El Salvador is home to powerful street gangs like Mara Salvatrucha (MS-13) and the 18th Street (Barrio 18) (Bruneau et al., 2011; Wolf, 2012b). Governments at the opposite ends of the ideological spectrum have attempted a variety of iron fist (mano dura) security policies to combat these organizations and reduce crime and violence (Cruz & Durán-Martínez, 2016; Franco, 2008; Wolf, 2011). Since 2015, when El Salvador registered the world’s highest murder rate with 103 homicides per 100,000 people, the levels of violence started to decline (Brigida, 2021). Yet in 2019, Salvadorans elected Nayib Bukele to the presidency. Like other candidates in the post-civil war era, Bukele has vowed to fight criminal organizations and has responded to public opinion and the population’s desire for drastic measures to tackle insecurity. Anti-gang initiatives have included extrajudicial killings, the torturing of suspected gang members, mass incarceration, and the militarization of El Salvador’s domestic security policy. Critics fear that Bukele’s authoritarian practices will not only damage democracy but could result in major human rights abuses and higher levels of violence (HRW, 2020; WOLA, 2020).

How might fear and criminal violence as well as the social and political context affect the public’s appetite for mano dura policies in El Salvador? This article examines recent trends in El Salvador and utilizes regression analysis to evaluate several relevant factors associated with punitivism. Based on an evaluation of theoretical and empirical scholarly works, together with the political, social, and historical context of present-day El Salvador, we show that fear of crime plays a vital role in support for tough on crime policies. The public does not trust the state and its ability to prosecute the guilty and is willing to take drastic measures to reduce crime, violence, and insecurity—actions that ultimately erode institutions and democracy. Since security concerns can pervade the population’s policy preferences, we also hypothesized that people from all ideological beliefs support mano dura policies in El Salvador. In a context characterized by dissatisfaction with political parties, ideological differences become less relevant, favoring only those candidates that run on a tough on crime policy regardless of their political affiliation. We argue that the convergence of political parties from all sides of the ideological spectrum around mano dura is likely to perpetuate punitivism as the electoral competition becomes symbolized by “punitive Darwinism” or the survival of the toughest against crime. This concept is akin to the race to the bottom for politicians trying to convince the public that they are the toughest and are willing to take the necessary steps to combat crime and violence.

This discussion is divided into five sections. It begins with a short review of the theoretical approaches to understand punitivism. With a special emphasis on crime-related scholarly works, the next section addresses prior empirical research evaluating different predictors of support for mano dura. The third section describes the broader context of violence in El Salvador as well as the policy responses adopted between the end of the civil war and the Bukele administration. The methodology section outlines the data analysis procedures, describes the survey data used in the article, and presents the regression model. The final section reviews the empirical results and discusses the implications drawing on the proposed concept of punitive Darwinism. The article concludes by summarizing the major findings of this work.

Understanding punitivism

Rising concerns about insecurity among many Western industrialized countries accelerated in the 1970s (Beckett, 1997; Garland, 1990), reaching developing regions such as Latin America and the Caribbean over the past decades (Dammert & Malone, 2006; Swanson, 2013). A growing body of work has attempted to provide theoretically grounded explanations to untangle how emotions and other cognitive factors affect the population’s punishment preferences. Moral and utilitarian accounts have been at the forefront of the theoretical debate across different social science disciplines (Johnson, 2009; Tyler & Boeckmann, 1997).

Utilitarian explanations emphasize that punitivism emerges when individuals or communities feel threatened (Maruna et al., 2004). The source of concern lies in tangible risks such as being the victim of a crime and the ineffectiveness of public institutions to control violence (Tyler & Boeckmann, 1997). From this perspective, contexts of insecurity, whether characterized by high victimization rates and or by fear of crime, pave the way for the emergence of punitive attitudes among the population. Thomas and Foster (1975), for example, contend that support for capital punishment can be interpreted as an understandable consequence of the widely held beliefs that crime rates were increasing rapidly during the second half of the twentieth century, that the average citizen was in danger of becoming a victim of crime, and the cultural notion that the death penalty was an effective means by which deviant and criminal behavior can be controlled (Thomas & Foster, 1975).

The study of the moral or symbolic determinants of punitive responses have an extensive history in social sciences, including the classical works of scholars (Durkheim, 1933; Mead, 1918; Ranulf, 1938). Durkheim (1933) sustains that crime provokes a psychological reaction of passionate feelings against the perpetrator. According to this view, the emotional response of punishment over those who have violated or infringed certain rules of conduct help to reestablish the values and order of societies. In other words, punitive reactions to rule breaking are intimately linked to public concerns about the cohesiveness of the family, the community, and society (Tyler & Boeckmann, 1997).

A more recent strand of relevant theorizing highlighting the symbolic function of punitivism has concentrated on the social and cultural contexts of late modern societies. In a comprehensive analysis of the social responses to crime, Garland (2001) addresses the cultural and political forces that gave rise to punitivism in the United States and Great Britain. He argues that punitivism was not a mere consequence of rising crime rates or the loss of faith in penal-welfarism, but rather the product of a series of responses to the cultural and political conditions of late modernity. More specifically, modern institutions of crime control and criminal justice are deeply marked by cultural formations, reactionary politics, and the new social relations that emerged around the changing structures of work, welfare, and market exchange (Garland, 2001). Factors such as the labor market fragmentation, the marginalization of the population, and the transformation of the family life have contributed to create a sensibility and fear about crime and the ultimate emergence of a “culture of control” (Garland, 2001).

Chevigny (2003) also highlights the symbolic relevance of the social, cultural, and economic contexts of late modern societies. He contends that free-market relations and the weakening of the service state provoke anxiety and fear, leaving citizens with the sense that there is no safety to protect them (Chevigny, 2003). Wacquant (2009) emphasizes the dynamics and negative consequences of neoliberalism. From this theoretical perspective, fears of crime and the associated rise of punitivism during the last quarter-century was a direct response to the diffuse social insecurities produced by the fragmentation of wage labor and the transformation of the ethnic hierarchy (Wacquant, 2009). Furthermore, Karstedt contends that the penal law and the criminal justice system have become increasingly “emotionalized” in late modern societies (Karstedt, 2002; Karstedt et al., 2011).

Scholars have shown how fear and emotions have contributed to the problem rather than helping the population to feel more secure. Simon (2007) sustains that institutions use citizens’ emotions and anxieties to promote governance by legitimizing or providing content for the exercise of power. It is argued that the United States has become more racially polarized and less democratic because of crime control policies and governance (Simon, 2007). Along these lines, a growing body of research has responded to the great diversity of institutional consequences of punitivism by increasing their analytic differentiation. Alternative conceptual forms such as “populism of the fear of crime” (Chevigny, 2003), “authoritarian populism” (Hume, 2007), “punitive populism” (Bonner, 2009, 2019), “penal populism” (Roberts et al., 2002), and “democradura” (Pérez, 2003), have been developed to describe the changing nature of criminal law, the criminal justice system, and the practices of government officials in Latin America and the Caribbean.

Prior empirical research

Using local and national samples, empirical research evaluating the link between emotions and support for mano dura has produced mixed results. While some scholars focusing on the United States disregarded the role of emotions as a predictor of punitiveness (Kleck & Jackson, 2017; Tyler & Boeckmann, 1997), a second strand of empirical work found a positive relationship (Costelloe et al., 2009; Dowler, 2003; Johnson, 2009; Unnever et al., 2005). The effects of emotions such as anxieties and fear of crime on punishment preferences have been also tested in other Western industrialized countries, including Great Britain (King & Maruna, 2009), Germany (Armborst, 2017), and Canada (Hartnagel & Templeton, 2012; Wanner & Caputo, 1987). Furthermore, various scholars test empirically how punitive attitudes influence criminal justice policy (Enns, 2014; Nicholson-Crotty et al., 2009).

In the Latin American context, victimization has been a central feature in many empirical accounts examining support for tough on crime policies. Bateson, for example, shows that the region’s victims of crime are far more likely to support hardline strategies such as mano dura (Bateson, 2010, 2012). Other scholars also demonstrate the positive association between victimization and attitudes towards punitive policies (Price et al., 2019). Rosen and Cutrona (2020), on the other hand, did not find a relationship between crime victimization and demands for hardline policies. In their study of Colombia and Brazil, they argue that other economic, political, and social insecurities operating at the aggregate-level may explain mano dura’s growing popularity (Rosen & Cutrona, 2020). These results are consistent with the analysis developed by other scholars (Míguez, 2013). By considering the causes behind crime in Argentina, Míguez (2013) shows that victimization does not lead to demands for harsh policies. According to this scholar, this relationship is reversed if citizens feel that drug addiction and minimum sentences are the driving-forces of insecurity (Míguez, 2013).

Other quantitative studies on support for tough on crime policies in the Americas have emphasized the relevance of alternative socioeconomic and demographic predictors. Scholars conducting survey research have revealed that people who are female and younger are more likely to favor the implementation of mano dura strategies (Price et al., 2019). This relationship is likely connected to the subjective or perceived proximity to crime rather than victimization itself (Miller et al., 1986). Empirical evidence shows that both women and the elderly have higher probabilities of being victimized in their home or neighborhood (Singer et al., 2012), but only the former group experiences comparatively high levels of fear of crime (Singer, 2017). In other words, these findings suggest that the feeling of insecurity rather than the objective crime proximity is likely associated with individuals’ punitive attitudes.

Yet the relationship between fear of crime and support for mano dura has not been widely examined in Latin America. Using data from the 2012 Americas Barometer survey, Singer et al. (2020) show that fear of crime is associated with increased punitiveness. In their regression analysis, the social sources of punitiveness are not limited to crime-related factors (Singer et al., 2020). Similar to other studies developed in the United States (Costelloe et al., 2009; Hogan et al., 2005; Singer et al., 2020) reveal that those who believe that the national economy is poor are more likely to support increased levels of punishment, although this correlation is reversed if the personal economic situation is measured.

Various empirical studies have also shown that crime fears can also translate into non-democratic policy preferences among Latin American citizens. The evidence demonstrates that growing crime rates and associated fear decrease the likelihood of support for political institutions and democracy (Carreras, 2013; Cruz, 2003a), erode the social capital necessary for democracy to work (Moser & McIlwaine, 2003; Caroline Moser & Holland, 1997), and favor authoritarian responses among the citizenry (Smithey & Malone, 2014; Bateson, 2012; Briceño-León & Zubillaga, 2002; Briceño-León et al., 1999; Pérez, 2003; Cruz, 2008).

The non-democratic responses to criminal violence and associated fear are particularly evident in countries such as El Salvador. In examining the post-civil war scenarios in this Central American sub-region, scholars find that both victimization and insecurity in one’s neighborhood have a negative impact on citizens’ level of satisfaction with democracy in El Salvador (Cruz, 2003b). Survey researchers have shown that approximately 55 percent of Salvadorans in 2000 from 18 to 95 years old would justify a coup in the presence of high levels of crime (Seligson et al., 2000). This situation is particularly evident among women, citizens with low-income and low education levels, and people living in small municipalities (Seligson et al., 2000). Smithey and Malone (2014) extend this claim to the aggregate level. According to their regression analysis, not only personal experiences with crime significantly reduce support for the rule of law in El Salvador but also fear of crime at the national level (Smithey & Malone, 2014).

Moreover, other empirical studies have shown that the impact of crime on democracy is often mediated by other relevant variables. Scholars have demonstrated that the attrition of the political culture that supports democracy in El Salvador comes from the government’s loss of legitimacy due to its inability to tackle crime and not only from the levels of violence and insecurity, respectively (Cruz, 2008). Malone, for example, suggests that crime does not unilaterally influence citizens’ public support for the rule of law in El Salvador, as the linkage is mediated by the national context (Malone, 2010, 2012). Diverse factors such as the role of the media (Bonner, 2019; Krause, 2014; Marroquín, 2007; Vasilachis de Gialdino, 2007), partisan competition and ideology (Holland, 2013; Yashar, 2011), the actions of different civil society actors (Bonner, 2019), and even the influence of foreign countries such as the United States (Cutrona, 2017, 2019; Zilberg, 2011) have also played a significant role in understanding policy preferences and mano dura strategies across Latin America and the Caribbean region.

Criminal actors, mano dura, and violence

El Salvador has a lengthy history of violence, as it had a civil war that lasted for more than a decade between the leftist guerilla group known as The Farabundo Martí National Liberation Front (Frente Farabundo Martí para la Liberación Nacional—FMLN) and the right-wing government. The civil war led to more than 70,000 deaths (Byrne, 1996; Wade, 2016; Wood, 2003) and has had long-lasting impacts on the socioeconomic and psychological well-being of El Salvadoran citizens (Allison, 2010).

During the armed conflict, many families fled to the United States to escape the violence afflicting El Salvador (Boerman, 2007; Logan, 2009). Salvadoran youth living in Southern California often had a difficult time fitting into schools as many did not speak English (Brenneman, 2011; Grascia, 2004; Ward, 2013). Youth that felt stigmatized and discriminated against by their peers formed MS-13 in the 1980s in Los Angeles. The gang provided marginalized Salvadorans with a sense of belonging as well as protection (Cruz, 2010; Ward, 2013).

MS-13 transformed over time. Due to their association with violence, the United States began deporting Salvadoran youth with criminal records back to their country of origin. The two governments, however, did not share pertinent information, and the deported youth entered El Salvador without criminal records. The MS-13 members returned to a country that many of them had not been to since their childhoods. The deported gang members faced high levels of poverty coupled with state fragility (Cruz et al., 2000; Cruz, 2007; Rosen & Kassab, 2016), and they continued what they knew best: the gang life (Banks, 2000). As a result of these deportation practices and the country conditions, MS-13 began to spread throughout the Northern Triangle and Southern Mexico (Seelke, 2016; Bruneau et al., 2011).

Various El Salvadoran governments implemented tough on crime strategies to combat gangs and violence (Bruneau, 2014; Wolf, 2012b). Right-wing presidents from the Nationalist Republican Alliance party (Alianza Republicana Nacionalista—ARENA), including Francisco Flores (1994–2004) and later Antonio “Tony” Saca (2004–2009), adopted what became referred to as mano dura and super mano dura strategies to combat gangs, crime, and violence (Hume, 2007; Rodgers, 2009; Wolf, 2011). Authorities rounded up youth involved in these organizations in large raids and incarcerated them. Critics contend that the jails became revolving doors as police arrested many of the same youth multiple times (Cruz, 2010). These tough on crime policies emerged from the public demanding a response to crime and violence as they felt unsafe and sought drastic measures. According to the 2006 LAPOP survey, 41.78 percent of the population maintained that the country needs a government of mano dura. Under these circumstances, politicians scored points with the electorate for their law-and-order platforms and promises to “clean up” the streets (Jutersonke et al., 2009; Wade, 2019; Wolf, 2017).

As the prison population increased, gangs began to better organize within the penitentiary system—the same system designed to punish them. Prisons served as “universities of crime” for gang members. The government separated people by gangs as having cells housing rival organizations could result in violence and even riots. This enabled gang members from diverse cliques to improve their structures (Cruz, 2010; Dudley, 2010). Currently, much of the top leadership of gangs are incarcerated in Salvadoran prisons.

However, tough on crime policies against gangs are not only a practice of the conservative Arena party. The FMLN’s Mauricio Funes, an award-winning journalist who served as president from June 1, 2009 until June 1, 2014, continued with this approach regardless of his political affiliation (Wade, 2019). One may have hypothesized that the left-wing party consisting of many former guerilla members would be less prone to implementing mano dura policies given their experiences and backgrounds as guerilla fighters against a right-wing authoritarian government. Yet President Funes continued tough on crime practices, and the El Salvadoran prison population increased from 19,814 in 2008, a rate of 324 per 100,000 inhabitants, to 28,334 in 2014, a rate of 450 per 100,000 people (ICPR, 2020). This trend suggests that tough on crime strategies have remained a bi-partisan issue not only restricted to the right-wing party.

In 2015, the Salvadoran Supreme Court classified gangs as terrorists, which helped the conservative Salvador Sánchez Cerén administration (2014–2019) justify hardline strategies to tackle high levels of crime and violence. The police forces continued to arrest gang members and could even charge them with illicit association. This policy has impacted many citizens as more than 450,000 people in El Salvador have connections to gangs (Bargent, 2013). The police could arrest youth for living in gang neighborhoods—or having friends who are members of these organizations—and for being “terrorists.” This has led to stigmatization and a sensation among some youth that it is a crime in El Salvador to be young and from a marginalized community (Rosen & Cruz, 2018; Wolf, 2012a).

The tough on crime policies had collateral damages and contributed to the country’s increasing levels of violence. In 2015, El Salvador became the most violent country in the world and recorded many grave human rights abuses (The Guardian, 2015). The levels of violence, in part, are a result of failed tough on crime strategies. The deployment of the military to combat gangs led these criminal organizations to respond by increasing the number of attacks against the police and military. Gangs have also battled among each other for control of territory as their principal source of revenue is extortion, which has only deteriorated the country’s general security landscape (Cruz et al., 2017). Despite attempted reforms, the police and military have been involved in extrajudicial killings and the torturing of suspected gang members. Investigative journalists have revealed that police officers formed death squads responsible for homicides and human rights violations (Asmann, 2019).

Notwithstanding efforts to combat crime and violence and declining homicide rates, security continued to be one of the top issues during the 2019 Salvadoran presidential elections. Nayib Bukele, the former mayor of San Salvador, formed his own political party, New Ideas (Nuevas Ideas), after being kicked out of the FMLN during the presidential campaign. He ran as a populist leader who vowed to improve the country’s economic and security conditions while also addressing the high levels of corruption. Upon assuming office, President Bukele has responded to the desire for tough on crime strategies and has implemented many controversial policies. He vowed to crack down on the gangs after they carried out various murders on the streets. President Bukele has also defied Supreme Court rulings and detained individuals who violated the Covid-19 quarantine. He even encouraged the police to use lethal force during the pandemic (HRW, 2020) and sealed off the doors and windows of prison cells.

In April 2020, the Bukele administration released photos of shirtless gang members in the penitentiary system lined up in rows, one on top of another (Pineda, 2020). The images reveal the cruel and unusual punishment of gang members and the violation of accepted social distancing norms during the Covid-19 pandemic (WOLA, 2020). Bukele contended that prison authorities will no longer separate gangs according to their respective organizations, but rather rival gang members will be placed in the same prison cells. Given the levels of animosity between rival organizations, putting people from different gangs in the same cell could result in violence, torture, and death. Not surprisingly, the photos and statements of Bukele led to an outpouring of criticism by experts and the international community for human rights violations (WOLA, 2020).

Moreover, Bukele has given the green light to the police, as he indicated that the government will no longer record extrajudicial killings (Asmann, 2020) and even vowed to pay the legal fees for police officers if they are investigated by authorities. These recent events, coupled with the deployment of the military to Congress in February 2020 to intimidate politicians who did not support his security law (Agren, 2020), have led to an outpouring of condemnation for Bukele’s undemocratic practices and human rights violations. In response to his critics, the president tweeted a photo of a soldier helping an individual in April 2020 and put the hashtag que bonita dictadura (what a beautiful dictatorship), further fueling the debate about his authoritarian practices.

Bukele is justifying his tough on crime policies and contending that they are necessary, as they have contributed to the recent decreases in violence. While violence has declined in recent years, human rights abuses remain high. In 2019, El Salvador had 2,390 murders, which is equivalent to a homicide rate of 36 per 100,000 inhabitants. This is a significant drop from a rate of more than 100 per 100,000 individuals in 2015 (Asmann & O’Reilly, 2020). While the decline in homicides enables the government to tout the “successes” of tough on crime strategies, the data is more troubling when considering forced disappearances. Taking into consideration the number of disappearances, abductions, and unexplained missing persons, El Salvador has more than 3,600 registered cases (HRW, 2019, 2020). Combining this number—which includes people who have been tortured and killed—with the total number of homicides would produce a much higher homicide rate than 36 per 100,000 inhabitants. The police have also registered more than 10,000 disappearances since 2010, demonstrating the grave human rights abuses (HRW, 2019). Because of the high levels of impunity, many of the forced disappearances have remained unsolved, creating tensions among the public and further eroding confidence in institutions.

In September 2020, investigative reports by El Faro revealed that the Bukele administration negotiated secretly with MS-13 for a year to reduce the levels of violence, indicating the government’s pragmatism and willingness to combat crime at any cost (Martínez et al., 2020), even if it includes making a deal with the most powerful gang in the Western Hemisphere. The Bukele administration has denied such negotiations and maintains that it is continuing its hardline strategies to combat gangs, crime, and violence. President Bukele understands that negotiating with gangs could costs his party votes at the polls, as the population is desperate for the government to take drastic measures to reduce gangs and criminal activity plaguing the country.

Methodology

We now turn to providing an initial empirical assessment of the correlates of support for mano dura in El Salvador. First, theoretical accounts and previous research emphasize the role of fear of crime in driving punitivism. We would therefore expect that measures of fear would be correlated with punitive attitudes among Salvadorans. Another key insight emerging from the above discussion is that in El Salvador, given the high levels of violence and the extent to which it appears to have already moved towards punitive Darwinistic politics, we would not expect political ideology and/or party affiliation to be correlated with support for mano dura policies. Similarly, the mixed results of previous analyses with respect to the impact of actual victimization on attitudes towards tough on crime policies suggest that this factor should be examined, but that it may not be a correlate of support for mano dura policies in El Salvador. Finally, trust in institutions plays a role in these accounts of punitivism, with citizens having low levels of trust in most government institutions. Considering the state’s inability to provide security in El Salvador, we looked at two possible institutions: the military and the judiciary system. Given the central role that the military has played in mano dura policies in El Salvador, we might expect that trust in the military is associated with support for tough on crime policies. Conversely, it is possible that a lack of trust in “ordinary justice” through the court system could be a factor driving public punitivism.

To assess some of the correlations between these theoretically relevant factors and actual support for mano dura policies in the El Salvadoran context, we utilize the 2018/2019 AmericasBarometer Survey conducted by The Latin American Public Opinion Project (LAPOP) at Vanderbilt University. LAPOP uses a sophisticated methodological approach, polling the adult voting age population using a multi-stage cluster sampling design based on regions, urban and rural areas, and the size of municipalities. The El Salvador survey had 1,511 participants and a sampling error of 2.5 percent. We use the data from this survey to operationalize these hypotheses. The operationalization of our dependent variable is the respondent’s answer to the LAPOP survey question about the need to increase penalties for crime. Respondents were presented with the statement, “To reduce crime in a country like ours, punishment of criminals must be increased”Footnote 1 and given a seven-point Agree/Disagree Likert Scale to categorize their response, with one representing “strongly disagree” and seven representing “strongly agree.”

The variables of interest are operationalized as follows. “Fear of crime” is measured with a binary indicator that the respondent has limited places for recreation for fear of crime. Other measures are possible—for example, Singer et al. (2020) uses “how safe do you feel in your neighborhood?”—but we feel that limiting recreation out of fear has a particularly universal potential resonance with respondents, is emotional, and has a social component which all resonate clearly with the theoretical work relevant to understanding punitivism. Ideology is measured on a ten-point scale, one for left and ten for right. Though this variable is sometimes collapsed into a simple “left and right” indicator, we follow other scholars (Polga-Hecimovich, 2019) and leave the scale uncollapsed in our analysis, so as not to miss people in the ideological center. Crime victimization, as opposed to the fear of crime, is the answer to a yes/no question asking whether the respondent has been the victim of crime in the last 12 months. In our data, this is a dummy variable coded “no” and “yes.”Footnote 2 Institutional trust in the military, and confidence that the judiciary will punish the guilty are used based on their relevance to the case in El Salvador. Trust in the military is measured on a seven-point Likert Scale, with one being “none” and seven being “a lot.” Confidence that the judiciary system will punish the guilty is the measurement for judicial institutions and is coded on a four-point scale of “a lot,” “some,” “little,” and “none.”

In our model, we also control for age, sex, education, urban, and monthly household income to account for standard socio-economic factors. The age variable is numerical, ranging from 18 to 99 years old, while sex is a dummy indicator for men and women.Footnote 3 Education is from zero years of school to 18 or more. Monthly household income has 16 categories from zero income to more than $900. Descriptive statistics for all variables are presented in Table 1.

Table 1 Descriptive statistics
Full size table

The purpose of this analysis is to examine some initial empirical evidence regarding support for punitivism in El Salvador. Multiple regression allows us to see what cluster of social, political, and attitudinal variables move together in this context. We initially ran an ordinary least squares (OLS) regression model using Stata 17 to assess the hypotheses about the factors that do and do not influence support for tough on crime policies.Footnote 4 After running the regression, we ran the linktest, which produced at hatsq term that is not statistically significant. Thus, we determined that our model did not have any specification errors. The model had a Mean Variance Inflation Factor (VIF) of 1.20 or less, indicating that the model did not have issues with multicollinearity. Finally, the Breusch–Pagan test produced a Prob > Chi2 that was statistically significant, indicating that the basic OLS model had issues with heteroscedasticity. We adjusted the model utilizing robust standard errors. The results from this standard-error corrected model ultimately form the basis of our analysis and discussion below (Table 2).

Table 2 Factors influencing perceptions of whether penalties for crime need to increase
Full size table

Descriptive statistics

The survey data indicates that the population continues to feel unsafe despite the country’s declining homicide rates and the tough on crime strategies implemented over the past decade. According to the 2018/ 2019 LAPOP data, 20.64 percent of the population answered that they have been a victim of a crime during the last 12 months. Moreover, 54.37 percent of Salvadorans contended that they have limited places for recreation for fear of crime.

While the Salvadoran public continues to be distrustful of institutions and the ability of the state to implement the rule of law, they support extreme measures to combat insecurity. This finding is consistent with trends over the past two decades (Córdova Macías & Cruz, 2008; Cruz, 2011; Pérez, 2003). When asked if there is a need to increase the penalties for people who commit crimes, 47.57 percent of the population stated that they “strongly agree,” while only 3.24 percent responded that they “strongly disagree.” In addition, 31.01 percent of the population maintained that a military coup is justified when crime is high. Finally, 35.80 percent of the population responded that they have no trust that the judicial system with punish the guilty, while 16.87 percent responded “a lot.”

The military remains one of the most trusted institutions and is perceived as more professional than the police (Pérez, 2015). On a seven-point Likert Scale, with one being “none,” and seven being “a lot,” 9.03 percent of the population answered “none” when asked their level of trust in the armed forces, while 19.13 percent responded “a lot.” On the other hand, 10.73 percent of the population stated that they had “a lot” of trust in the police, while 13.46 answered “none.”

Findings

In the model, age, fear, and trust in the military are statistically significant at the 95 percent confidence interval. For every one-unit increase in age, we expect a -0.008 shift in the dependent variable, holding all other variables constant. This suggests that people who are older are less likely to support the need to increase penalties for crime. This finding is consistent with previous quantitative studies, indicating that risk of crime victimization is higher earlier in life and decreases over time (Price et al., 2019). The negative relationship between aging and support for harsher sentencing could be also associated with past experiences with violence. El Salvador’s civil war persisted for more than 12 years, registering more than 70,000 deaths, people disappeared, and grave human rights violations (Moodie, 2011). The negative consequences of the confrontation between the government and domestic actors, albeit their different nature, could still be fresh on the minds of the elderly.

Moreover, for every one-unit increase in limited recreational activity because of fear of crime (i.e., moving for no to yes), we expect to see a 0.360 increase in the dependent variable, holding all other variables constant. It appears that people who are afraid to be in public spaces because of fear of crime are more likely to support tough on crime policies. While emotions seem to affect El Salvadoran’s policy preferences, victimization is not statistically significant in the regression model. The results suggest that punitivism is highly contextual and likely associated with economic, political, and social insecurities occurring at the aggregate level. This is consistent with studies focused on the United States and other Western industrialized countries, which emphasize that fear of crime affects the public’s punitive attitudes (Costelloe et al., 2009; Dowler, 2003). Singer et al. (2020) also find this correlation in Latin America.

The model also shows that for every one-unit increase in trust in the military we can anticipate a 0.094 increase in the dependent variable, holding all other variables constant. It appears that people who have higher levels of trust in the military are more likely to support the need to increase the penalties for crime. Our findings are consistent with the work of other scholars who have examined the high levels of trust in the armed forces (Pion-Berlin & Carreras, 2017). The military is perceived as better trained, more professional, and less corrupt than the police forces (Pérez, 2015; Pion-Berlin, 2003). Public support for the military also makes it easier for presidents to justify their participation in internal security operations despite the many criticisms by academics who argue that there should be a separation between the military and the police, as involving the armed forces in internal security affairs has led to human rights abuses (Amaya Cobar, 2012; Bagley, 1991; Isacson, 2001; Kruijt & Koonings, 2012; Main, 2014).

Interestingly, ideology is not statistically significant in the regression model. There are several reasons that could explain this finding. Survey data indicates that people feel insecure in El Salvador and are willing to take drastic measures to reduce crime and violence. Tough on crime policies are supported by people on all-sides of the political spectrum and appears not to be an ideological issue in this country. Prior empirical work has demonstrated that conservative constituencies are not necessarily more likely to support mano dura (Rosen & Cutrona, 2020). Under these circumstances, anxieties and security concerns may help to bridge the classical ideological divide between progressive and conservative individuals under a political competition that drives parties together rather than apart.

The other institutional variable in our model, confidence that the judiciary will punish the guilty, is not statistically significant. As noted in the descriptive statistics, the El Salvadoran public has low levels of trust in government institutions. Besides the trust in the military, the other institutions do not influence support for tough on crime strategies. The population-level evidence shows that people are desperate for change and are tired of the high levels of corruption and crime. Since institutions seem to be less relevant in a context of high security concerns, strongmen like Bukele, who was kicked out of the FMLN and formed his own political party, succeeded in tapping into the public’s high levels of fear and vowed to implement mano dura strategies to combat gangs and violence.

Discussion

Our study contributes to understanding punitivism in three different ways. First, we demonstrate that punitivism among Salvadorans is associated with fear of crime. Neither the end of the military rule that governed much of the twentieth century nor the peace accords following the civil war of the 1980s brought peace to El Salvador. The rise of Mara Salvatrucha and the 18th Street changed the nature of violence, and gang-related crimes proliferated during the 2010s. Interestingly, punitivism among Salvadorans is not associated with personal experiences with crime, as indicated by the lack of a statistically significant correlation between victimization and punitive attitudes, but with emotions and other cognitive factors affecting the public’s policy preferences. The election of Nayib Bukele in 2019—albeit homicide rates started to decrease after 2015—highlights the need to explore the different ways in which economic, political, and social insecurities operating at the aggregate-level exacerbate punitivism in El Salvador. More attention should be given to understanding the role that media outlets and the politicization of crime play in public support for mano dura policies.

Second, this study also shows that trust in ordinary law-and-order institutions is not statistically significant for predicting punitive attitudes. The levels of violence in El Salvador have exposed the serious problems of the state to provide security to its citizens. Whether violence was associated with the civil war or to the role of new criminal actors such as MS-13 and the 18th Street, the public does not trust institutions and their ability to prosecute the guilty. In a context of growing dissatisfaction due to high levels of impunity, the findings of this work suggest that Salvadorans may be willing to tolerate the concentration of power in the hands of the executive and lower levels of horizontal accountability as long as the problem of insecurity is addressed. The scenario is likely to favor the emergence of strongmen leaders promising to recreate the political system like Nayib Bukele, who tap into the populace’s fears of crime to promote radical security policies such as involving the country’s military forces in domestic policing operations.

Fear of crime and the inability of the state to maintain the rule of law are clearly associated with punitivism. Yet security concerns and distrust in institutions are necessary but not sufficient conditions for mano dura policies to persist over time. A third contribution of this work lies in the electoral implications of punitivism. The findings of this study suggest that the population is willing to support candidates who are tough on crime regardless of their political affiliation. Since the electoral competition becomes increasingly de-ideologized in contexts of high security concerns, political parties are likely to converge around mano dura platforms, thereby reinforcing continuity and discouraging change. El Salvador’s past presidential elections reflected how being “soft on crime” was not an alternative for candidates coming from both sides of the political spectrum, including Francisco Flores and Antonio Saca from ARENA and Mauricio Funes and Nayib Bukele from the FMLN and Nuevas Ideas, respectively. Even once in office, Bukele understood that negotiating with gangs could cost his party votes and rapidly denied the accusations following the investigative reports released by El Faro (Martínez et al., 2020). The empirical results of this article suggest, in other words, that only the toughest politicians on crime, or what we refer as punitive Darwinism, are likely to survive in El Salvador if ideological cleavages give way to security concerns and distrust in political institutions.

Although our study is limited to El Salvador, results are relevant to countries with high security concerns, dissatisfaction with political institutions, and, perhaps most importantly, where policy preferences become increasingly de-ideologized. When ideology loses ground in the face of heavy-handed demands, political parties, especially those less predisposed to implement radical solutions to crime, find greater incentives to adapt their electoral platforms according to the prevailing social climate. Punitivism is not only likely to affect the quality of democracy, but it could also become a defining platform of all political parties.

Conclusion

The Bukele administration is responding to El Salvador’s high levels of gang activity, crime, and violence. The president is tapping into the fear of the populace and perceptions of crime to implement militarized internal policing strategies. President Bukele’s populist agenda has also given the police permission to use lethal force against gang members, maintaining that extreme measures are necessary to reduce the country’s high levels of crime and violence. This gives the police and military more opportunities to partake in human rights abuses with impunity given the lack of accountability and the growing concentration of power in the hands of the executive.

Salvadorans have low levels of trust in institutions, except for the military. In a context of limited confidence in the country’s criminal justice system, the Supreme Court, and law enforcement institutions, cognitive and emotional factors such as fear of crime appear to be pervasive in the population’s policy preferences. The relevance of different predictors of support for mano dura such as limited recreational activity because of fear of crime illustrates citizens’ concerns and anxieties about gangs, crime, and violence.

If security concerns and the perceived ineffectiveness of public institutions to control violence can influence individuals’ punitive attitudes what does this mean for El Salvador’s democracy? We maintain that the implications of the population’s beliefs about tangible risks such as being a victim of crime and the lack of legitimacy of the country’s political institutions are likely to affect the quality of democracy as citizens may be willing to accept the concentration of power in the hands of the executive, lower levels of accountability, and the use of the military forces in policing operations if the government tackles insecurity.

Since ideological differences are blurred in a context of high security concerns and dissatisfaction with political institutions, the landscape in El Salvador suggests that the prospects for policy change are limited. The population is willing to support candidates who are the “toughest” on crime and vow to change the system and reduce crime and violence regardless of their political affiliation and institutional implications for democracy. Not only Bukele’s New Ideas party, but also the ARENA and the left-wing FMLN parties have harnessed fear and perceptions of insecurity to gain votes. In other words, the evidence suggests that El Salvador’s punitive Darwinism, in which only the political survival of the toughest is guaranteed, is likely to perpetuate the presence of mano dura policies.