The Welfare State amid Crime: How Victimization and Perceptions of Insecurity Affect Social Policy Preferences in Latin America and the Caribbean *

Abstract

Criminal violence is one of the most pressing problems in Latin America and the Caribbean, with profound political consequences. Its effects on social policy preferences, however, remain largely unexplored. This article argues that to understand such effects it is crucial to analyze victimization experiences and perceptions of insecurity as separate phenomena with distinct attitudinal consequences. Heightened perceptions of insecurity are associated with a reduced demand for public welfare provision, as such perceptions reflect a sense of the state’s failure to provide public security. At the same time, acknowledging the mounting costs and needs that direct experience with crime entails, victimization is expected to increase support for social policies, particularly for health services. Survey data from twenty-four Latin American and Caribbean countries for the period 2008–12 show that perceptions of insecurity indeed reduce support for the state’s role in welfare provision, whereas crime victimization strongly increases such preferences.

Keywords

crime social policy preferences victimization insecurity perceptions Latin America

One of the most important changes in the Latin American and Caribbean context in the last decade has been the dramatic increase of criminal violence. According to the 2014 survey of the Latin American Public Opinion Project, one-third of Latin Americans identified crime as their country’s main problem.¹ Such violent reality has already had important attitudinal and behavioral consequences across the region, from diminished support for democracy and trust in democratic institutions to increased support for iron-fist policies and varying effects on political participation.²

Prevailing evidence therefore suggests that crime-related violence shapes citizens’ views on politics and, subsequently, their support for various state policies and their engagement with democratic institutions. The literature analyzing the effect of crime on citizens’ policy perceptions has mainly focused on security issues. However, it is likely that crime also shapes citizens’ views of state policies beyond those related to security.³

In this article, we examine how criminal violence affects citizens’ perceptions of the state’s role regarding welfare policies. Exploring this issue is crucial given that attitudes toward redistribution are relevant for understanding patterns of welfare provision and influential in the policymaking process.⁴ Furthermore, as exposure to certain contextual conditions can either critically enhance or depress support for social policy provision among certain segments of the population,⁵ understanding the impact of crime is imperative. We argue, however, that in order to unravel the impact of crime on welfare preferences, we need a disaggregated approach that distinguishes insecurity perception from actual experiences with crime.

We contend that the nature and logic of crime in Latin America and the Caribbean have implications for citizens’ perceptions about the state and, in consequence, shape preferences about welfare provision. Criminal activity—particularly that conducted by organized crime, the main perpetrator of violence in the Latin American region—expands the public’s distrust in the role of the state in the provision of public security. High crime rates, thus, serve as an indicator of the state’s willingness to protect its citizens’ well-being.⁶ Crucially, a diminishing belief in the state’s capabilities to handle major tasks and functions can also destroy fiscal contracts.⁷

In this context, we argue, those with a heightened perception of insecurity are less likely to support the state’s having a larger role in the provision of welfare and public goods. We suggest that, in the face of severe crime, citizens feeling more exposed to violence may become more likely to reject interactions with the government, out of distrust or skepticism. Turning to the experience of crime victimization, we must acknowledge, however, its consequences in terms of new objective economic and health needs, from the restoration of damaged property and productivity costs to medical and healthcare services. Grievances derived from personal experiences with crime should, in contrast, positively alter demand for social policy provision, even if victims are skeptical of public institutions. Actual victimization turns diffuse fears into actual costs and needs. We therefore expect that, despite the state’s failure to provide public security, victims are more likely to demand government action as a way of coping with such new needs and costs.

Our study advocates for the disaggregation and correct identification of crime on social policy preferences, separating perceptions of insecurity from victimization experiences. We theorize and provide evidence on the varying dynamics linking crime exposure to policy attitudes. The article therefore contributes to the growing literature that seeks to examine the different behavioral and attitudinal consequences of crime exposure,⁸ albeit with a more disaggregated account. Ultimately, we contend that the spread of negative perceptions of insecurity as a result of growing criminal violence limits the possible expansion of the Latin American and Caribbean welfare state, despite the growing needs that victims of crime face as a result of their direct exposure to violence. Perceptions of insecurity tend to be more widespread than victimization.⁹ A negative effect of heightened perceptions of insecurity on social policy support, therefore, poses a severe threat to the expansion of the welfare state, contributing to the ongoing debate on the challenges of welfare state development in developing countries.¹⁰

To understand how a violent context transforms citizens’ perceptions of welfare policies, we first briefly review previous work on the determinants of social policy preferences and then explore the extant studies of the connection between welfare state policies and crime. Second, focusing on criminal violence, we propose two contrasting hypotheses about how crime might shape preferences for welfare provision in Latin America and the Caribbean (LAC), via perceptions of insecurity and victimization experiences, but by very distinct processes. Third, we present our research design, data, and methods. Fourth, we discuss the statistical tests and the robustness checks. Finally, in the conclusion, we provide an assessment of the political implications of growing organized crime activity for the development of the welfare state in LAC.

Welfare State Policies and Crime

A vast political economy literature has explored the different social and economic determinants of preferences for redistribution and welfare policies. At the micro level, the most influential insight has been that individual positions in the income distribution strongly affect preferences, with individuals below the mean preferring higher levels of redistribution.¹¹

In the Latin American context, scholars have found a diluted effect of relative income on preferences for redistribution.¹² They have provided evidence for the effect of other individual characteristics, such as occupation and place of residence,¹³ revealing potential limitations of the classical income models to account for preferences beyond advanced democracies. Moreover, low expectations among the poor critically diminish their demand for welfare policies.¹⁴

Exposure to certain contextual conditions can also lead individuals to support or oppose public social spending. Racial inequalities seem to have a depressing effect on mass support for redistributive policies.¹⁵ Similarly, a growing informal sector might reduce support for welfare spending.¹⁶ In line with this second strand of studies, our work explores the effect of individual exposure to criminal violence, a highly salient contextual characteristic in Latin America.

Previous research addressing the nexus between crime and welfare has shown that social and penal policies intersect and exhibit crucial interdependencies.¹⁷ For the specific study of the relationship between crime and attitudes toward the welfare state, extant works have mainly focused on the European context, analyzing the role of fear of crime. Evidence suggests that individuals in developed democracies can be supportive of redistribution in unequal contexts as a response to a heightened fear of crime.¹⁸ As David Rueda and Daniel Stegmueller show, fear of crime increases with higher levels of inequality. To address this negative externality of inequality, citizens—especially the rich—express greater demand for redistribution in order to reduce crime through welfare policies, which are considered a suitable response in Western Europe. In developed democracies, welfare measures seem to help individuals protect themselves from social and economic risks and reduce feelings of insecurity.¹⁹

As noted, these works have been exclusively limited to advanced democracies, where there is an effective level of state intervention and social welfare efforts (as well as criminal activity) have a very different nature and profile than in LAC. When we think outside of the European context and place ourselves in the region, we must consider that—despite improvements in the provision of noncontributory benefits and the expansion of universal social policies²⁰—welfare systems tend to be truncated, with slowly growing coverage rates and stagnating poverty-alleviation outcomes.²¹

Citizens in high-income countries turn toward the state in demand for protection, but such an effect is less evident in developing countries, where state capacity is an issue of debate and the state colludes with special interests and criminal groups.²² Although economic inequality in Latin America affects the distribution of crime,²³ crime itself might play a different role than the effect argued by Rueda and Stegmueller in advanced industrial democracies. Increased levels of crime might weaken mass support for state efforts. Previous works have shown that trust, expectations, and willingness to contribute to the tax effort are essential components for sustaining public support for the welfare state.²⁴ If citizens have low expectations of the potential of social policies to reduce inequality (and therefore crime) because of corruption, interest group capture, or lack of capacity of government officials, then state efforts will not be perceived as the best strategy to cope with crime. It can be more efficient to simply invest in private protection if one can afford it, rather than try to enforce a public solution via redistribution. As detailed elsewhere in this special issue, recent findings from Gustavo Flores-Macías and Mariano Sánchez-Talanquer for the case of Mexico corroborate this thought, showing that a decline in public safety reduces individuals’ willingness to sustain the fiscal contract.²⁵

Although the literature linking the welfare state and crime is still fairly limited, studies do suggest the importance of getting a better understanding of this relationship. It is also important to note that although the works that explicitly link crime and social policy preferences refer to a positive relationship between these variables,²⁶ they largely fail to consider the role of victimization experiences, to differentiate them from perceptions of insecurity, and to distinguish their potentially distinct consequences for the welfare-providing role of the state. In this article, we argue that such a distinction is relevant for several reasons. Perceptions of insecurity do not closely reflect the local crime rate or individual victimization rates.²⁷ At the same time, perceptions of insecurity are socially constructed,²⁸ whereas victimization experiences become embedded in an individual’s identity²⁹ and imply a continuous interactive process between the individual and the surrounding actors, constantly shaping the victim’s needs and attitudes.³⁰ Therefore, as will be discussed further, to understand citizens’ perceptions about welfare policies, perceptions of insecurity must be distinguished from victimization experiences. When crime is partly a result of corrupt state practices, as in LAC, and the consequences of victimization are taken into account on their own, attitudes toward public goods provision can be profoundly affected through diverging mechanisms.

Preferences for Welfare Provision amid Crime in LAC

In order to comprehend how crime shapes social policy preferences in LAC, we must first understand the logic and nature of such criminal activity, contextualize it, and consequently derive its implications for individual attitudes toward the role of the state in the provision of public goods, distinguishing the different roles and consequences of crime violence for victims and nonvictims. Among victims, we must account for the mounting problems, costs, and needs that criminal violence entails and that most likely make their attitudes toward welfare provision different from attitudes of those who have not experienced crime directly. Nonvictims may, nonetheless, perceive insecurity in their surroundings, but with fundamentally distinct consequences for their needs and attitudes toward the state. To understand this distinction, we explore the different logics and implications behind perceptions of insecurity and victimization experiences. On the basis of this research, we derive distinct attitudinal consequences in relation to the welfare state.

Criminal Violence in LAC

Growing violence in LAC over the past decade is largely the result of growing criminal activity, led mainly by organized crime,³¹ which depends greatly on the protection rackets that it is able to establish with different state actors.³² The close relationship between organized crime groups (OCGs) and state authorities largely enables organized criminal activity. It is only through their collusion with government authorities—exerting corruption, bribery, intimidation, or violence—that OCGs are able to flourish, protect their territories, and buy impunity. Therefore, organized crime violence not only reflects the vulnerability of state institutions but also the political control that criminal groups can exert on them.³³

Beyond organized crime activity—which may include homicides, kidnappings, disappearances, and extortion, among other criminal acts—ordinary street crime in LAC, such as robbery, has also expanded considerably, with a majority of such events involving violence.³⁴ Still, we must not ignore that assault, robbery, and coercion are tools also used by organized criminal groups, through local gangs, to guarantee profits and local compliance.³⁵ As we note in the following sections, such violent encounters with crime entail new economic challenges for victims.

Corruption and the weakness of Latin American judicial institutions have resulted in a regional failure to stop even ordinary crime. Latin America is ranked by the 2017 Global Impunity Index as the region with the worst global impunity scores.³⁶ In Mexico, reporting a robbery to the prosecutor’s office can take more than three hours, which discourages many from reporting crimes. In fact, only 6 percent of crimes in Mexico are reported and, on average, less than 1 percent of those crimes are actually solved.³⁷ In Guatemala likewise, organized crime has deeply infiltrated the judicial system, resulting in few arrests and investigations, along with quick releases “because of a prosecutor’s ‘weak’ case.”³⁸ These situations are prevalent throughout the LAC region. Ultimately, the prevailing impunity implies further violence and lack of justice, which then magnify perceptions of insecurity and of the state’s (in)ability to provide effective public security.³⁹

It is in the context of organized criminal violence in the LAC region that we seek to understand its consequences for welfare policy. We argue that this requires acknowledging how citizens perceive said insecurity and distinguishing those perceptions from actual experiences with crime, which greatly transform an individual’s needs beyond any subjective notion of the general situation of violence.

Perception of Insecurity versus Victimization Experiences

Research in diverse disciplines points to the importance of differentiating perceptions of insecurity and victimization experiences. A copious literature in sociology, criminology, and political science has explored the role and logic of perceptions of insecurity.⁴⁰ As argued by Wesley Skogan, perceptions of insecurity involve cognitive and evaluative components of an individual’s notions regarding crime.⁴¹ The perception of insecurity is based on a mix of objective information, cognitive processing (which can differ across individuals), and socially constructed beliefs and learning.⁴² Numerous works have shown that perceived risk is unrelated to personal experiences with crime and is in fact largely disproportional to actual rates of victimization. People perceive a much higher risk than is actually the case.⁴³ In her study of violence in Latin America, Rossana Reguillo finds that perceptions of insecurity are socially constructed and then culturally shared.⁴⁴ Works delving into individual characteristics associated with increased perceptions of insecurity habitually point to the socialization process of risk perception, by which, for example, the elderly and women express a heightened sense of insecurity because of a socialized vulnerability rather than higher victimization rates.⁴⁵

Recent works on victimization experiences further show that such experiences imply a fundamentally different process from perceptions of insecurity. Becoming a victim of crime involves a transformation of an individual’s identity. Direct experiences with crime redefine how a person regards himself or herself as well as how a person is regarded by others.⁴⁶ Foundational work on the politics of victimization by Eduardo Moncada has shown that becoming a victim is not only the experience of a one-time criminal act but a continuous interactive process between victims, criminals, the state, and society overall.⁴⁷

Given that perceptions of insecurity are socially constructed, we must understand how such socialization may also imply a reconfiguration of the way individuals perceive the broader issues around crime and insecurity, such as the role of the state. At the same time, given that victimization implies a transformation of the way an individual perceives himself or herself and relates with his or her surroundings, we must focus on how a direct experience with crime also transforms an individual’s reality and position regarding the state—not through socially constructed perceptions but through actual encounters that translate into specific needs and demands.

Perception of Insecurity and the Welfare State

From the perspective of political economy, an increase in perceived risk or uncertainty should be met with an increasing demand for redistribution to buffer such insecurity,⁴⁸ as the welfare state’s function is to hedge risks. However, the perceived uncertainty emanating from crime and violence is fundamentally different from worries about economic risks. Some Latin American regions may in fact be controlled by criminal groups. Across Brazil, Colombia, Jamaica, and Mexico, organized crime has been able to exert economic and political controls, collecting de facto taxes through extortion quotas, influencing electoral processes, and capturing and redirecting resources targeted at social programs.⁴⁹ In such circumstances, not only does distrust in the state expand but state welfare provision is largely ineffective.

When crime and violence are sustained by the collusion of state authorities with organized crime, as well as by the weakness and corruption of many of the police and judicial institutions—as in the case of LAC—citizens likely not only perceive higher physical insecurity but also reject further interactions with the state. According to Miguel Carreras, high perceptions of violence make individuals less respectful and proud of national political institutions.⁵⁰ Lack of support for the state’s institutions as a result of growing violence and feelings of insecurity can have profound consequences for the ways citizens interact with the state. Machado, Scartascini, and Tommasi show consistently that when institutions are weak, citizens do not engage with the state through official institutional mechanisms.⁵¹ Merolla, Mezini, and Zechmeister further note that during crises—possibly related to security—citizens are less likely to support democratic forms of governance.⁵²

In that light, we must consider that demanding a higher level of welfare provision entails not only an individual preference for the state having a larger role, as noted above, but also engaging in further interactions with the state, from paying taxes to following the procedures required to use public services. Social policy provision relies on an effective fiscal contract, but citizens may be less willing to pay taxes if they perceive that the government is not delivering adequate services or that it is corrupt, incompetent, or captured by special interests and criminal groups.⁵³ In other words, it is not simply incapacity due to lack of resources that repels citizens but the state’s unwillingness. Thus, if reciprocity breaks down, individuals will be less supportive of redistribution.⁵⁴ In addition, citizens form social policy preferences on the basis of expectations about the quality and fairness of public services.⁵⁵ Indeed, research suggests that perceptions about the quality of government are crucial for understanding levels of public support for welfare policies.⁵⁶ In the case of Latin America, previous analyses have shown that historic levels of exclusion from access to welfare provision have diminished citizens’ expectations about the redistributive role of the state. Alisha Holland notes that flat or regressive transfers and informal barriers to access reinforce skepticism of the benefits of social policy provision.⁵⁷ This effect is present even among poor voters, who, in developed economies, are generally supportive of welfare efforts.⁵⁸ Low public expectations in Latin America have led to a process of “coalition hollowing” from below, as low-income voters do not expect social policies to change their well-being fundamentally and therefore weaken their demand for redistribution.⁵⁹

As criminal violence in LAC erodes both citizens’ political trust and controls over state institutions, expectations about what the state can achieve regarding welfare provision are likely to decline. Therefore, we expect that individuals who feel more insecure—and who are consequently likely to be more suspicious of state institutions—will be less supportive of an extensive public welfare provision. In that regard, we must recall that the character of crime insecurity implies that the usual risk-insurance rationale is decoupled and individuals are likely to turn instead to private means of protection in response to heightened perceptions of insecurity. The mechanism might run through several channels, such as decreasing trust, disappointment in the state institutions, heightened skepticism of state capacity, or an increased perception of state corruption, induced by the perception that public safety is not adequately provided. We therefore propose the following hypothesis:

Hypothesis 1. Individuals with high perceptions of insecurity are less likely to demand public welfare provision than those perceiving their environment as more secure.

Victims of Crime and the Welfare State

Although crime may be associated with greater skepticism about the state and its capacity to provide public goods, we must also take into account that, despite such doubts, violence inevitably implies a profound change in the victims’ social and economic reality. This is part of the transformational process to which recent works on victimization have referred.⁶⁰ Studies have shown that violent crime has important consequences for employment and health outcomes. These two areas are at the core of welfare policies, and citizens who personally experience trauma and loss may acknowledge the role of the state in helping them address the economic and health consequences of crime victimization.

Robles, Calderón, and Magaloni demonstrate that criminal violence diminishes consumption and labor participation.⁶¹ In particular, criminal violence has had the greatest impact on the self-employed.⁶² It has resulted in fewer years of education for their children, as parents have been unable to keep them in school and have had to incorporate them into the labor market.⁶³ Violent crime has also affected health outcomes. Criminal violence results in low birth weight and prematurity.⁶⁴ In addition, a rise in violent crime is associated with an increase in blood pressure and hypertension rates.⁶⁵ And, of course, crime victimization and exposure to violence have important psychological effects that may require professional attention.⁶⁶

Victims of both violent and nonviolent crime face new and mounting costs, needs, and problems, including (a) out-of-pocket costs, such as for legal services and stolen or damaged property; (b) need for medical care, ranging from rehabilitation and prescriptions to premature funeral expenses; (c) need for mental healthcare by psychiatrists, psychologists, and social workers; and (d) productivity costs, such as lost wages and work days by victims and their families, in addition to long-term productivity losses in case of injury or lost body parts.⁶⁷ Even in the case of nonviolent crime, such as robbery or extortion, victimization means losing part of one’s income as well as replacing damaged or stolen property. Being a victim of a kidnapping can also mean surrendering family wealth to pay a ransom. Victimization of family members has significant costs for relatives. The rising number of disappearances across Latin America implies a major loss of family income—as a result either of the lost income of the household’s main financial provider or the new expenses invested in looking for the missing person—on top of the new mental health needs that the victim’s relatives will most likely face.

Overall, crime experiences entail numerous economic, health, and emotional costs that then render individuals unable to meet other basic needs. Such experiences, moreover, involve a high degree of stress that can affect one’s ability to work. Individuals who have been directly affected by crime may be especially motivated to demand government intervention and provision of social policies to help them cope with the multiple costs of crime. Even though the state may have failed victims by not protecting them from crime, the provision of public goods could be one of the “most tangible manifestations of the efforts by the state to remedy the harms they have suffered.”⁶⁸ Some victims—as in the case of disappearances—may require welfare provision even to continue their quest for truth and justice; without such help they would be unable to provide for their families and at the same time pay for legal services or travel to search for their missing loved ones.⁶⁹

Mexico—where violence has risen considerably over the past decade—is an interesting illustration of the impact of crime on victims’ needs.⁷⁰ Crime victims and their relatives can submit petitions to the Executive Commission of Attention to Victims (CEAV) for the specific assistance and support they require.⁷¹ The types of requests that the commission considers are revealing, as they show the kinds of needs that victims and their relatives face. Out of fifteen possible services the commission can help with or cover, thirteen are related to medical and mental healthcare, including surgeries and medical interventions, facial trauma and reconstructive surgery, prosthetic and orthotic devices, counseling and psychological services, and obstetric and pediatric services. This example points to a simple political economy rationale: an increase in costs must be met with an increasing demand for social protection through the state.

The foregoing discussion suggests that victims may consider that despite the failure of the state regarding security—or precisely because of such failure—it is more than ever the state’s responsibility to provide other basic public goods and help fill other basic needs. Moreover, there simply is no alternative to public social protection when victims are economically strained. Therefore, even if they are deeply skeptical of state institutions, victims may have increased demands for social policy provision as a consequence of their grievances, their new reality, and their transformed individual identities and relationship to their surroundings. We argue that as they face the burden of these unexpected costs, victims might be particularly supportive of an active role of the state in effectively providing the social services they require to cover their most basic needs. We then propose the following hypothesis:

Hypothesis 2. Victims of crime are more likely to demand public welfare provision than nonvictims.

Overall, our two proposed hypotheses present a puzzle we seek to address in this article. Although perceptions of insecurity may diminish trust in state institutions and thus depress the demand for an active role of the state in the provision of welfare, victims’ mounting costs and needs increase their demands for social protection. Untangling the effects of victimization experiences and perceptions of insecurity on citizen support for social policy can help us develop a more objective and comprehensive assessment of the political and social consequences of crime in LAC.

Research Design

We focus on the micro level. Because perceptions of crime differ from objective levels of criminal violence, we emphasize the need to study insecurity perceptions and crime victimization experiences independently from objective statistics. We test our hypotheses with survey data from three waves of the Latin American Public Opinion Project (LAPOP), conducted in 2008, 2010, and 2012, which contain the items needed to operationalize support for public welfare provision.⁷² The data cover countries in LAC so that a total of twenty-four countries are included.⁷³ We pool the survey waves and control for survey year and country fixed effects. Since levels of violence and crime vary at the very local level, we employ clustered standard errors at the municipality level.

Dependent Variable: Social Policy Demand

Building on the considerable scholarship studying preferences for social policies, we employ the standard items used in cross-country surveys to operationalize support for public social policies.⁷⁴ LAPOP asks respondents how much they support the statement that the “government, more than individuals, should be primarily responsible for ensuring the well-being of the people” (ROS2). The item reflects demand for public welfare, independent of how public goods are financed, deliberately leaving aside fairness and responsibility considerations that are more salient when respondents are asked about redistributive preferences. Respondents can answer on a scale from 1 (not at all) to 7 (a lot).⁷⁵

Levels of support for welfare policies are generally high across LAC countries, with the majority of responses concentrating on favorable scores regarding government provision of social policies. We therefore follow previous analyses and focus on studying the determinants of markedly positive attitudes.⁷⁶ We assign a value of 1 for strongly positive attitudes toward welfare provision (6–7) and a value of 0 for less favorable responses (1–5). Given the dichotomous nature of this dependent variable, we use logistic regression analysis.⁷⁷

In order to move beyond general support for welfare provision, we also test how far crime victimization and perceptions of insecurity influence preferences in a particular social policy field: healthcare. As we have already noted, because of the violent nature of crime in most cases in LAC, crime victimization usually leads to a series of immediate health needs. Furthermore, violent crime has long-lasting negative effects on health outcomes for victims and their families.⁷⁸ Healthcare needs concern not only the restoration of physical well-being but also psychological support to cope with trauma.⁷⁹

Our second dependent variable therefore focuses on a specific welfare policy: support for public healthcare (ROS6). The statement reads, “The (Country) government, more than the private sector should be primarily responsible for providing healthcare services,” with the same range of response options that we recode into a dichotomous variable focusing on strong support at the extreme.

Both survey items used here for the operationalization of our dependent variable refer to a trade-off between market and state as the responsible entity for the provision of welfare in general or healthcare in particular.⁸⁰ Support for public healthcare was asked only in 2010 and 2012, whereas the general public welfare preference was asked in 2008 as well.

Independent Variables: Perceptions of Insecurity and Victimization

To measure subjective perceptions of insecurity we employ an item asking respondents whether they feel safe in their neighborhoods. Insecurity perception ranges from 1 (safe) to 4 (very unsafe). The measure of insecurity perception available through LAPOP allows us to capture the individual’s risk perception. This personal assessment of insecurity is particularly useful for three reasons: (1) it addresses the neighborhood level,⁸¹ the most immediate political and social milieu shaping the way citizens view politics;⁸² (2) it is fundamentally different from a direct victimization experience; and (3) it is also different from fear of crime, which, as noted by Ferraro,⁸³ is more emotive in nature, whereas risk entails a cognitive judgment and involves people’s assessments of crime rates and the probability of victimization.

We capture victimization using a survey item asking respondents whether they have been a victim of a crime in the past twelve months. Although social desirability bias cannot be fully ruled out when using surveys to study sensitive issues, most criminologists largely use this technique because of its overall reliability.⁸⁴ The total nonresponse rate on this question is only 0.45 percent (LAPOP 2008–12). Moreover, the question is sufficiently broad that the difficulty of answering sincerely is less significant than for questions asking about victimization of specific types of crime. In addition, this question allows us to address and acknowledge the emotional and material costs that even regular crime such as a robbery entails: from stress, anxiety, and trauma to direct economic losses.⁸⁵ LAPOP also asks whether a family member has fallen victim to a crime, which can be categorized as a further form of victimization. We study both the direct question about personal victimization and a variable for victimization that conflates individual and family victimization as another operationalization of the victimization experience (see Supplementary Material, Table S1, available online on the publisher’s website).

Individual victimization experience through crime and the perception of living in an insecure neighborhood are only weakly positively correlated: ρ₁ = 0.19 (at the 0.1 percent level of significance). This is consistent with previously cited works that have persistently shown how perceptions of insecurity are not related to victimization experiences.⁸⁶ As noted, several works find that perceptions of insecurity are in fact more widespread than actual victimization rates.⁸⁷

Of course, victimization is not a randomized event, so we need to consider possible selection effects, which might make victimization endogenous to social policy preferences. The same holds true for insecurity perception. In order to address these concerns, we report cross-tables of insecurity perception and crime victimization with a set of standard socioeconomic covariates in the Supplementary Material (Tables B and C). Victims of crime tend to be young males with some level of education and living in urban areas, but they do not differ significantly from nonvictims in other sociodemographic characteristics that might be consequential for preferences regarding redistribution. In particular, correlates on education and urbanization, which can serve as a good proxy for demand for public welfare support, do not indicate that victims are more likely to require state-provided welfare and therefore to support tax-funded services. In addition, as noted before, we have no reason to think that the poor are particularly supportive of welfare expansion in the region.⁸⁸

Also, perceptions of insecurity do not seem to cluster in specific neighborhoods; they are widespread and often independent of the level of respondent’s reported experiences with crime (see Supplementary Material, Table C). As we would expect, respondents also perceive more risk when they live in urban areas. Women and young respondents tend to perceive a somewhat higher risk, but there are barely any differences for other covariates. Although some degree of sorting might take place (the rich select into safer neighborhoods), the spread of insecurity perception across the board demonstrates the deeply individual character of such perceptions. Again, this broad distribution of perceptions of insecurity indicates that there is no particular group opting into a given characteristic that could also be affecting preferences for redistribution and social policy.

When we aggregate LAPOP information for crime victimization and insecurity perception to the country level, the correlation between these variables is only moderately positive (see Fig. 1). Moreover, insecurity perception is weakly correlated with objective levels of violence. Using the homicide rate (homicides per 100,000 individuals) from the World Bank reveals that reported levels of violence even negatively correlate with insecurity perception at the individual level, with a correlation coefficient of ρ₂ = −0.04. This shows again that risk perception is a local measure that varies and therefore does not overlap with national violence figures. The discrepancy between objective and subjective levels of crime and violence further strengthens the need to understand perceptions of insecurity independently from actual experience as well as independently from official statistics that rely on reported crime information.⁸⁹

Figure 1.

Aggregated Insecurity Perception (Range 1–4) and Aggregated Victimization (Percentage).

Controls

Given that we expect trust in public institutions to be one of the key channels through which the effect of insecurity perception influences demand for public welfare, our models include a control for individual trust in political institutions. Using principal component analysis, we generate an index that summarizes individual trust toward the national legislature, the justice system, the supreme electoral tribunal, the armed forces, the national police, political parties, the president, and the supreme court. We add the institutional trust index as a control in a stepwise procedure.

Drawing on the literature previously discussed, we employ a set of standard sociodemographic characteristics. We factor in gender (female), age, age squared (to consider nonlinearities such as older respondents’ being more supportive of welfare provision because they rely on pensions), living in either urban or rural areas (rural is the reference category), and level of education (measured as years of education).⁹⁰ In our analysis we also take into consideration the individual’s employment status by introducing dummy variables for being a public employee, unemployed, retired, or responsible for housekeeping (employed in the private sector serves as reference category). Further, we add a variable for household status (married), since recent research has revealed important effects on welfare support following from intrahousehold redistribution or inequalities.⁹¹

Prioritizing a parsimonious model, we focus on sociodemographic characteristics as controls.⁹² The country and year fixed effects control for between-country variation so that cross-country variation in income inequality, welfare provision, and crime rates is accounted for.⁹³

Results

Results for the influence of crime experience and insecurity perception on the demand for public welfare provision, the dependent variable we discuss first, are reported in Table 1. Perceiving one’s neighborhood as insecure reduces the likelihood of expressing welfare demand at the 1 percent level of significance. The results from Model 1 corroborate our first hypothesis. In contrast, the personal experience of crime increases support for public welfare policies, as shown in Model 2, and thereby lends support to our second hypothesis. We report logistic coefficients in Table 1 and describe substantive effects next.⁹⁴

Table 1.

Logistic Regression: Support for Public Welfare.

	(M1)	(M2)	(M3)	(M4)	(M5)	(M6)
Insecurity perception	−0.028** (0.011)		−0.037*** (0.011)	−0.052*** (0.012)	−0.023* (0.011)
Victimization		0.103*** (0.021)	0.117*** (0.021)	−0.062(0.060)		0.111*** (0.021)
Victimization × insecurity perception				0.071** (0.022)
Institutional trust					0.019*** (0.005)	0.022*** (0.005)
Controls
Female	−0.071*** (0.017)	−0.072*** (0.017)	−0.068*** (0.017)	−0.069*** (0.017)	−0.070*** (0.017)	−0.071*** (0.017)
Age	0.009** (0.003)	0.009** (0.003)	0.010** (0.003)	0.010** (0.003)	0.010** (0.003)	0.010** (0.003)
Age²	−0.000(0.000)	−0.000(0.000)	−0.000(0.000)	−0.000(0.000)	−0.000(0.000)	−0.000(0.000)
Wealth index	−0.050*** (0.012)	−0.051*** (0.012)	−0.051*** (0.012)	−0.051*** (0.012)	−0.049*** (0.012)	−0.050*** (0.012)
Years of education	0.007** (0.003)	0.007* (0.003)	0.007* (0.003)	0.007* (0.003)	0.008** (0.003)	0.007** (0.003)
Employment status
Public employee (ref. cat.: private employee)	0.078* (0.033)	0.083* (0.033)	0.082* (0.033)	0.082* (0.033)	0.071* (0.033)	0.076* (0.033)
Unemployed	0.053(0.035)	0.052(0.035)	0.055(0.035)	0.055(0.035)	0.055(0.035)	0.054(0.035)
Nonemployed	−0.010(0.021)	−0.005(0.021)	−0.006(0.021)	−0.005(0.021)	−0.013(0.021)	−0.008(0.021)
Retired	−0.056(0.041)	−0.052(0.041)	−0.052(0.041)	−0.052(0.041)	−0.061(0.041)	−0.058(0.041)
Urban	0.008(0.029)	−0.006(0.029)	0.003(0.029)	0.004(0.029)	0.012(0.029)	−0.001(0.029)
Married	0.014(0.017)	0.015(0.018)	0.016(0.018)	0.016(0.018)	0.013(0.017)	0.014(0.018)
2010 (ref. cat.: 2008)	0.157*** (0.038)	0.154*** (0.038)	0.155*** (0.038)	0.154*** (0.038)	0.152*** (0.038)	0.148*** (0.038)
2012	0.131*** (0.037)	0.132*** (0.037)	0.129*** (0.037)	0.129*** (0.037)	0.125*** (0.037)	0.125*** (0.037)
Fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Constant	0.180+ (0.105)	0.109(0.102)	0.185+ (0.105)	0.216* (0.106)	0.147(0.106)	0.084(0.103)
Log-likelihood	−49,404.3	−49,395.8	−49,386.8	−49,381.0	−49,392.1	−49,379.4
Chi²	1,070.43	1,075.9	1,104.31	1,119.63	1,070.98	1.083.8
BIC	9,9202.5	9,9185.5	9,9178.9	9,9178.5	9,9189.4	9,9163.9

Note: Dependent variable is public welfare. We use a logistic regression with country and year fixed effects and clustered standard errors at the municipality level. Country fixed effects are not displayed. Observations = 77,321. BIC = Bayesian information criterion.

Source: Authors’ elaboration based on LAPOP 2008–12.

p < 0.10, *p < 0.05, ** p < 0.01, *** p < 0.001.

Becoming a victim of crime is positively associated with an increase in the incidence rate of individual consent to public welfare; the coefficient is highly significant at the 0.1 percent level of significance in Model 2 and Model 3. We report predicted probabilities for our two main independent variables in Figure 2, holding all other variables in the model at modes. Crime victimization increases the likelihood of supporting public welfare by roughly 2.3 percent. In contrast, the likelihood of demanding public welfare decreases by 1.9 percent when an individual goes from feeling safe to unsafe, which is a small but still substantive change, given that it refers to the average change. Effect sizes may differ (with smaller or larger magnitudes) for subgroups in our sample. Figure 2 shows a discernible effect for insecurity perception at the extremes of the scale—safe and unsafe—while 95 percent confidence intervals overlap largely for the middle categories. The graph reveals the uncertainty around the model estimates for the middle categories of this variable, and therefore the result for perceived insecurity should be interpreted with caution. Nevertheless, a change from feeling safe to feeling unsafe can occur rapidly as perceived risk increases, and the point estimates in Table 1 suggest that such change has a significant impact. Compared to the average marginal effects for education, where one additional year increases the probability of support by 0.2 percent, or wealth, which reduces support by −2.4 percent when the change from one standard deviation below to one standard deviation above the mean is estimated,⁹⁵ the magnitude of the effects for crime exposure is substantial.⁹⁶

Figure 2.

Predicted Probabilities for Crime Victimization and Insecurity Perception on Welfare Support.

Both variables seem to affect welfare state support independently; the effects remain significant in Model 3 when both variables are added jointly. In contrast to the positive effect of fear of crime on welfare support portrayed for the context of advanced industrial democracy,⁹⁷ perceptions of insecurity seem to reduce social policy demand in LAC, in line with findings for Mexico.⁹⁸ Latin Americans instead are more likely to opt for private welfare solutions when facing security failure of the state.

Insight into the economic and social consequences of victimization can explain the increased support for public welfare. Since crime experience leads to adverse economic effects—particularly in the LAC region, where crime entails some form of violence—such as unemployment, income shortage, private debt, or increased healthcare expenses, greater demand for public welfare is a logical step. Direct crime victimization then stands for a simultaneous increase in economic risk.⁹⁹

The effects are similar for preferences regarding the more specific social policy, that is, public healthcare. Average support for increased public healthcare provision is higher than for the more general welfare question, with a mean of 0.75 (standard deviation is 0.43). Table 2 reports our estimation results for the impact of crime victimization and the perception of insecurity on healthcare support. Again, personal experience of crime goes together with higher demand for public healthcare, whereas the perception of insecurity in one’s neighborhood reduces the likelihood of support for healthcare provision. The predicted probabilities in Figure 3 corroborate that result. The probability of supporting healthcare provision increases by 2.3 percent for victims, consistent with our argument. However, the coefficient for insecurity perception is only significant at the 10 percent level of significance in Model 7. Figure 3 shows overlapping confidence intervals across the categories of this variable. The effect of insecurity perception with a change in predicted probabilities of 1.4 percent is, hence, less robust for healthcare support than for general welfare.

Table 2.

Logistic Regression: Support for Public Healthcare.

	(M 7)	(M 8)	(M 9)	(M 10)	(M 11)	(M 12)
Insecurity perception	−0.027+ (0.015)		−0.038** (0.015)	−0.051** (0.016)	−0.024+ (0.014)
Victimization		0.135*** (0.030)	0.149*** (0.030)	−0.001(0.083)		0.140*** (0.030)
Victimization × insecurity perception				0.060* (0.030)
Institutional trust					0.010(0.007)	0.013* (0.007)
Controls
Female	−0.045* (0.021)	−0.046* (0.021)	−0.042+ (0.021)	−0.042* (0.021)	−0.044* (0.021)	−0.045* (0.021)
Age	0.008* (0.004)	0.008* (0.004)	0.009* (0.004)	0.009* (0.004)	0.008* (0.004)	0.009* (0.004)
Age²	−0.000(0.000)	−0.000(0.000)	−0.000(0.000)	−0.000(0.000)	−0.000(0.000)	−0.000(0.000)
Wealth index	−0.025(0.017)	−0.027(0.017)	−0.027(0.017)	−0.027(0.017)	−0.025(0.017)	−0.026(0.017)
Years of education	0.013*** (0.003)	0.012*** (0.003)	0.012*** (0.003)	0.012*** (0.003)	0.013*** (0.003)	0.012*** (0.003)
Employment status
Public employee (ref. cat.: private employee)	0.042(0.046)	0.048(0.046)	0.047(0.046)	0.047(0.046)	0.039(0.046)	0.044(0.046)
Unemployed	0.098* (0.045)	0.098* (0.045)	0.099* (0.045)	0.100* (0.045)	0.098* (0.045)	0.098* (0.045)
Nonemployed	0.001(0.028)	0.007(0.028)	0.006(0.028)	0.006(0.028)	−0.001(0.028)	0.005(0.028)
Retired	−0.062(0.055)	−0.056(0.055)	−0.058(0.055)	−0.058(0.055)	−0.065(0.055)	−0.060(0.055)
Urban	−0.039(0.043)	−0.056(0.043)	−0.047(0.043)	−0.046(0.043)	−0.038(0.043)	−0.053(0.043)
Married	0.036(0.024)	0.037(0.024)	0.038(0.024)	0.038(0.024)	0.035(0.024)	0.036(0.024)
2012 (ref. cat.: 2010)	−0.118** (0.042)	−0.114** (0.042)	−0.117** (0.042)	−0.117** (0.042)	−0.118** (0.042)	−0.114** (0.042)
Fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Constant	0.987*** (0.140)	0.907*** (0.136)	0.987*** (0.140)	1.013*** (0.140)	0.968*** (0.142)	0.889*** (0.138)
Log-likelihood	−30,052.59	−30,041.58	−30,036.26	−30,033.98	−30,050.78	−30,038.23
Chi²	807.38	811.88	820.8	825.013	804.036	808.39
BIC	60,477.5	60,455.5	60,455.8	60,462.2	60,484.9	60,459.8

Note: Dependent variable is healthcare. We use logistic regression with country and year fixed effects and clustered standard errors at the municipality level. Country fixed effects are not displayed. Observations = 57,051. BIC = Bayesian information criterion.

Source: Authors’ elaboration based on LAPOP 2010, 2012.

p < 0.10, * p < 0.05, ** p < 0.01, ***p < 0.001.

Figure 3.

Predicted Probabilities for Crime Victimization and Insecurity Perception on Healthcare Support (Table 2, Model 7 and Model 8).

The effect for insecurity perception is reduced in significance and magnitude for both dependent variables when we add a control for institutional trust in Table 1, Model 5, and Table 2, Model 11, while the effect for victimization (Model 6 and Model 12) only becomes stronger in magnitude. The Bayesian information criterion (BIC) value in the analysis of public welfare is slightly smaller in the model specifications that include institutional trust, indicating an improvement of the model fit. The reduced effect for insecurity perception on welfare preferences when institutional trust is held constant supports the assumption that the effect of insecurity perception works to some extent through the mechanism of trust in public institutions. The effect for victimization seems to be independent from institutional trust, aligning with our argument that suggests a role of grievances derived from experiences with crime.

As perception and actual crime experience indicated opposing dynamics regarding welfare preferences, we need to analyze which impact dominates in cases where both are present. We therefore build an interaction term of crime experience and insecurity perception. Estimation results are reported in Model 4 for public welfare demand and in Model 10 for healthcare support. We see that when both victimization and heightened perceptions of insecurity overlap, the positive impact of victimization is reinforced even further.¹⁰⁰

To interpret the interaction term more accurately, we plot the average marginal effect of victimization at different levels of insecurity perception in Figure 4. We find no discernable difference between the welfare preferences of victims who perceive their neighborhood as safe and individuals who did not experience any type of crime in the past twelve months. Victims and nonvictims who feel safe overlap in their probability to support public welfare with an incidence rate of 64 percent. However, as their perceptions of insecurity increase, victims continue to demand public welfare and healthcare, while rising perceptions of insecurity among nonvictims result in significantly diminished support for public welfare, falling to roughly 60 percent. The effect is similar for healthcare preferences as shown in the panel on the right in Figure 4. Overall, for differences in risk perception among nonvictims we see a clear renunciation of the state. But the victimization experience undermines the negative effect of insecurity perceptions on welfare preferences: victims who also feel unsafe remain as likely to demand public goods as victims who feel safe. Experiencing crime turns diffuse insecurity into real insecurity.

Figure 4.

Average Marginal Effect of Crime Victimization at Different Levels of Insecurity Perception on Public Welfare (Table 1, Model 4) and Healthcare (Table 2, Model 10).

Before we explore further the mechanism behind our findings for crime victimization in the subsequent section, some findings for the control variables deserve further elaboration. As indicated above, the more affluent respondents are, the less supportive they are of generous public goods provision (assuming a redistributive character), in line with findings of previous research that uses asset information to determine the respondents’ wealth status. Age significantly affects both general welfare support and preferences for healthcare, with an average change in probabilities of 0.1 percent. Welfare demand increases with age. Finally, the unemployed are significantly more likely to support public healthcare than private employees and public employees, which corroborates previous research.

Underlying Mechanisms

To shed light on the mechanism that forms the basis of support for welfare policies among victims of crime, we add further tests. Although the cross-national character of the data limits the means to test for the mechanism, we can explore correlations between victimization and further factors that reflect economic need. Family victimization equally exerts a positive effect on both support for public welfare and healthcare, which suggests that experiences with crime have a general negative impact on a household’s well-being, generating specific needs and subsequent demands (see Table S1 in the Supplementary Material). Moreover, the severity of the experience of crime should make a difference when economic needs drive support. Although nonviolent crime can be equally traumatizing, we can infer an increase in concrete needs among victims of violent crime with greater certainty. We therefore divide types of crime experience into violent (unarmed robbery with assault or physical threats, armed robbery, assault but not robbery, rape or sexual assault, kidnapping) and nonviolent (unarmed robbery, no assault or physical threat, vandalism, home burglary, and extortion) crime victimization.¹⁰¹ Victims of violent crime should have a stronger demand for public welfare, especially regarding healthcare compared to nonvictims. We report the predicted probabilities for our findings in Figure 5. Indeed, for public welfare we do not find discernible differences regarding the intensity of crime experience; both victims of violent and nonviolent crime are more supportive of public welfare (see Panel a in Fig. 5). But when it comes to healthcare, victims who suffered particularly violent encounters are significantly more likely to demand support than nonvictims (see Panel b). In contrast, victims of nonviolent crime are not significantly different from nonvictims. The findings on severity of crime thus correspond with our theoretical expectations regarding specific needs that (certain) victimization experiences create, particularly when related to the violent organized crime activity that permeates in LAC.¹⁰²

Figure 5.

Predicted Probabilities of Supporting Public Welfare and Healthcare by Experience with Crime (Violent, Nonviolent, Nonvictims) (LAPOP 2010, 2012).

In addition, interviews with Mexican public officials at CEAV help deepen our understanding of the types of benefits and services that victims demand when they turn to the state.¹⁰³ As noted, consistent with the results above, health is the most prominent and frequent social service provided by the commission to all victims. Beyond health, these interviews revealed further needs that victims and their families face. First, since many services require traveling to Mexico City to follow up and review their cases at the national prosecutor’s office, many victims and their relatives demand economic support for transportation, lodging, and food during such trips. Also, depending on the type of victimization and subsequent needs, direct and indirect victims may receive financial support for housing—for example in the case of victims of forced displacement. In the specific case of disappearances, indirect victims inevitably have a permanent need until their missing relative is found and therefore receive permanent economic support for room and board, in addition to scholarships for their children. Although these services do little effectively to repair the damage that the crime has caused, as noted by one of CEAV’s officials, in the face of prevalent impunity, “given the lack of legal justice, there should at least be social justice.”¹⁰⁴ Hence, although the state may have failed victims by not guaranteeing their safety and victims may consequently be skeptical of the state’s capacity, they are vocal in demanding welfare services that will at least help them continue their search for justice, which becomes their main priority. Although the evidence is limited and is not intended to be conclusive, it does help substantiate our argument regarding the needs that victims face and that translate into demands for the welfare state.

Discussion

The analysis of individual preferences on social policy reveals that criminal violence leaves its mark. Our evidence, however, points to two different directions. Individuals who become victims of crime are supportive of the welfare state and public healthcare in particular. In the face of rising risks, victimization increases the state’s mandate among individuals to improve welfare policies. In contrast, perceptions of insecurity reduce such demand for public policies, particularly among those with significantly contrasting views regarding their neighborhood’s (un)safety. Several mechanisms may be at work: increasing disapproval of the state’s performance in the provision of security, rising skepticism of the government’s capacity to control crime, or decreasing trust in the capabilities of the state to govern effectively. Looking at the joint effect of victimization and insecurity perception, experiences with crime seem to cancel out the negative effect of “mere” perception.

The empirical support for the victimization hypothesis potentially speaks of a mechanism in which policy support is generated through personal needs based on a change of circumstances. Our findings thereby contribute to the literature that has identified the personal experience of economic risk as a strong driver for redistributive preferences.¹⁰⁵ When they experience job loss and a period of unemployment, individuals tend to support public welfare policies at a higher rate.¹⁰⁶ The experience of crime might be similar to other egotropic perceptions. In contrast, the perception of insecurity speaks to a less personalized evaluation of circumstances. Feeling insecure might spur dissatisfaction with the performance of the state; perceiving a potential risk of crime and violence exposes the state’s failure to provide public security. However, the cross-sectional nature of our data and the obvious limitations regarding randomization of crime mean that we cannot test the mechanism that drives the two different responses to crime.

It is also important to note that our argument and findings speak to the specific characteristics of LAC, where violence has been characterized by the rising violent activity of organized crime and its implications for perceptions regarding the state. This situation is likely to contrast with the settings in which other works, both in Europe and other developing countries, have studied the relationship between demand for redistribution and crime. Experiences with crime and violence diverge from one region to another in their nature, the actors involved, and the state’s capacity to address them. Such diverging characteristics are also likely to have different implications for the welfare state in other regions.

Conclusion

In addressing the phenomenon of growing rates of violence and crime in LAC, scholarly work has revealed that a violent context can have deep attitudinal effects. Building on those insights, our study examines how crime and violence influence individual preferences about public policies. The question is relevant because investing in social policies, such as improved school enrollment rates that reduce the risk of unemployment or public healthcare that grants access to these services for all, has the potential to reduce violence and crime in the long run. Tackling the problem of crime through the means of the welfare state is therefore an alternative to iron-fist policies, which are favorably promoted by political candidates who campaign on the crime issue.

Our theoretical contribution is twofold. First, we move beyond the prevalent analysis of crime’s effect on citizens’ demand for iron-fist security policies, incorporating a dimension of policy that can be as important for preventing and addressing crime—welfare provision. Second, we show that a limited focus on perceptions of insecurity or victimization does not take us far enough to fully understand the impact of violence on citizens’ attitudes. Although crime victimization increases support for public welfare and healthcare, insecurity perception reduces the likelihood that the welfare state will be viewed favorably. When the two factors overlap and victims also perceive surrounding insecurity, support for welfare policies remains stable, pointing to a qualitative difference between actual experience and diffuse risk perception.

Our findings on victims also contribute to the incipient political science literature on the behavioral and attitudinal consequences of crime victimization. Victims can become more socially engaged while also withdrawing from electoral politics and political institutions.¹⁰⁷ Such departure from the political sphere in response to diminishing institutional trust is reasonable, but when it comes to social policy preferences, withdrawal is less likely. As we have shown here, crime victimization creates real costs and needs for which private options may be limited or insufficient.

Because the perception of insecurity is more widespread than actual victimization, however, our findings suggest that increasing crime rates might put welfare state development at stake in the long run. When evaluations of public safety are flawed, respondents have a lower likelihood to demand public welfare provision. If such evaluations influence vote choice, insecurity perceptions can boost economically conservative candidates into office so that a retrenchment of the already limited welfare programs is ushered in. Perceptions are, therefore, likely to be an important factor suppressing much-needed public goods provision. Of course, one could question how much preferences matter for welfare state development, but scholars have shown that preferences can be very decisive in shaping vote choices and policies.¹⁰⁸

Finally, this article points to potentially fruitful strands of future research. We elaborate on different mechanisms that link crime and social policy preferences. We propose that what drives victims of crime to mandate social policies is the increase in economic and health hardships due to the experience with crime. However, further research is needed to substantiate this mechanism and explore other intervening variables that may mediate the relationship between the victimization experience and views on the state, such as income or type of crime. The negative effect of insecurity perception on social policy preferences seems most likely to result from a simultaneous decrease of trust in public institutions and in government’s capabilities to carry out public policies, given a perceived security policy failure. Testing the mechanism behind such a negative impact should therefore also be at the core of future research on the crime–welfare state nexus. In addition, these findings should also push us to explore the role and distribution of welfare alternatives on which citizens living amid criminal violence—victims in particular—rely, within their private spheres of interaction, when facing the negative economic and social consequences of crime.

Supplemental Material

Altamirano_et_al_Supplementary_Material_WelfareStateCrime_220420 – Supplemental material for The Welfare State amid Crime: How Victimization and Perceptions of Insecurity Affect Social Policy Preferences in Latin America and the Caribbean*

Supplemental material, Altamirano_et_al_Supplementary_Material_WelfareStateCrime_220420 for The Welfare State amid Crime: How Victimization and Perceptions of Insecurity Affect Social Policy Preferences in Latin America and the Caribbean* by Melina Altamirano, Sarah Berens and Sandra Ley in Politics & Society

Footnotes

Acknowledgements

Preliminary versions of this article were presented at the Latin American Public Opinion Project at Vanderbilt University in 2019; the workshop “Societies under Stress” at The Ohio State University, November 2018; the PSAC seminar at Cornell University, 2018; the workshop “Criminal Violence” at CIDE–Mexico City, 2018; the “Criminal and Political Violence in Mexico and Central America” workshop at UC San Diego, 2018; EPSA 2018; MPSA 2018; and the workshop “The Left Hand and the Right Hand: Linking the Politics of Social Welfare and Crime Control” in Copenhagen, December 2017. We would like to thank Sarah Brooks, Miguel Carreras, Angélica Durán-Martínez, Peter Enns, Gustavo Flores-Macías, Omar García-Ponce, Sergio I. García-Ríos, Kristian Gleditsch, Yanilda González, Noam Lupu, Brian Phillips, David de Micheli, Javier Osorio, David Shirk, Mitchell Seligson, Peter Starke, Marianne Ulriksen, Christopher Way, Georg Wenzelburger, and Elizabeth Zechmeister for their helpful comments and suggestions. We gratefully acknowledge the use of data made available by the Latin American Public Opinion Project, which bears no responsibility for the analyses and interpretations presented here. The authors are exclusively responsible.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Melina Altamirano (maltamirano@colmex.mx) is an assistant professor at the Center for International Studies at El Colegio de México (Colmex) in Mexico City. Her research focuses on the political implications of economic vulnerability and the sources of public support for social policy in Latin America. Her most recent work was published in Latin American Politics & Society.

Sarah Berens (sarah.berens@uibk.ac.at) is an assistant professor of political economy at the University of Innsbruck. Her research focuses on the microfoundation of labor informality, social policy, taxation, and the behavioral impacts of crime experience in low- and middle-income countries. Her latest work has been published in Socio-Economic Review, Political Behavior, and Political Studies.

Sandra Ley (sandra.ley@cide.edu) is an assistant professor in the Political Studies Division at the Center for Research and Teaching in Economics (CIDE) in Mexico City. Her research addresses the political and behavioral consequences of criminal violence and has appeared in the Journal of Conflict Resolution and Latin American Politics & Society.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.19 MB