Semester Long Journal (50%)

Criminal Violence and
Democratization in Central America:

The Survival of the Violent State

José Miguel Cruz

ABSTRACT

Why does Nicaragua have less violent crime than Guatemala, El Sal-
vador, and Honduras? All these countries underwent political tran-
sitions in the 1990s. Many explanations point to the legacies of war,
socioeconomic underdevelopment, and neoliberal structural
reforms. However, these arguments do not fully explain why,
despite economic reforms conducted throughout the region, war-
less Honduras and wealthier Guatemala and El Salvador have much
more crime than Nicaragua. This article argues that public security
reforms carried out during the political transitions shaped the abil-
ity of the new regimes to control the violence produced by their
own institutions and collaborators. In the analysis of the crisis of
public security, it is important to bring the state back. The survival
of violent entrepreneurs in the new security apparatus and their
relationship with new governing elites foster the conditions for the
escalation of violence in northern Central America.

Violence still reigns over Central America. According to the UnitedNations Office on Drugs and Crime, Central American nations, par-
ticularly El Salvador, Guatemala, and Honduras, “may have recently sur-
passed the traditional world leaders in the number of murders commit-
ted per 100,000 members of the population” (UNODC 2007, 53). Yet this
violence differs from that which reigned not long ago, in the midst of
civil wars and political instability.

Starting with the triumph of the Sandinista revolution in 1979, most
countries in the region underwent a far-reaching process that would
overthrow the authoritarian governments that had ruled Guatemala, El
Salvador, Honduras, and Nicaragua for decades (Torres Rivas 2001). By
1996, all these countries had developed a new set of electoral institu-
tions, stopped the military conflicts that ravaged the region in the 1980s,
and created institutions to enforce law and respect for human rights.

However, the transitions from authoritarian rule also yielded
another type of wave, one of criminal violence, which has been flood-
ing Central America since the mid-1990s. More than ten years after the
transitions, this crime wave has turned these countries into the most vio-
lent in the world, producing the paradox of regimes that are electoral
democracies but that live under a de facto state of siege produced by

© 2011 University of Miami

violent crime. Guatemala, Honduras, and El Salvador have homicide
rates that double or triple the Latin American average, which itself is
considered very high (PNUD 2009).

Nevertheless, not all the Central American nations that underwent
transitions from authoritarian regimes have become equally violent.
Nicaragua stands as a special case, given its lower levels of crime in
comparison with the northern subregion of Central America (PNUD
2009). Although Nicaragua has critical issues regarding public security,
any comparison with Guatemala, El Salvador, and Honduras exhibits a
clear difference in terms of the prevalence of crime. Indeed, the levels
of violence in Nicaragua are closer to those of Costa Rica, a longstand-
ing democracy, than to those of other transitional countries (PNUD
2009; Programa Estado de la Región 2008).

Why is Nicaragua different? Why has the Nicaraguan society not
produced severe levels of violence, as did those of Guatemala, El Sal-
vador, and Honduras? Why, despite the long war Nicaragua suffered, the
social instability that accompanied its first posttransition years, and its
widespread poverty and inequality, did this nation remain less violent?
To answer this question, we need to pursue three themes. First, the
political conditions that made the northern part of Central America an
extremely violent region; second, the importance of transforming secu-
rity institutions into rule-of-law institutions; and third, the role of state
institutions and agents linked to them in the reproduction of criminal
violence.

In pursuing these themes, this article presents a conceptual frame-
work that emphasizes the role of the state and the persistence of infor-
mal brokers linked to the state in the management of public security and
the reproduction of violence. Clearly, the latter affect the performance
of security institutions, since they use them to obtain impunity and pro-
tection. Variations in the presence of these informal brokers can be
viewed largely as a function of the way political transitions were carried
out. In some cases, political pacts were reoriented or ignored, weaken-
ing public security institutions in order to privilege the surviving actors
and institutions of the old regimes. To demonstrate this, the article
focuses on four aspects of the transition processes: the continuity of vio-
lent entrepreneurs who survived the transition; the relationship between
the survivors and the new democratic governing elites; the variable role
of civil society in the transition processes; and the role of powerful for-
eign actors, particularly the United States.

In Nicaragua, despite the uncertain pace of political transformations,
the long process that started with the Sandinista Revolution eroded the
defining features of the authoritarian regime and set the conditions for
the removal of violent entrepreneurs acting on behalf of the state. In
contrast, in northern Central America, despite the appearance of water-

2 LATIN AMERICAN POLITICS AND SOCIETY 53: 4

shed institutional change and military withdrawal, the key features of
the old security apparatus—that is, the active collaboration of private
groups and civilians to enforce order (Holden 1996)—survived long
after the transition was over. In other words, state institutions have con-
tinued to be a significant source of violence.

CRIMINAL VIOLENCE IN CENTRAL AMERICA

The violence that is currently afflicting Central America is the most com-
plex this region has faced in its periods of peace. UNODC (2007) iden-
tifies eight areas where the problem of violence is especially serious:
drug trafficking, homicide, youth gangs, domestic violence, firearms
trafficking, kidnapping, money laundering, and corruption. Without
underestimating all those problems, this article focuses on the homicide
issue, as murder statistics are usually considered the most reliable indi-
cators of violence.

Table 1 shows some indicators of violence as of 2010: homicide
rates, percentage of victims of street crime, and gang membership per
one hundred thousand inhabitants. As is evident, the northern triangle
of Central America, Guatemala, Honduras, and El Salvador shows indi-
cators that far exceed those of Nicaragua and Costa Rica. The differ-
ences are especially striking with respect to homicide rates as they have
been documented by many organizations (e.g., UNODC 2007; USAID
2006). The UNDP Human Development Report for Central America
2009–2010 separates this region into two, according to crime levels:
one of high criminality, which includes Belize, El Salvador, Guatemala,
and Honduras; and other of low criminality, with Costa Rica, Nicaragua,
and Panama (PNUD 2009, 85–86).

This does not mean that Nicaragua, Costa Rica, and Panama are
crimeless. Research has shown that public insecurity is also a matter of
concern in these countries (Cuadra 2002; PNUD 2005). However,

CRUZ: VIOLENCE IN CENTRAL AMERICA 3

Table 1. Some Indicators of Violence in Central America, 2008–2010

Homicide rate Victims of Street Gang
2010 Street Crime Membership

(per 100,000 pop.) 2010 (%) (per 100,000 pop.)

Guatemala 41.5 23.3 111
El Salvador 64.8 24.2 153
Honduras 77.5 14.0 500
Nicaragua 13.2 19.2 81
Costa Rica 10.6 19.0 62

Sources: LPG Datos 2011; Americas Barometer database 2010.

Nicaragua, with a history of dictatorship, civil war, and profound eco-
nomic problems, seems to rank significantly below the northern trian-
gle in terms of homicide violence.

This gap between Nicaragua and the northern triangle has increased
in the decade since the end of the conflicts. The trends for homicide
rates from 1990 to 2010 show that murders increased in all countries but
the northern triangle levels soared (figure 1). Nicaragua tells a much dif-
ferent story, with a much smaller homicide rate in the immediate post-
war period, a brief increase in the early 1990s and then a slow decline
until 2001, followed by relative stability.1 Nicaragua’s relatively lower
violence is traceable from the 1990s and possibly dates back to the
1980s. However, the posttransition period has only widened the gap,
suggesting distinctive sociopolitical mechanisms behind the crime trends
(Moser and Winton 2002; PNUD 2009). Moreover, Guatemala, Hon-
duras, and El Salvador are set apart from Nicaragua not only in terms of
overall violence; they also suffer more serious problems of organized
crime and street gangs (USAID 2006).

In Guatemala, direct state-sponsored violence, although reduced
after the peace treaty, has been an important source of public insecu-
rity, along with organized crime, youth gangs, routine crime, and public
lynchings (Godoy 2006; Peacock and Beltrán 2003; Schirmer 1998). In

4 LATIN AMERICAN POLITICS AND SOCIETY 53: 4

Figure 1. Homicide Rates 1990–2006 in Postwar Central America
(per 100,000 population)

Sources: Chinchilla 2003; LPG Datos 2011; PNUD 2009; Raudales 2006; UNODC 2007.

postwar El Salvador, state-sponsored violence, although present
(Amnesty International 2008), has not been as noticeable as in
Guatemala; organized crime, youth gangs, and street crime constitute
the main features of violence (Moser and Winton 2002; UNODC 2007).
In Honduras, youth gangs, organized crime rings, and drug-trafficking
cartels have been accompanied by systematic violence against street
children and youth, committed by state-affiliated death squads (Amnesty
International 2003; U.S. Department of State 2008). Meanwhile,
Nicaragua, the least violent postwar society, had to deal with the activ-
ities of armed bands of demobilized combatants during the early years
of the posttransition and then with an increase in common crime, drug-
related crimes, and drug trafficking on the Atlantic coast (Moser and
Winton 2002; PNUD 2009; UNODC 2007).

What explains these varying levels of homicidal violence in Central
America? The literature points to diverse factors behind the crime surge
in the region. Poverty and inequality have consistently been recognized
as key variables behind crime (Chinchilla 2003; Moser and Winton 2002;
PNUD 2009; UNODC 2007). Economic globalization and the 1990s
implementation of neoliberal reforms exacerbated the effects of those
variables (Benson et al. 2008; PNUD 2009). They also produced social
disruptive processes of unemployment and migration (Rocha 2006;
Zinecker 2007), dynamics of translocation and segregation of urban
spaces (Baires 2003; Lungo and Martel 2003; Rodgers 2009), and the rise
of private security companies (Ungar 2007). Another set of factors com-
monly cited are war legacies. These range from the disbanding of
former combatants (Chinchilla 2003; Cruz 1997; Cuadra 2002) to the pro-
liferation of arms in the postconflict era (Godnick et al. 2002; Moser and
Winton 2002) to the continuation of a sort of culture of violence (Cruz
1997; Godoy 2006). The rise of criminal economies around the transna-
tional drug business, state weakness, and the existence of a predomi-
nantly young population have also been pointed to as driving factors
behind Central American violence (PNUD 2009; UNODC 2007).

All these factors play important roles, as the various research agen-
das have shown, but they do not really account for the differences
between Nicaragua and the northern triangle. After all, Honduras, which
did not face internal conflict, has higher levels of violence than
Nicaragua, which suffered a decade of internal war. Nicaragua is also
considered one of the poorest and most unequal countries in the region
(Programa Estado de la Región 2008), and many of its problems regard-
ing economic transformation, disruptive and segregating urbanization,
privatization of public security, arms availability, and young population
are to be found in the other countries as well (Lungo and Martel 2003;
PNUD 2009; Zinecker 2007). None of these factors—not even the drug-
trafficking enclaves that operate along Nicaragua’s Atlantic coast

CRUZ: VIOLENCE IN CENTRAL AMERICA 5

(Orozco Betancourt 2007; UNODC 2007)—have made Nicaragua as vio-
lent as the north. To be sure, there are specific processes that may
enhance some variables in one country and not in another, but none of
the previously quoted factors seems to render a definitive explanation
about the gap between Nicaragua and the rest.

One theory that has gained momentum and that taps into the dif-
ferences between Nicaragua and the northern triangle puts the blame
on the development of U.S.-type street gangs, locally known as maras.
According to this argument, maras are largely responsible for the sky-
rocketing levels of violence in northern Central America (Arana 2005;
Boraz and Bruneau 2006), and their activities may explain the notice-
able differences between these countries and Nicaragua, where maras
have been unable to gain a foothold. In 2007, the United Nations Office
on Drugs and Crime estimated 70,000 gang members roaming the streets
of Guatemala, El Salvador, and Honduras. These groups, made up of
two large gang networks, the Mara Salvatrucha 13 (MS-13) and the Eigh-
teenth Street Gang (la Dieciocho), have gained notoriety as the most
powerful youth gangs in the region. However, their role in the overall
levels of crime is a matter of contention. While the Comisión de Jefes y
Jefas de Policía de Centroamérica y el Caribe (2003) claims that gangs
might be responsible for most extortions committed in Guatemala and
El Salvador and up to 45 percent of homicides in Honduras and El Sal-
vador, the UNDP contends that no more than 15 percent of felonies can
be credited to maras in northern Central America (PNUD 2009).

Contention about figures aside, youth gangs are important actors of
violence in Central America (Jutersonke et al. 2009), and their role in the
reproduction of crime is more significant in El Salvador, Honduras, and
Guatemala than in Nicaragua (USAID 2006). Dennis Rodgers (2006a)
has documented the evolution of Nicaraguan gangs from youth barrio
groups to a more economy-driven type of criminal organization and has
shown that gang violence in Nicaragua is more prominent than usually
presumed.

Even so, any comparison between northern maras and Nicaraguan
pandillas yields a blatant difference not only in terms of the number of
gang members, but also the violence wielded by these groups (UNODC
2007). Different authors (Rocha and Rodgers 2008; USAID 2006) have
ascribed these divergences in gang growth to the distinct flows of
returned migration and deportation in Central America. MS-13 and
Dieciocho have flourished in the northern triangle, this argument goes,
because of the deportation and voluntary return of migrant youths from
the United States to their home countries. There, they have formed new
cliques and spread U.S. gang culture. Conversely, in Nicaragua, gangs
have not been affected by youth deportation from the U.S. because most
of Nicaragua’s poor migrant communities have settled in Costa Rica and,

6 LATIN AMERICAN POLITICS AND SOCIETY 53: 4

to a lesser degree, South Florida, where they have not joined street
gangs (Rocha 2007).

Although return migration certainly plays a part in the development
of gangs in northern Central America in contrast to Nicaragua, it is mis-
leading to blame the growth of maras and the prevalence of violence in
the northern triangle mostly on migration. If we accept the argument of
migration as it is usually raised, we cannot explain why, after years of
Mexican cyclical migration and deportation from the United States,
maras—and particularly la Dieciocho, which was formed by Mexican
immigrants long ago—have not gained a foothold in Mexico as they did
in Central America.2 As argued elsewhere (Cruz 2010a), it is important to
consider the domestic conditions that made Guatemala, El Salvador, and
Honduras fertile ground for powerful gangs, but not Mexico or Nicaragua.

Likewise, the problem with ascribing the increasing levels of criminal
violence to youth gangs is that it obscures the institutional features behind
the development of youth gangs and the reproduction of violence. To be
sure, gangs produce violence, but the rise of maras is not the main factor
behind the differences in the increase of crime, because the same condi-
tions that enabled gang growth have also facilitated the spread of crimi-
nal violence in general. In the analysis of violence, we have to bring back
another variable. We have to bring state institutions back.

Rather than internal war or poverty, one of the fundamental under-
lying differences between northern Central America and Nicaragua is
the manner in which these states have dealt with public security and
have responded to problems of violent crime. Violence and maras are
partly consequences of such management. That management is articu-
lated not only in the institutional capacities and policies carried out by
the state, but also in the utilization of informal violent entrepreneurs and
armed civilian collaborators to deal with crime and disorder.

These plans are the result of the specific mode of transition experi-
enced by each country. In the northern tier, the parallel security agents
and the elites who supported them substantially survived the transitions,
and they have continued to operate long after the transitions concluded.
Nicaragua is a distinctive case because the complex interplay of condi-
tions set by its historical process, beginning with the 1979 revolution,
followed a different path of institution building in the security appara-
tus, one that insulated it, to some extent, from criminal organizations
and enabled it to construct a different relationship with the population.

TRANSITIONS AND THE CONSTRUCTION OF
INSTITUTIONS OF SECURITY

Six ideas are fundamental to understanding the importance of the tran-
sitions in shaping public security in Central America. The first is that

CRUZ: VIOLENCE IN CENTRAL AMERICA 7

building democracy entails not only the construction of institutions
intended to ensure transparent electoral processes, thus fulfilling the
minimalist Schumpeterian requisite; it also entails the construction of
institutions intended to promote the rule of law and citizens’ basic rights
(O’Donnell 2004). As Diamond (2008, 46) states, “The most urgent
imperative [for effective democracy] is to restructure and empower the
institutions and accountability and bolster the rule of law.”

Security and justice institutions are fundamental in the development
of democratic governance (Karstedt and LaFree 2006), particularly in
societies that are emerging from authoritarian regimes, because they
help to provide the basic conditions to make governance possible. Fol-
lowing Koonings (2001), the biggest challenges for a democratic gov-
ernment trying to escape its authoritarian past lie not only in subordi-
nating security forces to civilian rule, but also establishing the public
monopoly on legal force in order to secure order, the rule of law, and
citizenship rights based on accountable norms and procedures.

As Karl (1990) notes for the modes of democratic transitions, the
formal arrangements are negotiated by elites; that is, those who wield
power. The agreements are conditioned not only by organized interests
(Haggard and Kaufman 1997), but also by the context of political inter-
actions produced by the transition itself. They shape new notions of cit-
izenship and stateness (Yashar 1999). In this phase, those interactions
will shape how the intended pacts are carried out and what their effects
will be on the security apparatus, as the latter is a key asset in the polit-
ical bargains. Popular mobilization and external influence also take part
in those interactions, reinforcing or eroding previous institutional pat-
terns (Stepan 1988).

Following Charles Tilly (2003) in assessing the politics of violence,
we know that regime interactions entail violent entrepreneurs. The pacts
made during the transition and immediately thereafter are conditioned
by agents who hold power derived from their capacity for violence
(army, police, paramilitaries, etc.), from their links to networks of inter-
est mediation and masses, and from their relationships with external
states. Some of these agents survive the transitions to constitute legal
players in the new institutional arena, while others do not. However,
some others maneuver and use their connections to colonize shadowy
areas within institutions and state bureaucracy, whereby they keep
operating (Cruz 2007).

The survival of authoritarian, violent entrepreneurs with connec-
tions to the new regime means that the state’s participation in the gov-
ernance of security may involve not only formal institutions but also ille-
gal armed actors; that is, private and informal agents who, in some
cases, may mutate into criminal operations using state institutions as a
cover-up (Davis 2009; Schlichte 2009). Paramilitaries, vigilante groups,

8 LATIN AMERICAN POLITICS AND SOCIETY 53: 4

death squads, and partisan gangs connected to state apparatus are some
of these groups. The participation of these actors, which do not fully
abide by the rules, are part of what Guillermo O’Donnell (2004) calls
the “brown areas” of rule.

Last but not least, the state’s use of informal collaborators to manage
violence is a practice long embedded in the relationship between Cen-
tral American governments and their populations. As Robert Holden
asserts, pretransition security institutions achieved an extraordinary
autonomy in the use of violence “by a correspondingly high level of tol-
erance for and collaboration with the state’s agents of repression, among
nonmilitary agents of the state and within civil society itself” (1996, 459).
In other words, it is impossible to understand Central American state
violence in the past without recognizing that violence was wielded not
only by formal institutions but also by informal groups acting in collab-
oration with the state (Alvarenga 1996). The democratic transitions did
not wipe out this important feature in some countries. Illegal groups and
rogue agents linked to the state contributed to the dynamics of violence
in posttransition Central America.

REGIME CHANGE IN CENTRAL AMERICA:
THE THREEFOLD FOUNDATIONAL TRANSITIONS

The key feature of Central America’s recent history is the widespread
internal wars and peace processes experienced by Guatemala, El Sal-
vador, and Nicaragua. In contrast to most of Latin America, which did
not face civil wars, the entire isthmus was influenced by these conflicts.
Guatemala, El Salvador, Honduras, and Nicaragua were ruled by author-
itarian regimes for most of their history. In the late 1970s, political unrest
led to the intensification of armed conflicts between leftist guerrillas and
military governments in Guatemala, Nicaragua, and El Salvador. How-
ever, the victory of the Sandinista revolution in Nicaragua was followed
by a counterrevolutionary war (Booth et al. 2010). Liberal democracy
came not only with the celebration of relatively free and fair elections
but, more important, with the end of armed conflicts in the 1990s.

There has been a great deal of discussion on when the transitions
started and ended in the region. This is an important debate, but
chronology is not the main interest of this article. Instead, the transitions
can be framed around three types of events: authoritarian breakdown
through military coups and popular uprisings; the electoral process fol-
lowing the breakdowns; and institutional reform aimed to transform the
security apparatus. This classification facilitates comparison and helps to
underline the importance of two dimensions of democratic transitions:
elections and the rule of law. These dimensions do not always concur,
and their separation has critical consequences in creating what some

CRUZ: VIOLENCE IN CENTRAL AMERICA 9

have called “democratization backwards”: democracies that have intro-
duced free elections before establishing the basic institutions of rule of
law and civil society (Rose and Shin 2001).

Central America’s transitions were actually conducted within signif-
icant political constraints, such as civil war or military oversight (Karl
1995). Elections led to alternation in power in the 1980s, but the reforms
that were aimed at guaranteeing full observance of human rights and
the rule of law did not come about until the 1990s, as a result of polit-
ical pacts. Table 2 shows the road to liberal democracy in each country,
highlighting the key events and the complexities of regime change.
Some authors (e.g., Torres Rivas 2001) have seen these regime changes
as comprising three parallel transitions: from war to peace, from military
to civilian rule, and from authoritarianism to democracy.

From War to Peace

The transition from internal wars to political peace is the main charac-
teristic of democratization in the region. In Guatemala and El Salvador,
peace was intrinsically linked with reforms included in the formal
treaties; whereas in Nicaragua, the 1988 Sapoá Treaty established the
conditions that led to the 1990 watershed elections (Torres Rivas 2001).

In Guatemala, the end of the war came in 1996 as a result of a series
of peace accords between the government and the URNG (Unidad Rev-
olucionaria Nacional Guatemalteca) guerrilla forces. However, since the

10 LATIN AMERICAN POLITICS AND SOCIETY 53: 4

Table 2. The Road to Liberal Democracy in Central America

Preceding First
Regimea Starting Point Election Ending Point

Guatemala Military Military coup 1984 Peace Accords
authoritarianism (1983) Regime Reforms

(1996)

El Salvador Military Military coup 1982 Peace Accords
authoritarianism (1979) Regime Reforms

(1992)

Honduras Military Elections (1980) 1980 Reforms in public
authoritarianism security

(1995–99)

Nicaragua Traditional Revolution 1984 Reforms as result
dictatorship (1979) of 1990 elections

aWith the exception of Honduras, the characterization of the preceding regime is
from Mahoney 2001.

military had virtually defeated the guerrillas by the mid-1980s through a
campaign directed against rural and indigenous communities (Schirmer
1998), the outcome was a treaty that did not diminish the real power
exercised by the military forces and the political elites associated with
them.

The resolution of the war in El Salvador is best described as a mil-
itary and political stalemate. Despite solid support from the United
States, the Salvadoran government was unable to defeat the FMLN
(Frente Farabundo Martí para la Liberación Nacional) guerrilla forces.
The international context and pressure generated conditions that led to
the 1992 Chapultepec Accords. Given the virtual stalemate, the Sal-
vadoran peace treaty was the most ambitious in terms of state reform
and democratization (Karl 1995).

The Nicaraguan civil war was the product of a counterrevolutionary
effort driven mostly by U.S. financial and strategic support. After the
overthrow of the Somoza dictatorship by the Sandinistas in 1979, the
United States helped to organize the National Resistance (Resistencia
Nacional), a force integrated by former Guardias Nacionales, Misquitos
(indigenous people from the Atlantic coast), and peasants. Although the
Sandinistas had practically won the war by the end of 1988, the perma-
nent U.S. blockade, the enormous human cost, and a shattered econ-
omy forced the regime to negotiate peace in 1988. The opposition won
the 1990 elections, and a long process of pacification started.

By 1996, political peace was the common feature in Central Amer-
ica. However, wars shaped the conditions under which political elites
negotiated the transitions and impinged on the quality and depth of the
peace agreements. State security apparatus, the military and police, were
substantially reinforced during the conflicts. Repressive institutions
gained an unprecedented political leverage, given their role in protect-
ing the regime. Even in warless Honduras, the U.S. utilization of the
country as a platform to project military assistance in the region led to
an enormous increase in military influence. Paradoxically, all this was
taking place at a time when electoral processes were promoted as sig-
nals of democratization. Consequently, the end of the wars was intrinsi-
cally tied to the need to confront military influence (Torres Rivas 2001).

From Military to Civilian Rule

Whether because of the internal wars or the authoritarian character of
the Central American regimes, the military played a central role in gov-
erning these countries. The military’s initial withdrawal from executive
offices in the early 1980s did not reduce its political power; instead, by
the end of the decade, Central American armed forces were wealthier,
more powerful, and more autonomous than ever before. For that

CRUZ: VIOLENCE IN CENTRAL AMERICA 11

reason, political transitions aimed to demilitarize the internal security
apparatus. Even in Nicaragua, the conspicuously partisan military that
emerged from the revolution was as powerful as the former Guardia
Nacional. As Kruijt has argued, the counterinsurgent campaigns of the
Sandinista army …