Uneven States, Unequal Societies, and Democracy’s Unfulfilled Promises: Citizenship Rights in Chile and Contemporary Latin America

Abstract

In contemporary Latin America, deep-seated social discontent with political elites and institutions has been, paradoxically, the counterpart of democratic stability and resilience. This paradox suggests that scholarly assessments of democracy are, at least partially, at odds with citizens’ own views of democracy. This article thus develops a framework to describe citizens’ everyday experience with civil, political, and social entitlements associated with democracy. It introduces the framework by analyzing the structural underpinnings of democratic discontent in Chile and then applying it to the analysis of perceived citizenship entitlements in 18 countries, using the AmericasBarometer data. Significant variance is observed across time and both across and within countries. The descriptive findings also imply that only a (declining) minority of Latin American citizens feel fully entitled to civil, political, and social citizenship rights. We advocate the need to bring the demand side of democracy back to the analysis of democratic shortcomings and crises.

Freedom will be instrumental … for seeking justice, all types of justice, that of ordinary law, but also social justice…. Democracy has greater value than that of a mere formula for making power legitimate, because with democracy not only do we vote, but we also eat, educate, and heal.

—President Raúl Alfonsín, Inauguration Speech, 1983.

Since democratization in the 1980s, the promise of democracy in Latin America has meant more than just the enactment of liberal democratic institutions. Raul Alfonsin’s famous 1983 inauguration speech reflects well what Latin Americans hoped for at the time. After four decades, liberal democratic regimes have endured in the region, which is a remarkable accomplishment. Popular support for democracy, notwithstanding recent marginal decreases, also appears stronger than in the past, as do expert judgments about the quality of democracy in the region (see, e.g., V-Dem 2022). Against the backdrop of democratic recession or backsliding elsewhere (Plattner 2015; V-Dem 2022), Latin American democracies look healthier than ever. ¹

At the same time, however, the region has observed cyclical turmoil, as well as marked processes of political alternation (toward neoliberalism in the 1990s, toward the left in the 2000s, toward the right in the mid-2010s, and, apparently, again toward the left in the 2020s.) These processes of alternation have largely destroyed traditional party systems in the region without having institutionalized new ones. Massive corruption, exposing “state capture” by political and business elites, has also erupted in recent years, engendering a series of political scandals that led to the early termination of presidents in Brazil and Peru. Public confidence in political and state institutions is at a low ebb (Cohen et al. 2017). And even if support for democracy remains high and stable, satisfaction with the way democracy works has declined sharply (Lupu et al. 2021).

Moreover, widespread riots and social unrest have suddenly erupted across the region, challenging the political class (Brazil 2013, Chile 2019, Colombia 2019, Ecuador 2019) and even the survival of democracy, with the political turmoil that led to Evo Morales’s resignation in Bolivia (2019). Recent elections in Peru (as well as in the most stable democracy in the region, Uruguay) displayed rising levels of political system fragmentation.

Deep-seated social discontent with incumbent political elites and political institutions has been, paradoxically, the counterpart of democratic stability and resilience. This paradox suggests that scholarly assessments of democracy and its quality are, at least partially, at odds with citizens’ own views, experiences, and frustrations with the promises of democracy. In addition, available explanations of persistent discontent with democracy in the region suffer several important weaknesses.

Recent political turmoil (e.g., large-scale, violent rioting) has taken place across dissimilar social and political contexts, thus defying simplistic explanation. On the one hand, riots took place against both leftist (Brazil 2013, Bolivia 2019) and rightist governments (Chile 2019, Colombia 2019, Ecuador 2019). On the other hand, rioting took place in cases with a recent legacy of mass mobilization (Ecuador and Bolivia) and weakly institutionalized political systems, as well as in cases with more institutionalized political systems and intermediate (Chile) to low (Brazil and Colombia) levels of mass mobilization in recent years. Moreover, rioting took place both in relatively affluent (Brazil, Chile, and Colombia) and poor (Bolivia, Ecuador) countries. Rioting did not take place in many other countries, several of which shared structural conditions and recent socioeconomic trajectories (Campello and Zucco 2016) similar to those in which serious crises took place.

Likewise, “state capture” by business and political elites, as well as corruption, undoubtedly have fostered social discontent with “the political class,” as it has done elsewhere (Przeworski 2019). In that context, another frequent culprit of democratic recession or backsliding has also been influential: social media were pivotal in disseminating information about political scandals and in providing organizational shortcuts for mobilization in the streets (Valenzuela et al. 2016). Yet corruption scandals and social media availability are jointly present in cases that did not witness massive rioting at the national level; these factors’ causal link to the type of rather spasmodic and anomic rioting observed in contemporary Latin America is, at best, partial. In a nutshell, Latin America’s recent experience with democratic discontent defies easy explanation and displays specific characteristics that challenge sweeping arguments about its drivers.

This article asserts that Latin Americans’ persistent discontent with democracy lies in the inability of formal democratic regimes (which enact political rights) to produce tangible parallel advances toward the enactment of a fuller and relatively evenly distributed package of civil and social rights. In other words, formal democratization has at least partially failed to bring about substantive democratization. ² In this regard, we rely on T. H. Marshall’s classical typology of citizenship rights (1992 [1950]). We also follow the lead of O’Donnell (2010, 180), who argued that to be meaningful, and thus valued by society, democracy requires agency, which, in turn, requires citizens to have access simultaneously to a “full package” of basic civil, political, and social rights. ³

In recent years, democratic stability and social progress have jointly occurred in ways that significantly limit citizens’ agency. Latin American democracies are today “violent” democracies, due to the expansion of organized crime and the escalation of criminal violence (Arias and Goldstein 2010; Bergman 2018). Moreover, inequalities, even if reduced, remain deep and multifaceted. Contemporary Latin American democracies are, as a result, characterized by a situation in which significant segments of the population lack social rights and the most basic civil rights (i.e., security and equal access to the judiciary system).

Segmented and partial access to citizenship rights thus provides a critical lens through which to address institutionally driven views on democratic quality. High levels of social inequality and weak and uneven state capacity call into question one of the two main assumptions of mainstream scholarship on democracy; namely, that the “experience of citizenship at any given point in time [and space] is, for analytical purposes, more or less constant among citizens of a given state” (Yashar 2005, 50). ⁴

We therefore propose to shift focus to the demand side of politics and to examine citizens’ perceptions of their access to the full package of political, civil, and social rights. Those perceptions, we shall claim, reflect the complex effects of two important contextual conditions that shape citizens’ everyday experience with democracy across the region: persistent social inequalities and weak and uneven state capacity. Those two conditions create significant unevenness in the ways access to social and civil rights is actually “lived” and also “perceived.” Unevenness in the access to civil and social rights, if successfully politicized, eventually renders the basic promise of democracy substantively empty for significant segments of the population.

Since state weakness and unevenness, as well as severe inequality, are relatively constant across time, we shall also emphasize the contingent nature of the process that leads to their politicization in each society. Enduring and multifaceted inequalities and unequal access to the rule of law and basic civil rights can coexist (and indeed, have empirically coexisted) with democracy for a long time. Yet persistently uneven citizenship rights eventually fuel discontent. The politicization of unevenness and the eruption of discontent can also eventually occur after a period of significant social progress.

In the last decade and a half, for instance, economic growth and social policy expansion translated into significant social progress (ECLAC 2018), and even into tangible reductions of social inequality in most countries. This social progress is the result, in part, of significant and unprecedented gains in state capacity to reach the poor and the informal sector, both in urban peripheries and in rural areas (CEPAL 2018). However, recent social incorporation might have made privilege and durable inequalities more politically salient. As is classically argued, relative deprivation (Runciman 1966) and progressive deprivation (Gurr 1970) produce rising discontent precisely in contexts of objective amelioration.

In sum, people’s enduring uneven access to civil and social rights, and cross-sectional differences related to such unevenness, eventually become politicized. Such politicization can occur even in the context of institutionalized democracies and even after sustained social progress, which might, paradoxically, help increase the salience of those differences in shaping political discontent. Depending on citizens’ recent experience with the supply side of politics (i.e., party system characteristics and trajectory), such politicization can occur either against incumbents (and in favor of intrasystemic challengers) or against the entire political system (and in favor of outsiders). Citizens’ perceptions of their access to political rights might be a proxy for systemic support or, conversely, for a demand for “antisystem” alternatives. The “anatomy” of the demand side of politics, and specifically citizens’ perceptions of their access to the full package of civil, social, and political entitlements, is therefore critical for advancing understanding of contemporary social discontent with democracy.

This article proposes an analytical and empirical framework focused on describing citizens’ everyday experience with democratic entitlements. It first motivates our proposed framework by looking at the territorial distribution of civil, social, and political rights in contemporary Chile. That country can be conceived as a “least likely case” for observing significant variance in citizens’ perceived civil, political, and social entitlements, due to the joint presence of a relatively strong and territorially even state apparatus, sustained economic growth and social progress over the last few decades, and long-term democratic stability and institutionalization.

Subsequently, we introduce our framework, discussing its theoretical and conceptual underpinnings. We then illustrate the usefulness of this framework by drawing on AmericasBarometer data to construct indicators of citizen perceptions of political, civil, and social rights. We discuss our operationalization and its current limitations and introduce our descriptive results. Our empirical analysis, even if far from ideal, lends significant support to the proposed analytical framework. We close by addressing theoretical accounts of contemporary democratic discontent and by outlining challenges that remain for advancing understanding of the demand side of current crises of democracy.

Short Distances, Huge Inequalities

Weak and uneven state capacity, as well as high levels of social inequality, importantly affect the operation of liberal democracies. Especially in highly segregated urban areas, such as those in which the great majority of contemporary Latin American citizens live, the effects of these factors are readily observable, even at a very short distance. Those effects condition citizens’ life prospects across a host of relevant dimensions (Kaztman 2021).

Even in Chile, a case seen until recently as a possible model for development in the region (Belaisch et al. 2005; Mainwaring and Scully 2008), those inequalities run deep. Indeed, the social outbreak of 2019 in Chile can be thought of as emerging from the politization of those inequalities. Figures 1 and 2 display the top 20 results returned by a Google Images search for two neighborhoods in Metropolitan Santiago, Bajos de Mena (in the municipality of Puente Alto) and Vitacura. Although the images seem to depict two completely different “societies,” the physical distance between them is less than 15 miles. While citizens residing in the neighborhood depicted in figure 2 are more likely to travel abroad than to visit the neighborhood depicted in figure 1, residents of the latter neighborhood are likely to commute daily to work in affluent areas of the city.

Figure 1. First 20 Google Image Search Results for “Bajos de Mena”

Source: Authors’ construction on the basis of google.com (retrieved March 25, 2019).

Figure 2. First 20 Google Image Search Results for “Vitacura”

Source: Authors’ construction based on google.com search results (retrieved March 25, 2019).

Twitter, through the GNIP application, also provides a simple way to visualize the scope and shape of segregation. This information is also interesting because it reflects citizens’ relative political engagement, as Twitter is usually considered the preferred social network for debate among political elites (Blank 2017). GNIP maps additionally provide an approximate estimate of socioeconomic segregation by plotting the type of device that predominates in each area of the city. ⁵ Red dots in figure 3, panel a correspond to tweets originating from (more expensive) iOS devices, while green dots were generated by either Android or alternative operating system devices. In Santiago, as in other major Latin American cities, segregation by socioeconomic status is striking. Let us zoom in to the two neighborhoods depicted in figures 1 and 2 (panels b and c in figure 3). Quite clearly, people in Bajos de Mena do not tweet very often; and the green-dominated space observed there contrasts with the reddish, and very active, Vitacura.

Figure 3. Twitter Activity by Type of Device Used in Metropolitan Santiago (panel a), Bajos de Mena (panel b), and Vitacura (panel c)

Source: Authors’ construction based on Twitter/GNIP API (retrieved March 26, 2019).

The contrast between the two “societies” has systematic implications for access to citizenship rights for the residents in each area. Figure 4 displays geocoded data for proxies related to civil, political, and social rights. Panels 4a and 4b plot the density of the top ten elite last names and of 117 Mapuche last names (Bro and Mendoza 2021) in the Metropolitan Region of Santiago. ⁶ The Mapuche ethnic group is the largest indigenous minority in Chile, and it is overrepresented among the country’s poorest citizens. In turn, the top ten elite last names are significantly overrepresented among Chilean political elites. ⁷ Panel 4c jointly depicts the average socioeconomic status of each area in the city—classified as high or low—and hotspots where police and judicial procedures related to narcotrafficking frequently take place. It is worth noting that enforcement does not necessarily overlap with the prevalence of trafficking, which also occurs in wealthy areas of the city, where profit margins and demand are higher. In poorer areas, enforcement is also contingent on the eventual collusion of police forces with different gangs. Panel 4d displays observed voter turnout in Chile’s presidential and congressional elections in 2017, focusing on citizens aged 18 to 29, an age cohort that exhibited a steady decrease in turnout from the late 1990s until 2020.

Figure 4. Geographical Distribution of Proxies for Civil, Political, and Social Rights in Metropolitan Santiago, Chile

Source: Authors’ construction based on maps drawn by Juan Correa Palma (a, b, c) and the Centro Producción del Espacio UDLA (d), using Bro‘s dataset (2020) and Bro and Mendoza (2021) for maps a and b; data from INE (2018), Centro de Inteligencia Territorial UAI (2013), and Fiscalía Nacional (2017) for map c; and data from SERVEL (2013) for map d.

To be sure, these are rough and incomplete proxies for the territorial distribution of civil (exposure to violence related to narcotrafficking), political (levels of engagement in political debate, turnout rates), and social rights (socioeconomic welfare and its interaction with salient ethnic categories) in contemporary Santiago. However, these maps reflect sizable inequalities in citizens’ access to civil, political, and social citizenship rights in contemporary Chile. We can infer from this set of maps the following patterns. First, poor citizens and ethnic minorities live in less secure areas of the city. Second, they tend to vote much less frequently than do their wealthier fellow citizens. Third, the rich are overrepresented among members of Congress, while poor citizens lack political representatives who look and live as they do and have needs similar to theirs.

Although inequalities are sizable across neighborhoods, significant differences also exist at a much closer range. When researching microtrafficking dynamics in Santiago’s shantytowns, one often comes across statements like the following, which reflects gaps in the access to basic civil rights: “Carabineros [the police] raid my house all the time. They enter and destroy things, scare the kids, destroy their toys. My kids hate carabineros. Yet they always go against the weak. They never raid the home of the guy I buy [drugs for reselling] from, who lives three blocks from here. They don’t touch that guy.” ⁸

Multifaceted inequalities also become more apparent in Chile’s protest outbreak of 2019. Whereas police forces were relocated to protect upscale areas of the city, popular sectors were left largely unprotected. In that context, riots were concentrated in the periphery of the city. In those same areas, between October and the end of November 2019, nearly two hundred instances occurred of citizen groups stoning police stations (La Tercera 2019).

During the COVID-19 pandemic in 2020 and 2021, differential access to social rights also became readily apparent. If the first recorded cases were concentrated among the upper sectors of society (who imported the disease during the summer holiday season), lower-class neighborhoods soon became the leading sites in terms of the number of cases and fatalities. The COVID fatality rate in people younger than 40 was three times higher in Santiago’s poorer neighborhoods than in its wealthier ones (Mena et al. 2021).

Also during the pandemic, Chile’s health minister declared that the government’s highly criticized quarantine strategy failed to account for the amount of overcrowding in poor households, due to the lack of information on living conditions in the country’s shantytowns (La Tercera 2020). At the same time, an interviewee dwelling in a shantytown in Santiago stated, “We take everything they give us, but the food boxes that the narcos give away are better. They include fish, for instance…. Even though they help everyone, they know well who needs more. They help old people, for instance, by buying them medicine and so on.” ⁸

Politics also works very differently across socioeconomic groups. Besides the differences in citizens’ turnout patterns and participation in public debate, political parties and representation also work differently when relating to either poor or wealthy neighborhoods. Whereas political campaigns are more programmatic and conducted via broadcast media in wealthy areas of the city, campaigns are more personalistic and clientelistic in poor neighborhoods (Luna 2014).

The configuration and distribution of citizenship rights in Chile stands in contrast to the country’s role, until recently, as a possible model for democracy and development in contemporary Latin America. Since its return to democracy, Chile has seen a sharp decline in poverty rates. Even though the country remains one of the most unequal in Latin America—Chile’s 0.45 Gini coefficient in 2017 was the sixth-highest of 18 countries in the region—it has become less unequal since 2000, when its Gini coefficient was 0.485. Moreover, considering V-Dem’s five democracy indexes (deliberative, egalitarian, electoral, liberal, and participatory) featured in figure 5, Chile consistently ranks above the Latin American average for most of the 1990–2021 period.

Figure 5. Chile and Latin America in Five V-Dem Democracy Indexes

Source: Authors’ construction based on Twitter/GNIP API (retrieved March 26, 2019).

Chile’s democratic trajectory, according to expert scholarly assessments, is obviously inconsistent with the country’s highly segmented and unequal access to civil, social, and political citizenship rights described here. The protest outbreak of 2019, in our view, indicates the shortcomings of overly institutional and top-down evaluations of democracy, which fail to account for democracy’s unfulfilled promise to provide more substantive citizenship entitlements. That unfulfilled promise is especially salient in highly unequal societies and in contexts of weak and uneven state capacity. The remainder of this paper proposes a framework for elucidating how democracies “actually” work in that type of setting and what drives contemporary democratic discontent.

Conceptual Approach: Perceived Entitlements to the Full Citizenship Package

Although extant works theorize the perils of disjointed citizenship rights and unequally distributed democratic agency—(see, e.g., PNUD 2004; Yashar 2005; O’Donnell 2010), empirical assessments of democracy remain blind to this theoretical perspective. This study begins to tackle that deficiency by providing a first descriptive measure of unequal agency in contemporary Latin American democracies, taking T. H. Marshall’s citizenship categories as a guide for operationalizing the measure. Moreover, in what we consider a proof of concept exercise, the study implements an operationalization strategy that can identify territorial and cross-sectional unevenness in citizens’ perceived access to civil, political, and social rights.

We are not interested in validating the temporal sequencing of rights proposed by Marshall, which has been challenged both for advanced capitalist societies (Mann 1987) and for Latin American countries (O’Donnell 2010). Indeed, O’Donnell (2010) claims that social incorporation in some Latin American countries predated civil and political incorporation, thus reversing the British sequence stylized by Marshall. ⁹ Rather than focusing on sequence, we seek to analyze the cross-sectional distribution of citizenship rights in contemporary societies. In this way, we seek to detect variance in citizens’ perceived access to the “full package” of citizenship rights. Our results indicate that the hypothesized preconditions for democratic agency to be fully realized are largely absent in the region.

Our conceptual approach accords with Somers’s criticism (1993) that Marshall conflates rights-as-status with rights-as-practice. Somers’s seminal critique opened the way for relational approaches to citizenship (see, e.g., Tilly 1996; Heller and Evans 2010; Bertorelli et al. 2017; Appadurai 2001), which focus on the interaction between citizenship, the state, and inequality. However, a relational approach to citizenship does not necessarily measure effective access to citizenship, which is, of course, normatively and substantively crucial for individuals’ life chances.

Instead, we measure perceived access to citizenship rights and its fluctuations over time. We claim that these perceptions, at the aggregate level, might be even more pivotal than effective access to citizenship rights in causing political mobilization and democratic discontent. While the gap between perceptions and objective reality might be seen as a weakness of our approach, we claim that it is actually one of our approach’s major strengths.

Operationalization Strategy

There is no universal principle that determines what those rights and duties shall be, but societies in which citizenship is a developing institution create an image of an ideal citizen against which achievement can be measured and toward which aspiration can be directed.

—T. H. Marshall, 1992 [1950], 19

The AmericasBarometer includes a series of indicators that allows us to study individuals’ perceptions about their access to different citizenship rights. On that basis, this study presents a series of descriptive results for Latin American societies. Our operationalization strategy is based on a latent variable approach to each citizenship type: social, civil, and political. We therefore built linear indexes for each type by drawing on a set of potential (partial) indicators available in the AmericasBarometer sample.

Because the scale reliability of these indicators is reasonable (.77 for political citizenship, .64 for social citizenship, and .72 for civil citizenship), we used a principal components algorithm to compute individual scores that recover the shared variance in our indicators. For that analysis, we drew on the pooled dataset, which yielded an N of 113,056 observations from 18 countries that pertain to the 2012, 2014, 2016, and 2018 pooled sample of the AmericasBarometer. The obtained country-level results were generally stable across time.

A general assumption underpinning our operationalization strategy is that respondents’ assessments of their access to different citizenship rights involve personal experiences more than general evaluations of the country as a whole. We justify this assumption based on our data, which indicate the presence of significant within-country variance across our set of indicators. Such variance, which is systematically associated with territorial and individual-level factors, should be less pronounced if general evaluations have predominated over (or have significantly blurred) personal experiences concerning access to citizenship rights.

The specific codes and question phrasings can be found in table 1. For social citizenship, we drew on two survey questions that are available for every country included in LAPOP: respondents’ evaluations of the quality of public health and education. The Cronbach’s alpha for the two indicators we used is .64, and the principal factor solution yields a unique meaningful factor (see table A1 in the online appendix). Although we considered including other variables, such as the household receipt of government transfers, the diverse nature of transfer programs across the region and their socioeconomic targeting goals might introduce biases into a cross-national comparison of the kind we are pursuing. That said, assessing social citizenship on the basis of satisfaction with state-provided health and education services carries its own perils. On the one hand, not all respondents interact on a daily basis with publicly provided services. Their opinions might thus reproduce prejudice and stigma, particularly in unequal societies. On the other hand, quality assessments rely heavily on subjective expectations. The interpretation of our results should therefore be sensitive to these factors.

Table 1. Indicators for Types of Citizenship

Source: LAPOP 2014–2018.

To operationalize political citizenship, we drew on four indicators but discarded voter registration, due to the impact of different laws across countries. The selected indicators tap into respondents’ trust in electoral processes, in the executive, in the congress, and in political parties as the main representative agents in a democracy. The question phrasings are shown in table 1. The Cronbach’s alpha for the four indicators we used is .77, and the principal factor solution yields a unique meaningful factor (see table A1 in the online appendix).

Again, this operationalization might seem inappropriate for gauging the effectiveness of political rights. In that regard, indicators of effective political disenfranchisement would be ideal for assessing Marshall’s notion of political citizenship. Given their absence, however, we think a measure that recovers citizens’ trust in electoral processes, agents, and the political system as guarantor of their rights can act as a possible surrogate. ¹⁰ The interpretation of our results regarding other measures, however, must be sensitive to its own possible biases.

To operationalize civil citizenship, we combined three indicators of perceived access to justice and a general measure concerning the extent to which citizens feel that the political system works to protect their basic rights. In other words, our chosen index represents a good mixture of respondents’ more abstract sense of having equal access to justice and the functioning of the rule of law. The question phrasings are shown in table 1. The Cronbach’s alpha for the four indicators we used is .72, and the principal factor solution yields a unique meaningful factor (see table A1 in the online appendix).

Marshall’s notion of citizenship rights is linked to the very emergence of the nation-state as the locus of legitimate authority and the implications of this emergence for the end of privilege (Bendix 1964). In this regard, an explicit measure of equality before the law could have been used to assess civil citizenship. However, responses to such a question are arguably contextually politicized. We therefore decided to operationalize civil rights by using indirect assessments of equality before the law.

For certain analyses, such as the computation of subnational information, we also built linear additive indexes for each citizenship dimension. ¹¹ The factor scores predicted for political and social citizenship correlate with their corresponding linear additive index at about 0.9. ¹² Working with the retained factors for each citizenship dimension and their highly correlated linear additive indexes, we then computed subnational indexes (at the provincial level) for each country. On that basis, we used multilevel regression and poststratification MrP (Kastellec et al. 2010) to estimate municipal-level citizenship indexes for Chilean municipalities included in the AmericasBarometer sample. Furthermore, to illustrate the observed levels of subnational variation, we relied on MrP to estimate variance at the regional and municipal level for Chile. (Further details and full documentation are available in the online appendix.) The results of both subnational analyses suggest that perceived access to citizenship rights is highly “localized” in contemporary Latin American societies.

If variance by locality is high, factor solutions and scalability tests are extremely stable. Individual country analyses for two randomly chosen cases (Chile and Mexico) replicate the results obtained for the entire region almost exactly. The same holds for a replication sample that combines observations from five other randomly chosen countries.

On the basis of these highly reliable factor scores, we ran a series of k-means clustering procedures. A solution with eight groups best achieved the twin goals of representing different relevant citizenship packages (distinct combinations of civil, political, and social citizenship) and also representing a sizable segment of the respondents. The online appendix shows additional figures for these clusters, the scores and citizenship distribution of each cluster, the average frequency distribution of the clusters, and the frequencies for the different waves. ¹³

As an additional robustness check, we divided samples for the three randomly chosen cases into two groups, contingent on interview dates (first half of the data collection vs. last half of the interviews). In each case, and for each “half” sample, the seven clusters obtained essentially the same relative membership (i.e., the percentage of cases classified into each of the seven groups was virtually the same). Moreover, with only one exception, (the case of Brazil), cluster membership was also stable within cases across time when comparing the classifications obtained for 2012, 2014, 2016, and 2018. Figure A1 in the appendix illustrates our operationalization and modeling strategy.

Results 1: Perceived Access to Citizenship Rights in Contemporary Latin America

Our results suggest that only a small minority (15 percent) of Latin American citizens perceive that they have access to the “full citizenship” package (see figure B2 in the online appendix for frequencies plot). A similarly small minority (12 percent) holds the extreme opposite view, perceiving that they lack access to all three types of citizenship. In other words, 73 percent of Latin Americans, most of whom live in stable democratic regimes, perceive themselves to enjoy only partial access to modern citizenship rights. Figure 6 displays all the possible combinations of access to citizenship rights, as well as the observed combination in contemporary Latin America. ¹⁴

Figure 6. Citizenship Packages and Citizens’ Perceived Access to Them in Contemporary Latin America

Source: Authors’ construction based on AmericasBarometer 2012, 2014, 2016.

Another frequently observed (12.1 percent) citizenship package is the one representing citizens who perceive themselves as having access to political and social citizenship rights while lacking proper access to civil citizenship rights. The polar opposite group—that is, those who feel an entitlement exclusively to civil citizenship rights—accounts for 9.9 percent of the sample. Two other groups also are defined as feeling entitled to only one of the three types of citizenship rights: while 14.4 percent perceive themselves to be entitled to political citizenship only, an additional 13.6 percent feel entitled only to social rights.

The remaining groups that complete the set of possible combinations are two. In the first group, 14.3 percent of respondents perceive themselves to be entitled to social and civil rights but do not perceive themselves to be politically enfranchised. Meanwhile, 8.7 percent of respondents perceive themselves to be entitled to political and civil citizenship rights while lacking social incorporation.

To summarize, we find widespread empirical support for the notion that contemporary political regimes operate in a context in which citizens perceive uneven access to basic citizenship rights. These combinations are arguably connected to citizens’ experiences (and the expectations those experiences generate over time) with politics and with different facets of the state that are in charge of granting social and civil citizenship entitlements across time and space.

Figure 7 presents the distribution of citizenship packages by country for the pooled (2012–18) sample. The figure illustrates the existence of significant country differences in citizens’ perceived entitlements. For instance, Brazil, a country that has recently witnessed political crises, and Haiti have the highest incidence of perceived “empty citizenship.” Black and mulatto citizens are a sizable group in both countries, and, as shown in figure B4 in the appendix, ethnic minorities are overrepresented in the empty citizenship package. Uruguay, Guyana, Ecuador, and Costa Rica, by contrast, show a relatively high percentage of respondents who perceive themselves to enjoy the “full citizenship” package. Yet the perceived lack of civil citizenship entitlements is very high in three of those four countries (Uruguay, Ecuador, and Costa Rica), arguably resulting from the recent but highly visible increase in threats to individual safety and property observed in each case. The relative incidence of that package (i.e., lacking civil citizenship) is also high in the Dominican Republic, Argentina, and Paraguay.

Figure 7. Distribution of Perceived Access to Citizenship Packages by Country

Source: Authors’ construction based on AmericasBarometer 2012, 2014, 2016, 2018.

While Chile, Bolivia, and Peru display the highest percentage of respondents in the “political citizenship only” category, the “civil citizenship only” group is especially prevalent in Haiti, Brazil, Colombia, and Bolivia. The category “social citizenship only” is especially prevalent for Paraguay, the Dominican Republic, and Honduras. Perceived lack of political citizenship is especially prevalent in Costa Rica, Ecuador, and Uruguay.

To be sure, we do not claim that perceived access to citizenship rights conforms to objective conditions in each country. Our claims in that regard are limited to two descriptive premises. First, there is substantial longitudinal and country-level variance in citizens’ perceptions of their citizenship entitlements. Second, those perceptions might result from the interaction between objective conditions related to the provision of each citizenship right and citizens’ evolving expectations (i.e., “running tally”) regarding their day-to-day experiences with politics and the state.

Results 2: Within-country Variation and the Case of Chile

Given our opening discussion of the Chilean case, we extend our analysis to patterns observed in this case. Here, we explore subnational variance a bit further by relying on MrP estimates of municipal-level variance in the Metropolitan Region of Chile. Chile is the Latin American country most consistently identified as having the most capable and territorially even state apparatus (see, e.g., Kurtz 2013; Soifer 2015). Moreover, it is a country that has seen significant improvements in social incorporation, reducing poverty and extreme poverty since the transition to democracy by record amounts in the Latin American context. Chile is, however, a case in which inequality and its correlation with territorial segregation is among the highest in Latin America, especially in its urban areas (ECLAC 2016, 70).

Figure 8 displays the evolution over time of perceived access to citizenship packages for the 2012–2016/7 period. (Information on the evolution of each citizenship index is included in the appendix.) As we can observe in the figure, the empty citizenship group more than doubled in size during that period. A parallel yet inverse evolution is observed regarding the full citizenship package, which shrank in size. Interestingly, while the perceived lack of social citizenship declines sharply over time, the percentage of those who feel socially entitled but perceive a lack of access to civil and political citizenship increased significantly over time.

Figure 8. Distribution over Time of Perceived Access to Citizenship Packages in Chile

Source: Authors’ construction based on AmericasBarometer 2012, 2014, 2016, 2018.

Let us now focus on territorial variance in perceived access to citizenship rights. The three scatterplots presented in figure 9 display the bivariate correlations among each pair of citizenship rights at the municipal level in the Metropolitan Region of Santiago. Although access to different types of citizenship rights is always positively and moderately correlated (especially in the case of perceived entitlements to civil and political citizenship rights), each scatterplot displays many municipalities that significantly deviate from the linear trend. Nevertheless, by focusing on, for instance, the relatively strong correlation observed between perceived civil and political citizenship, it is possible to identify substantively meaningful trends.

Figure 9. Correlations of Citizenship Packages at the Municipal Level in Santiago, Chile

Source: Authors’ construction based on AmericasBarometer 2014.

Observed variance is not perfectly correlated across citizenship dimensions. The residents of some districts perceive relatively more access to one type of citizenship than to other types. For instance, wealthier municipalities (e.g., Las Condes, Vitacura) and more rural municipalities (e.g., Talagante, Melipilla) display a relatively greater level of perceived political citizenship than expected, according to their levels of perceived civil and social citizenship. Meanwhile, the poorest urban municipalities display consistently weaker perceived entitlements to both types of citizenship (e.g., La Pintana, Puente Alto, Conchalí, etc.). Therefore, subnational variance is probably not stochastic but related to different socioeconomic configurations (e.g., levels of inequality, levels of territorial segregation), as well as to longitudinal trajectories relating to key determinants of perceived social and civil political rights, such as the evolution of state capacity and social incorporation throughout the country’s subnational units.

To sum up, our evidence suggests that perceived access to citizenship rights varies significantly across time and space, even in a country commonly perceived as a democratic exception in the region. Even if the fit between our conceptual categories and the empirically available indicators for each citizenship right leaves a lot to be desired, our results suggest that different types of rights theoretically associated with democracy are very differently perceived by citizens in contemporary Latin America (and probably elsewhere).

Discussion

Empirically, our results reveal great heterogeneity in the perceived distribution of citizenship rights in Latin America. First, only a small minority of citizens perceive themselves to have access simultaneously to civil, political, and social rights; that is, the “full package” of democratic citizenship. Second, the cluster analysis, as well as the presence of significant variance among country averages for each dimension, suggests that each national case presents specific configurations in relation to each particular type of citizenship right. Third, notwithstanding the study’s important limitations given available sample sizes, the subnational exploration also suggests the need for future research to analyze and understand subnational patterns and socioeconomic correlates that shape perceived access to each citizenship type.

We tentatively suggest two broad sets of hypotheses. On the one hand, the citizen perceptions we track empirically might relate to their objective conditions that influence effective access to entitlements. For instance, at the individual level, ethnic minorities, historically subject to discrimination, perceive themselves to be entitled to their formal civil, political, and social rights significantly less than do nonminorities. Moreover, if we jointly consider national and subnational drivers of citizenship rights, we might hypothesize that two objective conditions shape citizens’ access to civil, political, and social rights in ways that are consistent with the subnational variance we have detected. Those objective conditions are (territorialized) socioeconomic inequality and the uneven territorial reach of state institutions at the subnational level, as driven by the interaction between state infrastructural capacity and the presence of local challengers (such as local gangs with territorial control) who constrain the state’s monopoly on coercion. In this regard, we could posit the following type of explanation for social unrest: while political incorporation stabilized and created more room for protest activity and contestation (eventually delegitimizing political elites), civil and social incorporation were perceived as stagnant or declining. Since unrest is driven by socioeconomic inequalities and vulnerabilities and by states’ incapacity to assert a monopoly of coercion in (highly populated) areas of their territory, our proposed explanation traces back to classic works by Karl Polanyi (2001) and Max Weber (1978).

On the other hand, respondents’ perceived entitlements might be driven by their expectations and subjective experiences with citizenship rights. In this regard, the expansion and duration of democracy in Latin America might have raised expectations for civil and social citizenship entitlements beyond states’ capacities to intertemporally fulfill them. We know that individuals’ life trajectories and expectations shape their evaluations of different outcomes, which might, in turn, drive how they experience the citizenship rights formally associated with democracy. For instance, “fear of crime,” triggered by press coverage of violent events in the country, might also shape citizens’ perceptions about their access to rights in ways that seem detached from objective indicators. Moreover, economic crises and corruption scandals should also play a strong role in delegitimizing political representation and key democratic institutions and agents.

Objective or subjective drivers of perceived citizenship rights might be more or less relevant for different types of entitlement. For instance, one could argue that state capacity at the local level might increase the provision of social policies and public security, thus being pivotal for inducing more favorable perceptions toward access to civil and social rights. By contrast, political citizenship entitlements might be more closely determined by subjective orientations and processes, related to a political system’s capacity to provide meaningful and legitimate representation to citizens. If this were the case, perceived political citizenship might also display less territorial variation than perceived civil and social citizenship.

Appropriately addressing and testing these alternative accounts requires moving beyond public opinion analysis. Better and more accurate descriptive assessments also require a different empirical approach, probably combining analysis of “objective” indicators and in-depth fieldwork. For now, we have attempted only to illustrate citizens’ perceptions of uneven access to citizenship rights in contemporary Latin America, while speculating on the implications of such unevenness for assessments of democracy and its contemporary perils.

Conclusions

This article has proposed a conceptual and descriptive contribution that yields new evidence on citizens’ perceptions of citizenship rights in contemporary Latin America. How does such evidence speak to the region’s recent political trajectory? Although we have purposely avoided causal claims, we conclude by speculating on the possible drivers of citizens’ widely divergent perceptions of their citizenship entitlements. In general, we hypothesize that the configurations we detect in our empirical analysis are contingent on each country’s trajectory in terms of state capacity, socioeconomic trajectory, and political representation structures. ¹⁵ Those trajectories are not easy to capture because they are subject to both temporal and subnational variation.

Subnational heterogeneity is produced by the overlap between territorial and socioeconomic segregation and the unevenness of state capacity at the local level. Temporally, case trajectories are neither univocal nor linear. Consider, for instance, the case of state capacity. In recent years, the expansion of conditional cash transfer programs seems to have expanded access to social citizenship in most of the region. Access has also expanded into each country’s peripheral areas, which were historically characterized by state weakness or its complete absence. At the same time, the expansion of organized crime and a widespread fear of crime have deteriorated states’ perceived capacity to grant basic civil rights to citizens. This is particularly the case in metropolitan areas, where citizenship was most fully granted in the past.

We have also argued that bringing the demand side back into the analysis of democracy and its discontents is crucial to understand contemporary crises of democracy. Democratic capitalism is in crisis; as a result, assessments regarding democratic recession or backsliding have mushroomed (Diamond 2015; Lührmann and Lindberg 2019). However, those assessments often concentrate on procedural and formal institutional dimensions and focus on top-down factors, such as the behavior of political elites (Diamond and Plattner 2015). Our results might eventually be useful for developing more complete assessments of the particular drivers of democratic discontent in other advanced capitalist societies that have seen increasing social inequality (Piketty 2013) and increasing territorial unevenness (Hooghe et al. 2016).

To be consequential (especially at the national level), the demand side of politics requires political aggregation. This is especially the case given the subnational differences we have identified in terms of citizens’ perceived entitlements to their rights. In other words, similar cross-national “demand sides” (in terms of countries’ observed distribution across “citizenship packages” perceived to be available to their citizens) might either aggregate at the national level and produce similar political outcomes across countries or remain atomized at the local level (in some or all cases) by virtue of social and territorial fragmentation.

The demand side is thus a necessary but insufficient element for explaining national-level outcomes, which also need to be explained by incorporating the supply side of politics. The interaction between demand and supply is pivotal for understanding which inequalities and grievances get mobilized and aggregated at the national level and who effectively translates them into the institutional arena (i.e., either systemic or antisystemic actors). Norris and Inglehart (2019), for instance, provide an explanation for democratic backsliding in capitalist advanced societies by looking at elites’ strategizing in a context shaped by the cultural backlash induced by the rise of the postmaterialist “silent revolution.” Although this type of process might be related to the political activation of conservative groups (such as Evangelicals) that have played an important role in contemporary Latin American cases, such as Brazil and Bolivia (see Boas 2013; Smith 2019), those groups and their political allies have not played a decisive role in other cases.

In other instances, such as contemporary Chile, specific government decisions triggered protest that eventually grew into an antiestablishment movement, via the activation of a mobilization cascade catalyzed by the (negative) synergy between demand and supply side actions (Granovetter 1978). In this case, no leadership has yet been able to articulate and aggregate the demand side. Recent rioting in Bolivia and Ecuador, in turn, represents a more traditional interaction between civil society movements in the opposition and an incumbent government facing declining popularity. In still other cases, such as contemporary Argentina, Mexico, and Uruguay, the political system may have been able to electorally channel demand for change, with social discontent remaining latent and protest fragmented at the local level, still without reaching nationally visible outcomes.

Although our conclusions remain tentative, we expect our results to render the analysis of the demand side of democratic governance more visible and eventually better integrated into current top-down assessments of the nature and deficits of contemporary democracies. In a social context in which experience with citizenship is segmented and territorialized, as well as being patterned by different groups’ experience with democracy and the state, our ability to capture regime dynamics through a handful of aggregate indicators is probably falling short.