteen variables examined.
TABLE 1 HERE
As Table 1 indicates, three factors – which I have named “Human rights”,
“Contestation”, and “Inclusiveness”, are clearly recognizable in the data. The factors can be
easily interpreted since they delineate readily noticeable patterns among the variables. The first
factor, for example, groups most of the civil and political rights that serve to distinguish
democracies from autocracies such as the freedoms of speech, association, and electoral selfdetermination, as well as the rights to physical integrity, workers’ rights, and freedom of
religion. The factor loadings imply that it is impossible for governments to engage in extensive
civil rights violations without also abrogating some political rights. The high loadings of freedom
of the press and the competitiveness of participation on this factor indicate that these are both
features of the political system that help guarantee and uphold these rights.20
Component number 2, which I have labeled “Contestation”, correlates highly with
features of the political system that help make political participation more competitive such as
having the smaller parties capture a greater share of the votes cast in parliamentary or
presidential elections (witness the high loading of competition on this factor), or an executive
that is subordinate to or faces parity with other branches of government (executive constraints).
16
The variables indexing openness and competitiveness of executive recruitment also load highly
on this factor.21
Finally, the third factor groups the variables suffrage, women’s political rights,
participation, women’s economic rights, and women’s social rights. The latter two are the only
rights not significantly loading in the “Human rights” dimension and it is noticeable in this
regard that they correlate strongly with other measures of participation. We live in times when
suffrage restrictions based on gender, race, educational or other ascriptive qualifications have
become increasingly rare. As a result, my interpretation of these patterns is that by granting
women economic, social and political rights, countries elevate their status vis-à-vis that of men
and hence ensure greater and perhaps qualitatively more meaningful political participation.
The factors also correlate moderately with one another, their pairwise correlations
ranging from 0.42 in the case of factors 1 and 2, to 0.51 in the case of factors 1 and 3.
Furthermore, with the exception of women’s political rights and workers’ rights, which have
relatively high uniqueness scores, all other variables seem very relevant to the analysis. Worker’s
rights will of course merit its own analysis apart from other human rights, the latter being
defined empirically by the factor scores from the first dimension.
This initial phase of the analysis helps me then generate scores for the two regime
dimensions previously discussed, contestation and inclusiveness, and an additional dimension –
human rights. The factor analyses allow me thus to verify the validity of selecting two individual
components from these factors – workers’ rights and executive constraints – for more in-depth
analysis. The scores from the exploratory factor analysis can also be used as independent
variables in the subsequent stage of the analysis.
17
Workers’ rights is originally coded as an ordinal variable with three categories – “[a]
score of 0 indicates that workers’ rights were severely restricted; a score of 1 indicates that
workers’ rights were somewhat restricted; and a score of 2 indicates that workers’ rights were
fully protected during the year in question” (Cingranelli and Richards 2010). “Executive
constraints” is originally a seven-point categorical variable, but in our analysis only categories 5,
6, and 7 are present. Category 5 is reserved for countries in which “the executive has more
effective authority than any accountability group but is subject to substantial constraints by
them” (Marshall et al. 2011, 25); level 6 is an intermediate category, and 7 is reserved for
countries in which there is executive parity or subordination to other groups or branches of
government. Since these categories are not numerous enough to approximate a continuous
variable but the distances separating them are not precisely known, I created two new
dichotomous variables indicating full protection of workers’ rights (versus the presence of some
restrictions), and full constraints on the executive (versus the presence of some executive
prerogatives) respectively.
Estimation and Results
In determining the population of countries suitable for the analysis, I decided to employ a
qualitative judgment of what countries would at least qualify as minimally or procedurally
democratic. According to the Polity IV codebook, 5 is the first value at which there are
substantial limitations on executive power. To be a democracy then, a country had to have a
value of 1 in Cheibub’s dichotomous coding of democratic regimes and a value of 5 or more in
Polity’s measure of executive constraints (“xconst”). Finally, I also excluded those observations
where authority characteristics are not stable due to civil war, foreign occupation, or a protracted
18
regime transition. If a Polity score of 6 or greater, a commonly used metric, is used instead to
delineate the relevant population, the sample only increases by 4 observations, providing
confidence in the procedure used to select the population. Table 2 presents a list of countries
included in the analysis.
TABLE 2 HERE
Turning now to the empirical analysis, the regression equations estimated are of two
kinds: random effects logit models for the two dichotomous variables – workers’ rights and
executive constraints – and OLS with random effects for human rights and inclusiveness. The
choice of logit in the case of workers’ rights and executive constraints is dictated by the nature of
these dependent variables.22
For the OLS models, coefficients have been standardized so as to be able to compare
their magnitudes directly, reflecting the number of standard deviations a dependent variable will
change, per standard deviation increase (decrease) in the predictor variables. Since the models
include many stock variables, I expect most of the variation they explain to be cross-sectional
rather than time-series.
OLS models estimated on panel data are prone to three familiar threats to valid causal
inference: serial correlation, contemporaneous correlation, and heteroskedasticity. I use a robust
cluster estimator for the covariance matrix to correct standard errors in the presence of any
pattern of heteroskedasticity and correlation of the errors within units. I also add year fixed
effects to control for any common shocks countries may experience that would cause
contemporaneous correlation in the errors. As a check on the robustness of the results, I also
estimate mixed effects logit and OLS models that more appropriately control for another source
19
of heterogeneity in the data: country-specific, time-invariant unobserved factors. These factors
are usually modeled using varying intercept parameters for countries.
Multiple intercepts are usually interpreted as offsets from a dropped group intercept. A
more efficient method, however, is to model these intercepts directly by placing a common
distribution around them. By employing varying intercepts for countries, variation in country
effects is estimated conditional on the data and parameters of the model. This partial pooling is
particularly desirable for unbalanced panels since it allows more accurate estimates of country
effects. Partial pooling also alleviates the problem of slow-moving or completely time-invariant
predictors that are correlated with the country effects (Shor et al., 2007:168).
Finally, the models control for endogeneity between social protest and worker’s rights
on the one hand, and social protest and executive constraints on the other. An argument could be
made that protests, rather than the cause of more or less protection for workers and more or less
constrained executives, could be a response to a lack of rights or heavy-handed executives.
Although this endogeneity cannot be completely removed, I address the problem by lagging antigovernment demonstrations, general strikes, and riots by one year in these two models. I also lag
the variables policy disagreement and judicial independence in the model for executive
constraints, removing their interactions with demonstrations, strikes, and riots. This also
removes from the analysis the possibility that, in response to a heavy-handed executive, the
judiciary acts less independently or citizens vote in a divided government. Table 3 presents the
results of the four regression models estimated.23
TABLE 3 HERE
20
As the R^2s in Table 3 indicate, the models for human rights and inclusiveness explain
approximately two-thirds of the variance in the data. Twice as much of that variation is crosssectional than time-series in the first model, thrice as much in the second. The results for the
most part support the hypothesized relationships, particularly for the stock variables expected to
control for the historical tendency of certain countries to be more democratic than others. The
one major exception is the negative and very significant relationship between contestation and
all three dependent variables it helps predict. A lowess smoother plot (not shown) indicates that
contestation is for the most part positively associated with inclusiveness; not so, however, for the
relationship between contestation and human rights. This, moreover, is not simply an artifact of
pooling established and new democracies, as the relationship holds among established
democracies as well (driven in the latter case by countries such as Colombia, Venezuela, India,
and Israel). This finding merits further investigation.
Since the analysis includes both stock and flow variables for seventy-eight countries,
we would expect the results to be weighed against significant effects of protest activity on
democratic reforms, particularly since the protest event variables have been interacted with
features of the political opportunity structure. When either a continuous or dichotomous variable
is interacted with another continuous or dichotomous variable, the coefficient on the individual
variables return their effect when the other variable is 0. This combination of values is
technically implausible in the case of GDP per capita since this variable by definition cannot
take the value of 0. Consequently, we can go ahead and focus on statistically significant findings
(for interactions with GDP per capita) even in the absence of statistical significance for the
individual coefficients. The results indicate that higher levels of development help democracies
become less repressive in the presence of demonstrations and more inclusive in the presence of
21
strikes, but more repressive in the presence of riots. Riots, however, increase respect for human
rights in the presence of judicial independence. Riots in the previous year and governments that
are more ideologically fragmented are also associated with less constrained executives in the
following year. An independent judiciary in the previous year, however, helps safeguard
executive parity in the current year.
I next performed robustness checks using mixed effects models (not reported) for all
dependent variables except executive constraints.24 All results are robust to country specific
unobserved factors except the finding that general strikes increase the level of inclusiveness in
more developed countries, which loses its significance.
Conclusion
This chapter has systematically examined the contribution of social protest to
democratic reforms around the world. The effect of protest on the democracy can be small and
inconsistent, as the results make clear. As expected, other variables seem more important to the
quality of democracy around the world. Nevertheless, in at least two areas, social protest is
having profound effects in the quality of democracy: respect for human rights, and executive
constraints. Peaceful protests and general strikes are prompting countries to increase respect for
human rights and executives to behave less heavy-handedly, particularly in the more developed
democracies. Not surprisingly, riots can sometimes diminish the quality of democratic
governance by prompting governments to engage in human rights violations or executives to
behave more heavy-handedly. The upside is that judicial independence can significantly prevent
governments from repressing riots.
22
With respect to the political opportunity variables that can be expected to condition the
effect of social protest on democracy, the results are also in line with expectations and augur well
for the future of democratic governance. By the standards laid out in this chapter, countries
become more democratic, that is, they guarantee and protect human rights more and
circumscribe the power and prerogatives of the executive more, the more developed their
economies and the more independent their judiciaries. Somewhat surprisingly, workers rights
seem to be more immune to these effects, tending to increase in response to a divided
government and the absence of large income inequalities, oil rents, and ethnic fractionalization.
It would have been encouraging to find evidence of social protest leading to more respect for
worker’s rights and more inclusiveness, which was not the case. Nevertheless, the results
presented in this chapter demonstrate that social protest is helping realign democratic politics
around the world. They also call for more attention to the precise mechanisms linking social
protest to more respect for human rights and more constrained executives.
23
Table 1. Dimensions of Political Regimes, 1994-2007
Component1 Component2
Component3
Uniqueness Source
Variable
0.6661
0.2612
0.1546
0.176
Freedom of the press
Freedom House
-0.5831
-0.4895
-0.0714
0.1012
Political rights
Freedom House
0.5478
0.2604
0.1796
0.3367
Competitiveness of participation
Polity IV
0.4201
0.6621
0.0202
0.1305
Executive constraints
Polity IV
-0.2299
-0.5878
-0.2108
0.2783
Regime
Cheibub
0.6955
0.1939
-0.0005
0.3662
Freedom of speech
CIRI Human Rights
0.4126
0.5496
0.0421
0.2966
Competition
Vanhanen
0.7428
0.2105
-0.0904
0.3507
Freedom of association
CIRI Human Rights
0.6324
0.3936
-0.0556
0.2897
Freedom of electoral self-determination
CIRI Human Rights
0.23
0.8339
-0.0121
0.1031
Competitiveness of executive recruitment
Polity IV
-0.1137
0.3006
0.7363
0.2595
Suffrage
IDEA
0.1782
-0.1034
0.5346
0.6427
Women's political rights
CIRI Human Rights
-0.1604
0.2379
0.7755
0.3006
Participation
Vanhanen
-0.2496
0.8678
0.0629
0.3266
Openness of executive recruitment
Polity IV
0.5583
-0.165
0.3834
0.434
Physical integrity rights index
CIRI Human Rights
0.5399
-0.1019
0.2937
0.5255
Workers' rights
CIRI Human Rights
0.8677
-0.0783
-0.2403
0.4336
Freedom of religion
CIRI Human Rights
0.0382
-0.0605
0.7999
0.3717
Women's economic rights
CIRI Human Rights
0.0967
-0.0137
0.8099
0.2667
Women's social rights
CIRI Human Rights
53
8.46
7
Variance explained (%)
Notes: Scores calculated using data for all countries regardless of their political regime (autocracy or democracy). High loadings,
indicative of belonging to a factor, are in bold. Uniqueness is the variance that is not shared with other variables, that is, an indicator
of how relevant the variable is to the factor model (the more unique, the less relevant).
24
Table 2. Countries included in the analysis
Albania
Argentina
Armenia
Australia
Austria
Bangladesh
Belgium
Benin
Bolivia
Brazil
Bulgaria
Canada
Chile
Colombia
Costa Rica
Croatia
Czech Republic
Denmark
Dominican Republic
Ecuador
El Salvador
Estonia
Finland
France
Georgia
Germany
Germany West
Ghana
Greece
Guatemala
Honduras
Hungary
India
Indonesia
Ireland
Israel
Italy
Jamaica
Japan
Kenya
Korea South
Kyrgyzstan
Latvia
Lithuania
Macedonia
Mexico
Moldova
Mongolia
Nepal
Netherlands
New Zealand
Nicaragua
Nigeria
Norway
Pakistan
Panama
Papua New Guinea
Paraguay
Peru
Philippines
Poland
Portugal
Romania
Senegal
Slovakia
Slovenia
Spain
Sri Lanka
Sweden
Switzerland
Thailand
Trinidad and Tobago
Turkey
Ukraine
United Kingdom
United States
Uruguay
Venezuela
25
Table 3. Determinants of Political Reforms, 1994-2007.
Independent variable
Contestation
Durability
Fractionalization
GDP per capita
Market inequality
Oil rents
Globalization
Demonstrations
General strikes
Riots
GDP*demonstrations
GDP*strikes
GDP*riots
Policy disagreement
Disagreement*demonstratio
ns
Disagreement*strikes
Disagreement*riots
Judicial independence
Independence*demonstratio
ns
Independence*strikes
Workers’
Rights
-9.354***
(1.457)
0.019
(0.015)
-3.691*
(1.947)
0.000
(0.000)
-0.118***
(0.042)
-0.134**
(0.060)
0.024
(0.018)
-0.015
(0.619)
0.252
(1.418)
1.821
(1.271)
-0.000
(0.000)
0.000
(0.000)
-0.000
(0.000)
3.842**
(1.787)
0.711
Human
Rights
-0.892***
(0.081)
0.079*
(0.041)
-0.136**
(0.064)
0.152***
(0.048)
0.023
(0.025)
-0.172**
(0.070)
0.186***
(0.045)
0.537
(0.396)
0.127
(0.298)
-1.371**
(0.631)
0.047***
(0.018)
-0.002
(0.014)
-0.044*
(0.023)
0.002
(0.032)
-0.020
(1.191)
-2.886
(3.071)
-7.875**
(3.134)
2.088***
(0.633)
-0.170
(0.025)
-0.020
(0.020)
0.048*
(0.027)
28.193***
(4.698)
-4.554
(0.024)
-0.008
(0.017)
0.002
(0.027)
12.354*
(6.620)
3.840
(3.368)
-0.929
(2.462)
(3.809)
-1.500
(2.469)
(0.347)
0.217
(0.884)
26
Executive
Constraints
-0.046*
(0.027)
0.265
(4.392)
0.001***
(0.000)
-0.054
(0.050)
-0.285***
(0.084)
0.049*
(0.028)
0.521*
(0.313)
0.660
(0.527)
0.140
(0.322)
-0.000*
(0.000)
-0.000
(0.000)
-0.000
(0.000)
Inclusiveness
-0.610***
(0.095)
0.158***
(0.055)
-0.305***
(0.088)
0.294***
(0.073)
-0.143***
(0.036)
-0.076
(0.072)
0.165***
(0.050)
-0.444
(0.451)
0.175
(0.299)
-0.468
(0.457)
0.001
(0.025)
0.020*
(0.010)
0.010
(0.022)
-0.024
(0.036)
-0.034
Independence*riots
1.012
(0.687)
11.092**
(5.310)
Policy disagreement t−1
-3.981**
(1.570)
1.068**
(0.453)
0.099
(0.131)
0.335
(0.264)
-0.386**
(0.160)
-0.028
(0.047)
Judicial independencet−1
Demonstrationst−1
General strikest−1
Riotst−1
4.073
(3.785)
0.250
(0.170)
-0.098
(0.358)
-0.249
(0.205)
Economic growtht−1
0.68
0.62
𝑅2
Number of countries
76
76
78
76
Number of observations:
731
731
1411
731
Notes: Coefficients reported in columns 1 and 3; standardized coefficients in columns 2 and 4.
Standard errors in parenthesis in columns 1 and 3; robust cluster standard errors in columns 2
and 4. Models in columns 2 and 4 also include year fixed effects (not reported). All models
include a constant (not reported).
27
References
Alemán, José, and David D. Yang. 2011. "A Duration Analysis of Democratic Transitions and
Authoritarian Backslides." Comparative Political Studies 44 (9):1123-51.
Alesina, Alberto, Arnaud Devleeschauwer, William Easterly, Sergio Kurlat, and Romain
Wacziarg. 2003. "Fractionalization." Journal of Economic Growth 8 (2):155-94.
Banks, Arthur S., and Kenneth A. Wilson. 2007. Cross-National Time Series Data
Archive. Jerusalem, Israel: Databanks International. Retrieved from
http://www.databanksinternational.com.
Bueno de Mesquita, Bruce, George W. Downs, Alastair Smith, and Feryal Marie Cherif. 2005.
"Thinking Inside the Box: A Closer Look at Democracy and Human Rights." International
Studies Quarterly 49 (3):439-57.
Carey, Sabine C. 2008. "The use of Repression as a Response to Domestic Dissent." Political
Studies 58 (1):167-86.
Carothers, Thomas. 2006. Promoting the rule of law abroad: In search of knowledge.
Washington, D.C.: Carnegie Endowment for International Peace.
Casper, Gerhard. 2007. "Caesarism in Democratic Politics: Reflections on Max Weber”. Social
Science Research Network. Retrieved from
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1032647.
Cingranelli, David L., and David L. Richards. 2010. "Short Variable Descriptions for Indicators
in the Cingranelli-Richards (CIRI) Human Rights Data Dataset." Retrieved from
http://ciri.binghamton.edu/documentation/ciri_variables_short_descriptions.pdf.
Collier, David, and Steven Levitsky. 1997. "Democracy with Adjectives: Conceptual Innovation
in Comparative Research." World Politics 49 (3):430-51.
Coppedge, Michael. 2012. Democratization and research methods: The methodology of
comparative politics. New York: Cambridge University Press.
Coppedge, Michael, Angel Alvarez, and Claudia Maldonado. 2008. "Two Persistent Dimensions
of Democracy: Contestation and Inclusiveness." Journal of Politics 70 (3):632-47.
Dahl, Robert Alan. 1971. Polyarchy: Participation and opposition. New Haven, CT: Yale
University Press.
———1989. Democracy and its critics. New Haven, Connecticut: Yale University Press.
Dalton, Russell, Alix van Sickle, and Steven Weldon. 2010. "The Individual–Institutional Nexus
of Protest Behaviour." British Journal of Political Science 40 (1):1-23.
28
Davenport, Christian. 2007. State repression and the domestic democratic peace. New York:
Cambridge University Press.
Diamond, Larry. 2008. The spirit of democracy: The struggle to build free societies throughout
the world. New York: Times Books/Henry Holt & Company.
Diamond, Larry J. 1999. Developing democracy: Toward consolidation. Baltimore: Johns
Hopkins University Press.
Doorenspleet, Renske. 2005. Democratic transitions: Exploring the structural sources of the
fourth wave. Boulder: Lynne Rienner Publishers.
Dreher, Axel. 2006. "Does Globalization Affect Growth? Evidence from a New Index of
Globalization." Applied Economics 38 (10):1091-110.
Flanagan, Robert J. 2006. Globalization and labor conditions: Working conditions and workers
rights in a global economy. New York: Oxford University Press.
Foweraker, Joe, and Roman Krznaric. 2000. "Measuring Liberal Democratic Performance: An
Empirical and Conceptual Critique." Political Studies 48 (4):759-87.
———. 2002. "The Uneven Performance of Third Wave Democracies: Electoral Politics and the
Imperfect Rule of Law in Latin America." Latin American Politics and Society 44 (3):29-60.
———. 2003. "Differentiating the Democratic Performance of the West." European Journal of
Political Research 42 (3):313-40.
Foweraker, Joe, and Todd Landman. 2004. "Economic Development and Democracy Revisited:
Why Dependency Theory is Not Yet Dead." Democratization 11 (1):1-20.
Giugni, Marco. 1999. "Introduction: How Social Movements Matter: Past Research, Present
Problems, Future Developments." In How social movements matter, eds. Marco Giugni,
Doug McAdam and Charles Tilly. University of Minnesota Press: Minneapolis.
Giugni, Marco G. 1998. "Introduction: Social Movements and Change: Incorporation,
Transformation, and Democratization." In From contention to democracy, eds. Marco G.
Giugni, Doug McAdam and Charles Tilly. Lanham, MD: Rowman and Littlefield.
Helmke, Gretchen, and Frances Rosenbluth. 2009. "Regimes and the Rule of Law: Judicial
Independence in Comparative Perspective." Annual Review of Political Science 12 :345-66.
Henisz, Witold J., and Bennet A. Zelner. 2010. "The Political Constraints Index (POLCON)
Dataset." Retrieved from http://www-management.wharton.upenn.edu/henisz/.
Houle, Christian. 2009. "Inequality and Democracy: Why Inequality Harms Consolidation but
does Not Affect Democratization." World Politics 61 (4):589-622.
29
Inglehart, Ronald, and Christian Welzel. 2005. Modernization, cultural change, and democracy:
The human development sequence. Cambridge, New York: Cambridge University Press.
International Institute for Democracy and Electoral Assistance. 2012. "Unified
Database." Retrieved from http://www.idea.int/resources/databases.cfm.
Johnston, Hank. 2011. States and social movements. Malden, MA: Polity Press.
Kitschelt, Herbert P. 1986. "Political Opportunity Structures and Political Protest: Anti-Nuclear
Movements in Four Democracies." British Journal of Political Science 16 (1) (January):5785.
Koopmans, Ruud. 1999. "Political. Opportunity. Structure. Some Splitting to Balance the
Lumping." Sociological Forum 14 (1):93-105.
Levitsky, Steven, and Luncan A. Way. 2002. "Elections without Democracy: The Rise of
Competitive Authoritarianism." Journal of Democracy 13 (2):51-65.
Lijphart, Arendt. 1993. "Constitutional Choices for New Democracies." In The global
resurgence of democracy, eds. Larry Diamond, Marc F. Plattner. Baltimore, MD: Johns
Hopkins University Press.
Mackin, Anna E. 2008. "Protest and Liberal Democratic Performance: A Panel Analysis".
Retrieved from http://www.nuffield.ox.ac.uk/politics/papers/2008/Mackin.pdf.
Marshall, Monty G., Keith Jaggers and Ted Robert Gurr. 2011. Polity IV project: Political
regime characteristics and transitions, 1800-2010, dataset users' manual. College Park:
Center for International Development and Conflict Management, University of Maryland.
Mosley, Layna, and Saika Uno. 2007. "Racing to the Bottom or Climbing to the Top? Economic
Globalization and Labor Rights." Comparative Political Studies 40 (8):923-48.
Nam, Taehyun. 2007. "Rough Days in Democracies: Comparing Protests in
Democracies." European Journal of Political Research 46 (1):97-120.
O'Donnell, Guillermo A. 1994. "Delegative Democracy." Journal of Democracy 5 (1):55-69.
Przeworski, Adam. 1991. Democracy and the market: Political and economic reforms in Eastern
Europe and Latin America. Cambridge: Cambridge University Press.
Puddington, Arch. 2012. Freedom in the world 2012: The Arab uprisings and their global
repercussions. Washington, D.C.: Freedom House. Retrieved from
http://www.freedomhouse.org/article/freedom-world-2012-arab-uprisings-and-their-globalrepercussions.
30
Julio Ríos-Figueroa and Jeffrey K. Staton, "An Evaluation of Cross-National Measures of
Judicial Independence". Retrieved from
http://userwww.service.emory.edu/~jkstato/papers/RiosStaton.pdf.
Ross, Michael L. 2012. The oil curse: How petroleum wealth shapes the development of
nations, . Princeton: Princeton University Press.
Schatzman, Christina. 2005. "Political Challenge in Latin America: Rebellion and Collective
Protest in an Era of Democratization." Journal of Peace Research 42 (3):291-310.
Solt, Frederick. 2009. "Standardizing the World Income Inequality Database." Social Science
Quarterly 90 (2):231-42.
Tarrow, Sidney. 1994. Power in movement: Social movements, collective action, and politics, .
New York: Cambridge University Press.
———1998. Power in movement: Social movements and contentious politics. 2nd ed.
Cambridge: Cambridge University Press.
Tilly, Charles. 1978. From mobilization to revolution. Reading, Mass.: Addison-Wesley Pub. Co.
———2007. Democracy. Cambridge: Cambridge University Press.
Wimmer, Andreas, Lars-Erik Cederman, and Brian and Min. 2009. Ethnic Power Relations
Dataverse. Retrieved from http://dvn.iq.harvard.edu/dvn/dv/epr.
Zakaria, Fareed. 2004. The future of freedom: Illiberal democracy at home and abroad. New
York: W. W. Norton & Company.
31
1
The late twentieth- and early twenty-first century brought democracy to a greater number of countries than ever
before (Doorenspleet 2005). Since 1974, political protest has played a role in producing democratic transitions.
Witness, for example, the color revolutions in the post-communist world (Diamond 2008, 191) and the Arab Spring
that began in North Africa in 2011 and is still being felt throughout the Middle East.
2
For a regime to be truly democratic, citizens also have to be protected from arbitrary state action and engage in
binding consultation on state policies and personnel.
3
It should be clear by now that I am using the terms ‘equitable’ and ‘emancipative’ interchangeably.
4
Dahl used the term polyarchy – rule by the many – as a synonym for democracy, which he regarded as an ideal
political regime. Democracies, Dahl implied, must “have substance, quality, and meaning. They must, over time,
hear people’s voices, engage their participation, tolerate their protests, protect their freedoms, and respond to their
needs” (Diamond 2008, 292).
5
The distinction between illiberal and liberal democracies originated from the observation that during the latter part
of the Third Wave, “human rights violations have become widespread in countries that are formally democratic”
(Diamond 1999, 34). Countries where competition to determine who rules does not ensure outcomes such as “high
levels of freedom, equality, transparency, [and] social justice” (Diamond 2008, 23) are also referred to as formal or
electoral democracies (Inglehart and Welzel 2005; Munck 2009). Some refer to democracies that ensure these
outcomes as ‘liberal democracies’ (Freedom House 2012). Others go still further when they claim that corruption
can prevent a government from being institutionally effective and hence accountable to the people (Inglehart and
Welzel 2005, 154). They thus speak of ‘effective democracy’ as a combination of liberal democracy and an
environment free of corruption. For a full definition of the term ‘electoral democracy’ and a comparison with
‘liberal democracy’, see the methodological appendix in Freedom House’s “Freedom in the World 2012” report,
available at http://www.freedomhouse.org/report/freedom-world-2012/methodology.
6
Surprisingly, out of the 731 country-year observations in my dataset, 83 belong to countries that do not provide the
most minimal protection for workers’ rights according to Cingranelli and Richards’ (2010) coding.
7
The complete definition of this and other human rights examined here can be found at
http://ciri.binghamton.edu/documentation/ciri_variables_short_descriptions.pdf.
8
Dahl’s thinking has changed in this respect over the years. In Polyarchy, he seemed to privilege accountability
over representation when he stipulated that “government policies depend on votes and other expressions of
preference”. In Democracy and Its Critics, the requirement changed to the view that control over government
decisions about policy be vested constitutionally in elected officials (Dahl 1989, 221).
9
This is evident in how one of the most popular indicators of political regimes, the Polity IV project, codes
countries on one of their authority characteristics, executive constraints. See Polity IV Project: Political Regimes
Characteristics and Transitions, 1800-2010 Dataset Users’ Manual, page 25, available at
http://www.systemicpeace.org/inscr/p4manualv2010.pdf.
10
Other terms used to describe this phenomenon are hyper-presidentialism and “caesarism” (Casper 2007).
11
See Mackin (2008, 14), for example, for a similar interpretation.
12
For precise definitions of these variables, see http://dss.princeton.edu/access_data/codebooks/banks.pdf.
13
http://www.databanksinternational.com/
14
One possible mechanism through which a social movement organization may bring about change is through its
incorporation in the policymaking apparatus of the state. See Giugni et al. 1998 on this point.
15
http://data.worldbank.org/data-catalog/world-development-indicators.
16
More precisely, the index is calculated using the number of independent branches of government with veto power
over policy change combined with measures of heterogeneity in party composition within the legislative and
executive branches and congruence in party control across branches. A veto player is an individual, party, or
political institution that can stop a change in policy from the status quo. See Henisz and Zelner (2010) for a more in
depth discussion of this index. One of these veto players, the legislature, has been identified as a crucial explanator
of cross-national variation in protest across Western European democracies (Nam 2007). According to Kitschelt
(1986: 63), “the capacity of legislatures to develop and control policies independently of the executive” is a critical
measure of its power. Since a weak legislature is unable to act as in intermediary between citizens and the
32
government, dissidents in countries with a weak legislature cannot easily expect access to the government for the
purposes of making their demands. Therefore, they are more inclined to protest than citizens in countries with a
strong legislature. Although sympathetic to this view, I also note that the power of legislatures vis-à-vis executives
and their role in motivating social protests cannot be divorced from the larger question of the effect of executive
constraints on the quality of democracy. Furthermore, since executive constraints are correlated with policy
alignment between and within partisan and constitutional veto players, I opt to use the latter to try to explain
variation in the former.
17
http://globalization.kof.ethz.ch/
18
My data on voting age populations comes from the International Institute for Democracy and Electoral Assistance
(IDEA 2012). http://www.idea.int/vt/. In the interest of saving space, I do not include definitions of the other
thirteen variables. For descriptions of the remaining variables, see Coppedge et al. (2008).
19
See Cingranelli and Richards’ (2010) Human Rights Data Project.
20
The political rights index is negatively signed because higher values imply fewer rights.
21
Cheibub’s categorical regime variable is negatively signed because higher values imply more autocratic,
specifically less constrained, political regimes.
22
For a one-unit increase in contestation, for example, the expected change in log odds of workers’ rights is -9.35.
23
Because O’Donnell (1994) attributes the phenomenon of delegative democracy in part to the severity of the
socioeconomic problems that newly installed democratic governments have inherited, I also control for lagged
economic growth in the model for executive constraints.
24
It was not possible to achieve convergence for this model.
33