RULE OF LAW WORKING PAPER
OCTOBER 2017
I
nsecurity, crime, and state weakness are parts of
everyday life in much of Central America. Most
homicides and other crimes go unreported or unsolved
and law enforcement, judicial, and correctional systems
are overloaded, corrupt, and ineffective. Despite decades
of effort on remedies, the underlying security situation
remains largely the same, if not worse, in some countries.
Facing dangerous and daunting contexts, individuals
modify their behaviors in ways that have personal,
economic, societal, and even transnational consequences.
A focus on these dynamics can reveal opportunities for
strategic programming to curtail the damaging effects of
crime and violence to the region.
The problem is particularly severe in the three “Northern
Triangle” countries–El Salvador, Guatemala, and
Honduras–which regularly rank among the most violent
countries in Latin America and the world. (Note: This report
uses “Northern Triangle” to refer to these three countries
and “Central America” for those three plus Nicaragua,
Costa Rica, and Panama.) In 2016, El Salvador recorded
81.2 homicides per 100,000 people, the highest rate in the
Americas (with the possible exception of Venezuela, where
ofcial statistics are typically incomplete or unavailable).
Honduras and Guatemala reported rates of 59 and 27.3
By asking individuals not just about
crime victimization and perceptions
of crime, but also about their
behavior patterns in the face of
crime, the study delves deeper into
how and why insecurity aects the
social, political, and economic fabric
in Central America.
BENEATH THE VIOLENCE
HOW INSECURITY SHAPES DAILY LIFE AND
EMIGRATION IN CENTRAL AMERICA
A Report of the Latin American Public Opinion Project and the
Inter-American Dialogue
Ben Raderstorf, Carole J. Wilson, Elizabeth J. Zechmeister, and Michael J.
Camilleri
per 100,000, respectively—lower than El Salvador but
still high enough to rank as third and fth most violent in
the Latin America and Caribbean region.
1
The reality of
high homicide rates registers among the population: in
the Northern Triangle, more than half the population has
“some” or “a lot of fear” of being a victim of homicide, with
El Salvador the most affected (see Figure 1).
2
The main drivers of violence—beyond a long history of civil
war, political instability, and weak judicial institutions—are
Beneath the Violence
2
Foreword
We are pleased to present “The Toll of Crime on Daily Life
and Intention to Emigrate in Central America,” a joint report
by the Latin American Public Opinion Project (LAPOP)
at Vanderbilt University and the Peter D. Bell Rule of Law
Program at the Inter-American Dialogue.
This report, by Ben Raderstorf, Carole Wilson, Liz
Zechmeister, and Michael Camilleri, addresses critical
questions about how insecurity impacts everyday life in
Central America and how violence shapes behaviors from
economic activity to migration. Based on approximately
9,300 in-person interviews conducted across nationally
representative samples in Guatemala, Honduras, El
Salvador, Nicaragua, Costa Rica, and Panama as a part
of the 2016/17 AmericasBarometer, these ndings paint
a detailed portrait of the ongoing toll that crime takes
on countries in Central America. They also point the way
towards some potential solutions.
This report represents the culmination of a year-long
collaboration between LAPOP and the Inter-American
Dialogue that attempts to connect the dots between
opinion polling on security and actionable, concrete policy
recommendations. At the end of the report, the authors
provide a list of policy guidelines for decision makers in
Central America, the United States, and elsewhere.
LAPOP is a center for excellence in international survey
research. Its core project is the AmericasBarometer, a
regular study of how citizens experience and evaluate
democratic governance in 34 countries. LAPOP’s mission
is four-fold: produce high quality public opinion data;
develop cutting-edge methods in international survey
research; build local capacity in the eld of survey research
and analysis; and, disseminate timely results with rigor and
clarity.
Established in 2015 with support from the Ford Foundation
and named in honor of a founding Dialogue co-chair, the
Peter D. Bell Rule of Law Program of the Inter-American
Dialogue strives to elevate policy discussions around
democratic institutions, government accountability, human
rights, and citizen engagement in Latin America.
The data used in this report are available free of
charge at LAPOP’s website: www.lapopsurveys.org.
Extensive information on LAPOP’s methods and the
AmericasBarometer survey can be found at the same
website. The AmericasBarometer survey has been made
possible because of support from the United States
Agency for International Development (USAID), Vanderbilt
University, the Inter-American Development Bank, the
United Nations Development Programme, the Open Society
Foundation, and a network of other partners across the
Americas. The opinions expressed in this study are those
of the authors and do not necessarily reect the point of
view of Vanderbilt University, the Inter-American Dialogue,
USAID, or any other sponsor or partner to the study.
We are grateful to Kevin Casas-Zamora, a non-resident
senior fellow at the Dialogue, for his instrumental role in
helping initiate and shape this collaborative project and
for his comments on drafts. We also thank USAID and the
Igarapé Institute, headquartered in Brazil, for input into
some of the questions that are used in this study.
MICHAEL SHIFTER MITCHELL A. SELIGSON
President Founder and Senior Advisor,
Inter-American Dialogue LAPOP
Centennial Professor of
Political Science,
Vanderbilt University
The 2016/17 AmericasBarometer shows that
fear of crime leads large percentages of the
population to alter their daily activities—
avoiding public transit or making purchases,
keeping children at home, changing jobs or
place of study, moving neighborhoods, and
even considering emigration.
RULE OF LAW WORKING PAPER | OCTOBER 2017
How Insecurity Shapes Daily Life and Emigration in Central America
3
gang activity and drug trafcking.
3
There are an estimated
54,000 gang members across the three countries,
4
divided
among groups that compete, often violently, for territory
and resources. These gangs, known as maras, often seek
to extract value directly from the communities. As a result,
criminal extortion is rampant. A 2015 estimate found
that Salvadorans alone pay an estimated $400 million in
extortion and protection fees to gangs and other criminal
groups.
5
Extortion, in turn, leads to country-wide networks
of fear and intimidation, tightly constraining economic
activity in many areas and sectors. According to an
estimate by the Inter-American Development Bank, crime-
related costs in Guatemala, El Salvador, and Honduras are
equal to 3%, 6.2%, and 6.5% of GDP, respectively.
6
In recent years, violence in the region has received
signicant international attention as a major factor
in a wave of migration to the US
7
(see the 2014
AmericasBarometer Insights brief on “Violence and
Migration in Central America”
8
). In 2014, when the
crisis peaked, US Border Patrol agents apprehended
nearly a half million people at the US-Mexico border, in
large part families and unaccompanied children from
the Northern Triangle.
9
While migration rates have
since fallen somewhat—dropping sharply in 2015 but
bouncing back in 2016—much of that decline has come
from increased enforcement efforts on the part of the
Mexican government.
10
There are also anecdotal reports
that the Trump Administration’s policies and rhetoric
have discouraged migration, at least temporarily and
especially for unaccompanied minors.
11
In any case,
evidence suggests that the underlying desire to emigrate
remains strong. The 2016/17 AmericasBarometer nds
that intentions to move abroad have risen signicantly in
every country in Central America since 2014, especially in
Honduras.
12
At the same time, governments in the Northern Triangle
have been mostly unable to respond effectively to
problems of crime and security. Prison systems are
massively overcrowded, with one estimate placing El
Salvador’s current prison population at a staggering
348.2% of capacity.
13
Meanwhile, criminal justice systems
are widely seen as corrupt—and often with good reason:
11.6% of adults in the Northern Triangle report being asked
to pay a bribe to a police ofcer in a twelve-month period.
14
FIGURE 1: LEVEL OF FEAR OF BEING A VICTIM OF
HOMICIDE
Source: © AmericasBarometer, LAPOP, 2016/17
27.9%
27.5%
22.7%
22.0%
Guatemala
35.7%
20.9%
23.4%
20.0%
El Salvador
32.6%
20.1%
21.9%
25.4%
Honduras
Source: Ó AmericasBarometer, LAPOP, 2016/17
Level of Fear of Being a Victim of Homicide
A Lot of Fear
Some Fear
Little Fear
No Fear at All
This project asked questions in all six countries about “Out of fear
of crime, in the last 12 months…
Have you avoided leaving your home by yourself at night?”
Have you avoided public transportation?”
Have you prevented children from playing in the street?”
Have you felt the need to move to a different neighborhood?”
We also asked three questions in Guatemala, Honduras, and El
Salvador about “In the last 12 months…
Have you avoided buying things that you like because they may
get stolen?”
Have you changed your job or place of study out of fear of crime?”
Have you considered migrating from your country due to
insecurity?”
And one question in Guatemala only: “Out of fear of crime, in the
last 12 months…
Have you kept your minor children from going to school for fear of
their safety?”
Data collection for the Central American countries included in
the 2016/17 AmericasBarometer was conducted in late 2016
(Costa Rica, El Salvador, Honduras, and Nicaragua) and early
2017 (Guatemala and Panama). The samples are designed to be
nationally representative. In each country, approximately 1,550
adults were interviewed face-to-face.
Beneath the Violence
4
Together, all these factors paint a concerning picture of
the region. That is, the enormous burden of crime and
insecurity on the Northern Triangle—and how it generally
relates to migration, corruption, instability, and a lack of
economic opportunities—is well-documented.
However, understanding the more specic impact of
insecurity on individuals’ lives – who is more likely to
take precautions against crime and to what consequence
– is often hindered by a lack of granularity and clear,
consistent, and comparative data. National indicators,
especially crime trends beyond homicides, are often
inconsistent, of poor quality, and at times politicized.
There is also often a gap in understanding between those
statistics and the individual-level picture painted by the
many high-quality eld and ethnographic studies that have
been conducted over the years.
15
It is at times difcult to
elaborate clearly on how the security environment affects
the daily functions of a society as a whole and how it
drives economic activity, migration, and other trends.
By asking ordinary citizens to report on their own
circumstances, public opinion studies provide a window
into micro-level dynamics. However, a comprehensive
assessment requires sizeable modules on crime and
violence, in order to drill into the topic. For example,
the 2016/17 AmericasBarometer nds that nearly equal
proportions of the public in Uruguay and Honduras cite an
issue related to security as the most important problem
facing their country.
16
Yet across these diametrically
opposed countries – Uruguay among the safest in the
region and Honduras among the most violent in the world
– individuals’ specic experiences and concerns vary.
Via the inclusion of more detailed questions, surveys
provide a means to reveal exactly how problems of crime
and insecurity manifest in a particular context, and with
what consequences for individual behaviors and societal
outcomes.
This report, which presents ndings gathered as a part of
LAPOP’s 2016/17 AmericasBarometer surveys, begins to
ll in some of these gaps.
17
Analyses of the survey data,
which were collected in Guatemala, El Salvador, Honduras,
Nicaragua, Costa Rica, and Panama, more concretely
measure and diagnose the human and societal impact
of insecurity. By asking individuals not just about crime
victimization and perceptions of crime, but also about
their behavior patterns in the face of crime, the study
delves deeper into how and why insecurity affects the
social, political, and economic fabric in Central America.
In presenting key ndings from the project, we aim to shed
more light on the many complex and pressing security
concerns in the region—and inform discussions about how
to address them.
1. How Insecurity Shapes
Behavior in Central America
Measured by changes in behavior, the toll of insecurity
on daily life in Central America is widespread and
signicant. The 2016/17 AmericasBarometer shows that
fear of crime leads large percentages of the population
to alter their daily activities—avoiding public transit or
making purchases, keeping children at home, changing
jobs or place of study, moving neighborhoods, and even
considering emigration.
56.7%
67.6%
60.8%
69.2%
63.6%
65.9%
59.6%
36.9%
59.8%
46.8%
48.6%
56.6%
31.1%
42.2%
47.4%
25.5%
32.4%
50.2%
21.1%
15.0%
15.4%
19.4%
19.7%
19.5%
Panama
Costa Rica
Nicaragua
Honduras
El Salvador
Guatemala
Panama
Costa Rica
Nicaragua
Honduras
El Salvador
Guatemala
Panama
Costa Rica
Nicaragua
Honduras
El Salvador
Guatemala
Panama
Costa Rica
Nicaragua
Honduras
El Salvador
Guatemala
0 20 40 60 80
Prevented Children from Playing in the Street
Has Avoided Leaving House Alone at Night for Fear of Crime
Has Avoided Using Public Transportation for Fear of Crime
Has Felt Need to Move Neighborhoods for Fear of Crime
Percentage
95 % Confidence Interval
(with Design−Effects)
Source: Ó AmericasBarometer, LAPOP, 2016/17
FIGURE 2: CRIME AVOIDANCE BEHAVIORS IN CENTRAL
AMERICA
Source: © AmericasBarometer, LAPOP, 2016/17
RULE OF LAW WORKING PAPER | OCTOBER 2017
How Insecurity Shapes Daily Life and Emigration in Central America
5
As seen in Figure 2 and Figure 3, the distribution of impact
varies signicantly across behavior type and country
in Central America. Unsurprisingly, crime avoidance
behaviors are most common in Guatemala, Honduras,
and El Salvador. This is consistent with other statistics
on crime and security in the region, which generally nd
a sharp divide between the Northern Triangle and its
neighbors immediately to the south. Yet, surprisingly high
proportions of the population in Nicaragua, Costa Rica, and
Panama report having changed their behavior, particularly
when it comes to preventing their children from playing on
the street or feeling the need to move neighborhoods. This
suggests that fears about family safety are more rigid—and
less connected to the underlying crime rates—than fears
about individual safety. That approximately two-thirds of
Central Americans (63.9%) report having prevented their
children from playing in the street may reect a widespread
state of panic, or it may simply reect a general tendency
toward caution in the care of others, even in the face of
comparatively smaller risk.
18
Digging deeper into the Northern Triangle specically—
measuring purchasing habits, changing jobs or place of
study, migration, and school attendance—we nd that the
behavioral impact of crime is particularly strong when
it comes to economic activity. In Guatemala, Honduras,
and El Salvador, just over half of the adult population
(50.5%) reports having avoided buying “things they like”
in the past year because they may be stolen.
19
These data
afrm the assumption that underlies assessments based
on aggregate economic data: crime and insecurity deter
individuals from spending, to the detriment of the country’s
FIGURE 3: CRIME AVOIDANCE, ECONOMIC ACTIVITY,
AND MIGRATION IN THE NORTHERN TRIANGLE
Source: © AmericasBarometer, LAPOP, 2016/17
48.6%
51.2%
51.7%
35.1%
17.2%
37.1%
9.4%
12.4%
11.1%
Honduras
El Salvador
Guatemala
Honduras
El Salvador
Guatemala
Honduras
El Salvador
Guatemala
0 10 20 30 40 50 60
Has Avoided Buying Items for Fear of Robbery
Considered Migrating in Last 12 Months because of Insecurity
Has Changed Jobs or Study Locations for Fear of Crime
Percentage
95 % Confidence Interval
(with Design−Effects)
Source: Ó AmericasBarometer, LAPOP, 2016/17
68.6%
67.5%
69.2%
68.9%
64.3%
59.6%
61.3%
58.4%
62.6%
52.2%
46.3%
46.3%
53.1%
42.8%
45.5%
44.5%
51.8%
54.1%
52.2%
50.1%
29.8%
31.2%
27.4%
28.4%
32.9%
19.0%
20.3%
19.0%
22.0%
19.6%
9.4%
14.0%
12.2%
10.4%
8.0%
5
4
3
2
1
5
4
3
2
1
5
4
3
2
1
5
4
3
2
1
5
4
3
2
1
5
4
3
2
1
5
4
3
2
1
0 20 40 60 80
Prevented Children from Playing in the Street
Has Avoided Leaving House Alone at Night for Fear of Crime
Has Avoided Using Public Transportation for Fear of Crime
Has Avoided Buying Items for Fear of Robbery
Considered Migrating in Last 12 Months because of Insecurity
Has Felt Need to Move Neighborhoods for Fear of Crime
Has Changed Jobs or Study Locations for Fear of Crime
Percentage
95 % Confidence Interval
(with Design−Effects)
Source: Ó AmericasBarometer, LAPOP, 2016/17
FIGURE 4: CRIME AVOIDANCE BY QUINTILE OF WEALTH
IN CENTRAL AMERICA
Source: © AmericasBarometer, LAPOP, 2016/17
Beneath the Violence
6
nancial situation. Additionally, 11% of those residing in
the Northern Triangle have changed their job or place of
study in the past year because of fear of crime.
20
This type
of outcome likely takes its own toll on the economy, and
also may be socially disruptive and destabilizing.
These crime avoidance behaviors are somewhat more
common in urban areas than in rural ones. For example,
23.3% of urban respondents in the Northern Triangle
report feeling the need to move neighborhoods due to
crime, while only 16.2% of rural respondents say the same.
There is only a small variation in responses by gender,
with female respondents slightly more likely to prevent
children from playing in the streets, avoid leaving the
house alone, and feel the need to move neighborhoods,
and male respondents slightly more likely to avoid public
transit. Variation in crime avoidance also exists across age
groups, with middle cohorts slightly more likely to engage
in crime avoidance—especially when it comes to pressure
to move neighborhoods: 25.7% of respondents between 36
and 45 years old report feeling a need to move, whereas
only 11.8% of those 66 years and older feel the same.
Interestingly, in most cases there is little variance in crime
avoidance behavior across wealth quintiles (see Figure
4). Among the exceptions are that those in the poorest
quintile are signicantly less likely to avoid leaving the
house alone at night and the wealthiest are signicantly
more likely to avoid public transit. We note that, while there
is not a statistical difference between the least wealthy
groups in their likelihood of avoiding public transportation,
the fact that 44.6% of the lowest two quintiles do so is
still a signicant and concerning nding.
21
Many of these
individuals are unlikely to have alternative forms of transit
(for example, 80 percent of adults in Honduras report
not owning a car
22
) and therefore exclusion from public
transport is a signicant blow to mobility and economic
opportunities.
Finally, in Guatemala specically, approximately one in
three respondents (31.7%) report having kept children
at home out of fear of crime (see Figure 5).
23
This is in
keeping with recent studies and reporting about a growing
number of children routinely missing school or college
because of fear of violence and criminal gangs. As
Francisco Benavides, a regional education adviser for Latin
America and the Caribbean at UNICEF described, “In some
areas of Latin America, we are talking about a second lost
generation.”
24
2. Creating and Validating a
Crime Avoidance Behavior
Index
With these ndings in mind, we create a crime avoidance
behavior index—a single score that reects how much
single individuals or specic populations change their daily
routines to seek security. This index, which is constructed
using the questions that were asked in all Central American
countries (VIC71, VIC72, VIC43, and VIC74; see earlier text
box for wording), can be used to measure the aggregate
impact of insecurity between countries (see Figure 6) as
well as compare crime avoidance behavior with other data
gathered as a part of the 2016/17 AmericasBarometer.
In short, the crime avoidance behavior index is a way of
measuring how much each individual goes out of his or
her way because of insecurity. It is essentially a numerical
shorthand for how “crime averse” any one person or group
is. With each individual assigned a score between 0 and
100, we can compare the aggregate impact of crime
avoidance across countries, as well as compare subgroups
based on various other traits and responses.
As expected, high index scores are associated with
increased perceptions of insecurity, crime victimization,
and gang presence. Those who have been a victim of any
type of crime in the last 12 months—including robbery,
burglary, assault, blackmail, fraud, and extortion—score
39.1% higher on the index than those who have not (60.1
versus 43.2). In other words, crime avoidance is higher
68.3%
31.7%
No
Yes
Kept Children Home from School Out of Fear of Crime
Source: Ó AmericasBarometer, LAPOP, 2016/17
FIGURE 5: GUATEMALANS WHO REPORT KEEPING
CHILDREN HOME FROM SCHOOL BECAUSE OF CRIME
Source: © AmericasBarometer, LAPOP, 2016/17
RULE OF LAW WORKING PAPER | OCTOBER 2017
How Insecurity Shapes Daily Life and Emigration in Central America
7
among those who are genuinely at risk for crime. Those
who report feeling “Very Unsafe” in their neighborhood
score almost twice as high on the crime avoidance index
as those who report feeling “Very Safe” (60.9 vs 33.7).
Respondents in neighborhoods without gang-related
grafti—as evaluated by the interviewer—engage in
measurably less crime avoidance behavior. Experiences
with police corruption and longer perceived police
response time also correlate with more crime avoidance
behavior (see Section 4).
In summary, the crime avoidance behavior index is a useful
and valid measure of the burden of crime in a population.
It correlates in the expected ways with indicators of crime,
gang presence, and weak rule of law and it is consistent
with observations made in macro-level studies and by
policymakers. Therefore, we can use the crime avoidance
behavior index to test how the daily impacts of security
are related to—and potentially the drivers of—various other
trends, above all the intention to emigrate.
3. Crime Avoidance and
Intentions to Migrate
Among the most important consequences of crime
avoidance, from a policy perspective, is migration. In
fact, crime avoidance behavior is one of the strongest
predictors of intention to migrate. This individual-
level dynamic helps explain why, in the past ve years,
intentions to “live or work in another country in the next
three years” have spiked in all countries in Central America,
especially in Honduras (see Figure 7).
4%15%28%28%24%
4%14%24%30%28%
7%17%28%26%22%
9%25%29%23%14%
9%23%28%25%16%
9%25%29%23%14%
Costa Rica
Nicaragua
Panama
Guatemala
El Salvador
Honduras
None 1 Item 2 Items 3 Items 4 Items
Crime Avoidance Behavior
Source: Ó AmericasBarometer, LAPOP, 2016/17
34.1
36.5
40.0
45.8
47.4
48.0
Nicaragua
Costa Rica
Panama
El Salvador
Guatemala
Honduras
0 10 20 30 40 50
Crime Avoidance Behavior
95 % Confidence Interval
(with Design−Effects)
Source: Ó AmericasBarometer, LAPOP, 2016/17
FIGURE 6: THE CRIME AVOIDANCE BEHAVIOR INDEX
ACROSS COUNTRIES
Source: © AmericasBarometer, LAPOP, 2016/17
Crime avoidance behavior is one
of the strongest predictors of
intention to migrate and helps
explain why intentions to do so
have spiked in all countries in
Central America.
Beneath the Violence
8
In the Northern Triangle specically, intentions to migrate—
as expected—are robustly linked to factors related to
insecurity. In fact, analyses of the survey data reveal
that security concerns play a central role in individual
motivation to migrate. 29.8% of adults in the Nortern
Triangle have considered migrating in the last 12 months
specically due to insecurity, as seen in Figure 8.
25
The
pressure to migrate due to insecurity is especially high in El
Salvador and Honduras when compared to Guatemala. This
rate is comparable to the 34.7% of adults in the Northern
Triangle that intend to live or work in another
country in the future regardless of motivation,
which suggests that many potential migrants
are driven by security, not just a search for
economic opportunities or other factors such as
family unication.
This is reinforced by comparing crime avoidance
behavior index scores in the Northern Triangle
to intentions to migrate for any reason. To
perform this analysis, we predict individuals’
intention to migrate out of the country with
measures of their individual characteristics,
security evaluations and behaviors, corruption
experiences, trust assessments, and economic
situation. Interestingly, crime avoidance
behavior is as strongly associated or more
strongly associated with migration than almost
all other factors measured, including gender,
age, wealth, perception of
neighborhood insecurity,
interpersonal trust, trust
in local and national
governments, experience of
bribery, crime victimization,
fear of being murdered,
gang activity, and change
in household income. Only
unemployment and receiving
remittances—which means
the respondent likely has
relatives abroad—have larger
regression coefcients than
the crime avoidance index.
These results are presented
in Figure 9, where the dots
and associated numbers
indicate the estimated effect
of a maximum unit change
in the independent variable
(y-axis) on individuals’
intentions (0-100 likelihood) to migrate out of the country.
The lines on either side of the dots represent the 95%
condence interval for the coefcient. Solid dots are
statistically different from zero, whereas hollow dots are
not.
These ndings provide strong empirical evidence for
the chorus of arguments that recent migration from the
Northern Triangle has been driven by “push” factors in
addition to “pull” factors. The most-discussed example of
70.2%
29.8%
No
Yes
Considered Migrating because of Insecurity
17.2%
37.1%
35.1%
0
10
20
30
40
Considered Migrating because of Insecurity
Guatemala El Salvador Honduras
95 % Confidence Interval
(with Design−Effects)
Source: Ó AmericasBarometer, LAPOP, 2016/17
FIGURE 8: PRESSURE TO MIGRATE REGIONALLY AND BY COUNTRY
Source: © AmericasBarometer, LAPOP, 2016/17
0
10
20
30
40
Intends to Live or Work Abroad
2004 2006 2008 2010 2012 2014 2016/17
Survey Wave
Honduras
El Salvador
Nicaragua
Guatemala
Panama
Costa Rica
Country
Source: Ó AmericasBarometer LAPOP, 2016/17
FIGURE 7: A DRAMATIC INCREASE IN INTENTIONS TO MIGRATE
Source: © AmericasBarometer, LAPOP, 2016/17
RULE OF LAW WORKING PAPER | OCTOBER 2017
How Insecurity Shapes Daily Life and Emigration in Central America
9
this argument is the 2014 report “Children on the Run” by
the UNHCR, which interviewed 404 unaccompanied child
migrants and found “violence, insecurity and abuse” to be
the primary reasons they had risked the journey.
26
A 2014
Inter-American Dialogue survey of migrants from Central
America also found that “for Salvadorans and Hondurans,
violence in their country of origin was by far the leading
push factor, while for Guatemalans it was both violence
and the lack of opportunities” with family reunication
“secondary to the more immediate and pressing issues
of violence and poverty.”
27
The results in Figure 9 are
consistent with this pattern, showing that the toll of
insecurity on individuals’ daily lives – via their experiences
and their crime coping behaviors –is a key driver of
individual intentions to migrate, alongside factors that
indicate economic insecurity.
This suggests that US immigration control efforts that
focus exclusively on domestic policies and border security
are unlikely to be successful in deterring migration in the
long run. Even if changes in immigration policies or rhetoric
in the US result in a drop in border crossings—as many
have argued occurred in the initial months of the Trump
Administration
28
—the decline is likely to be temporary. As of
August 2017, there was already evidence that the number
of undocumented migrants apprehended at the US-Mexico
border was rising quickly back towards pre-Trump levels,
with a 22.5% month over month increase from July.
29
This is
especially true for families crossing the border, as opposed
to unaccompanied minors; apprehensions of family units
1.9
4.2
−1.4
2.0
11.7
9.6
12.3
−7.9
−6.2
5.4
9.5
5.5
10.6
Receives
Remittances
Loss in
Household
Income
Unemployed
Interpersonal
Trust
Police Response
Time
Police Officer
Asked for a
Bribe
Trust in
National Police
Trust in
Executive
Gang Signs or
Graffiti on the
Walls
Level of Fear
of Being a
Victim of
Homicide
Victim of Crime
in the Last 12
Months
Perception of
Neighborhood
Insecurity
Crime Avoidance
Behavior
−15 −10 −5 0 5 10 15
95% Confidence Interval
with Design−Effects
Source: Ó AmericasBarometer, LAPOP, 2016/17
FIGURE 9: PREDICTORS OF INTENTIONS TO MIGRATE
Source: © AmericasBarometer, LAPOP, 2016/17
68.5
60.1
49.6
58.1
64.4
45.1
56.2
59.9
56.2
52.6
51.3
37.5
54.2
46.6
52.2
46.5
38.9
46.0
45.0
33.9
31.5
41.7
36.6
44.1
47.9
4 Items
3 Items
2 Items
1 Item
None
4 Items
3 Items
2 Items
1 Item
None
4 Items
3 Items
2 Items
1 Item
None
4 Items
3 Items
2 Items
1 Item
None
4 Items
3 Items
2 Items
1 Item
None
0 20 40 60 80
Interpersonal Trust
Trust in Local Government
Trust in National Police
Trust in the National Legislature
Trust in Executive
Average
95 % Confidence Interval
(with Design−Effects)
Source: Ó AmericasBarometer, LAPOP, 2016/17
FIGURE 10: CRIME AVOIDANCE AND TRUST IN PUBLIC
INSTITUTIONS
Source: © AmericasBarometer, LAPOP, 2016/17
Beneath the Violence
10
rose 30% relative to a similar 2016 period. This should not
be surprising, as these results show that the pressure to
migrate because of insecurity is high even for adults.
These ndings suggest that in order to reduce the number
of undocumented migrants risking the journey from the
Northern Triangle in the long term, the most effective
strategy is one that attempts to improve conditions in the
three countries, particularly in terms of security, violence,
and crime. We also suggest that tracking crime avoidance
behavior can be a useful tool to identify the populations
most at risk and likely to migrate.
4. Countering the Toll on
Governments and Societies
Crime avoidance behavior also provides an important lens
into how security situations negatively affect democracy
and the state—and how governments can counter the tide.
Crime avoidance is associated with diminished trust in
the executive, the legislature, the national police, and local
government (see Figure 10, where crime avoidance appears
on the y-axis as a count of how of the crime avoidance
items in the index that received a positive response).
There is also clear evidence that crime avoidance behavior
is associated with lower levels of interpersonal trust, as
measured by whether the respondent thinks “people from
around here” are trustworthy.
30
According to the AmericasBarometer, those who have
experienced police corruption in the past year in the
Northern Triangle engage in signicantly
more crime avoidance behavior than those
who have not, as shown in Figure 11. There
are several possible explanations for this
relationship. On one hand, those who are
more at risk of being targeted by violent
crime may also be targeted more often
by police corruption. This would suggest
that police corruption is of most concern
in the communities that are most plagued
by violence. On the other hand, it may also
be that those who have had a police ofcer
ask for a bribe are less likely to trust law
enforcement and attempt to take control
of their own security. In any case, the link
between corruption and crime avoidance
behavior supports the growing consensus
that ghting corruption is critical to solving
the Northern Triangle’s pressing challenges, including
violence and migration.
31
There is also evidence, detailed in Figure 12, that longer
perceived police response time is associated with more
crime avoidance behavior. This makes sense for obvious
reasons: if individuals anticipate that the police take hours
to arrive (or won’t come at all), they can be expected to
take measures to protect their own safety. The relationship,
however, is less strong than one might expect, and is less
dramatic than having experienced police bribery. This
suggests that the integrity of law enforcement is more
important than the proximity of the police. Therefore, police
reform—including controls against corruption and the
implementation of high quality community-based policing
approaches—may be more effective than simply bolstering
the size and presence of police forces.
Finally, and perhaps most interestingly, there is strong
evidence that crime avoidance behavior (along with being
45.6
58.3
0
10
20
30
40
50
60
Crime Avoidance Behavior
No Yes
Police Officer Asked for a Bribe
95 % Confidence Interval
(with Design−Effects)
Source: Ó AmericasBarometer, LAPOP, 2016/17
FIGURE 11: POLICE CORRUPTION AND CRIME AVOIDANCE
Source: © AmericasBarometer, LAPOP, 2016/17
There is strong evidence that crime
avoidance behavior (along with
being a victim of a crime) correlates
with increased engagement and
activity in the community.
RULE OF LAW WORKING PAPER | OCTOBER 2017
How Insecurity Shapes Daily Life and Emigration in Central America
11
41.0
43.9
47.2
49.9
50.3
50.6
0
10
20
30
40
50
Crime Avoidance Behavior
Less than 10
minutes
10 to 30
minutes
30 to 60
minutes
1 to 3 hours Less than 3
hours
There are no
police / They
won’t come
Police Response Time
95 % Confidence Interval
(with Design−Effects)
Source: Ó AmericasBarometer, LAPOP, 2016/17
FIGURE 12: POLICE RESPONSE TIME AND CRIME AVOIDANCE
BEHAVIOR
Source: © AmericasBarometer, LAPOP, 2016/17
a victim of a crime) correlates with increased engagement
and activity in the community. Higher crime avoidance
behavior scores are associated with attendance at
meetings of religious organizations, meetings of a parents’
association at a school, and meetings of a community
improvement committee or association (see Figure 13).
32
Likewise, intentions to migrate are also associated with
greater attendance in these community organizations. In
other words, instead of self-isolating, those individuals
most alert to crime are more engaged with their community
even if they hope to escape. This nding ts with previous
research on the “associativity of distrust” in Latin
America—a positive correlation between fear and the
tendency to participate in local organizations.
33
As Kevin
Casas-Zamora argues, this engagement can be “based
more on reasons of convenience—to ght crime—than
on solidarity.”
34
This is part of a broader
pattern of engagement: research has also
shown that crime victimization leads to
increased participation in politics around
the world.
35
At the same time, increased participation
may be the product of informal
community organizations, dispute settling
mechanisms, and local norms that emerge
in poor communities outside the reach of
the state. It is important for future studies
to delve into variations in the types of
local participation and efforts that are
emerging in response to crime, since some
may be more conducive to democratic
deepening than others. A good example
is how drug trafckers maintain order in poor
neighborhoods, as documented by Enrique
Desmond Arias and Corinne Davis Rodrigues
in Rio de Janeiro.
36
They argue that “trafckers
create a ‘myth of personal security’ in which
individual residents believe they can guarantee
their own safety through their actions and
political connections to trafckers.”
37
In that
sense, individuals most vulnerable to crime
may feel the need to participate in community
organizations and institutions for the sake of
self-protection through personal relationships,
yet the extent to which those organizations
operate outside the connes of the rule of law,
or within it, varies.
In either case, this suggests a critical new
piece in this puzzle in the Northern Triangle:
even though the people most affected by security issues are
less trustful of their neighbors and more likely to want to
move out of their community, they may also be more willing
to work to try to improve it. This nding may help point the
way forward when it comes to lessening the burden of crime
and violence—and by extension, stemming the pressure to
migrate. These vulnerable populations may be turning to
local groups as a last resort. This may also make them a
potential focal point for policy interventions, either by the
state or by development organizations. The ndings from
the AmericasBarometer study suggest that the Central
Americans who are most affected by crime and violence
are not passive actors. They are turning to their community
institutions, either to get help or to try to improve the
community themselves. Those looking to assist them
should follow.
−0.3
−1.6
−0.6
5.4
4.3
Gang Signs or
Graffiti on the
Walls
Crime Avoidance
Behavior
Level of Fear
of Being a
Victim of
Homicide
Victim of Crime
in the Last 12
Months
Perception of
Neighborhood
Insecurity
−4 −2 0 2 4 6 8
95% Confidence Interval
with Design−Effects
Source: Ó AmericasBarometer, LAPOP, 2016/17
FIGURE 13: COMMUNITY PARTICIPATION BY VARIOUS FACTORS
Source: © AmericasBarometer, LAPOP, 2016/17
Beneath the Violence
12
RECOMMENDATIONS
Focus on “push factors” instead of “pull factors”
behind migration. These ndings provide clear
evidence for the argument that migration from the
Northern Triangle is driven to a large degree by concern
about crime and fear of violence. Policymakers in
the United States looking to stem migration pressure
should continue to focus on improving security
and economic conditions in the Northern Triangle
countries, rather than focusing exclusively on domestic
immigration and border security policies.
Think of security as an economic investment. The high
incidence of crime avoidance behavior—particularly
in employment and purchasing decisions—suggests
a clear link between insecurity and missed economic
opportunities. Measures aimed at improving citizen
security situations should be framed as long-term
economic investments.
Invest in communities. Evidence suggests that
community organizations and local institutions can
be an important resource for those most vulnerable
to crime and violence. Investing in communities at
the micro level may be a critical juncture, both when it
comes to lessening the burden of insecurity and as a
potential way to stem pressure to migrate.
RULE OF LAW WORKING PAPER | OCTOBER 2017
How Insecurity Shapes Daily Life and Emigration in Central America
13
Protect access to transportation and education. The
sheer number of Central Americans that avoid public
transit or keep their children out of school because
of fear of crime translates to a signicant loss of
opportunity with signicant long-term impacts on
economic productivity and other outcomes. Improving
secure access to both these services is critical when it
comes to lessening the burden of crime and violence.
Fight petty corruption and improve police to bolster
community security. While the exact mechanisms still
need to be studied, the link between petty corruption by
police and crime avoidance behavior is clear. Improving
the quality and efcacy of local police will help ease
the impacts of crime. Fighting corruption more broadly
may also help improve overall trust in institutions and
improve the quality of public services, which can also
help lessen the burden of insecurity.
Stay focused on Honduras. Within Central America, the
highest burden of crime avoidance falls on individuals
in this country. Despite the recent progress in reducing
homicide rates, Honduras remains the most affected
in terms of crime avoidance behavior and it has the
highest (and most sharply increasing) rates of
intention
to migrate.
Beneath the Violence
14
ENDNOTES
1. Gagne, David (2017). “InSight Crime’s 2016 Homicide Round-up.” InSight Crime. http://www.insightcrime.org/news-analysis/insight-
crime-2016-homicide-round-up
2. AmericasBarometer (2016/17). LAPOP. FEAR11. Thinking about your daily life, how much fear do you feel about being a direct victim
of homicide?
3. This assertion is supported by World Bank (2011). Crime and violence in Central America: A development challenge. Washington,
D.C., World Bank. https://siteresources.worldbank.org/INTLAC/Resources/FINAL_VOLUME_I_ENGLISH_CrimeAndViolence.pdf
4. International Crisis Group (2017). Maa of the Poor: Gang Violence and Extortion in Central America. Latin America Report no. 62.
Brussels, International Crisis Group. https://www.crisisgroup.org/latin-america-caribbean/central-america/62-maa-poor-gang-
violence-and-extortion-central-america
5. Dudley, Steven and Michael Lohmuller (2015). “Northern Triangle is the World’s Extortion Hotspot.” InSight Crime. http://www.
insightcrime.org/news-briefs/northern-triangle-world-extortion-hotspot
6. Jaitman, Laura et al. (2017) “The Costs of Crime and Violence: New Evidence and Insights in Latin America and the Caribbean.
Washington, DC, Inter-American Development Bank. https://publications.iadb.org/handle/11319/8133#sthash.A20urGQ5.dpuf
7. International Crisis Group (2016). “Easy Prey: Criminal Violence and Central American Migration.” Brussels, International Crisis
Group. https://www.crisisgroup.org/latin-america-caribbean/central-america/easy-prey-criminal-violence-and-central-american-
migration
8. Hiskey, Jonathan et al. (2014). “Violence and Migration in Central America.” AmericasBarometer Insights: 2014, no. 101. LAPOP.
https://www.vanderbilt.edu/lapop/insights/IO901en.pdf
9. Bennett, Brian (2016). “Illegal border crossings by Central American families increase again.” Los Angeles Times. http://www.
latimes.com/nation/la-na-border-crossings-20161017-snap-story.html
10. International Crisis Group (2016). “Easy Prey: Criminal Violence and Central American Migration.” Brussels, International Crisis
Group. https://www.crisisgroup.org/latin-america-caribbean/central-america/easy-prey-criminal-violence-and-central-american-
migration
11. Semple, Kirk (2017). “Central Americans, ‘Scared of What’s Happening’ in U.S., Stay Put.” The New York Times. https://www.nytimes.
com/2017/07/03/world/americas/honduras-migration-border-wall.html?mcubz=0
12. AmericasBarometer (2016/17). LAPOP. Q14. Note that the question asks about intentions to move abroad in general, and does not
reference a specic destination country.
13. World Prison Brief (2016). “El Salvador.” http://prisonstudies.org/country/el-salvador
14. AmericasBarometer (2016/17). LAPOP. EXC2.
15. See, for example, Nathan P. Jones (2013). “Understanding and addressing youth in ‘gangs’ in Mexico.” Woodrow Wilson International
Center for Scholars and the Justice in Mexico Project at the University of San Diego. https://www.wilsoncenter.org/sites/default/
les/04_youth_gangs_jones.pdf
16. AmericasBarometer (2016/17). LAPOP.
17. The analyses in this report are based on AmericasBarometer dataset version GM v07172017.
18. AmericasBarometer (2016/17). LAPOP. VIC74.
19. AmericasBarometer (2016/17). LAPOP. VIC40a.
20. AmericasBarometer (2016/17). LAPOP. VIC45N.
21. AmericasBarometer (2016/17). LAPOP.
22. AmericasBarometer (2016/17). LAPOP. R5 (Honduras).
23. AmericasBarometer (2016/17). LAPOP. FEAR6FA.
24. Grifn, Jo (2016). “Too afraid for school: Latin America is losing new generation to gang violence.” The Guardian. https://www.
theguardian.com/teacher-network/2016/jan/27/risk-life-qualication-education-latin-america
25. AmericasBarometer (2016/17). LAPOP. Q14A.
26. United Nations High Commissioner for Refugees (2014). “Children on the Run: Unaccompanied Children Leaving Central America
and Mexico and the Need for International Protection.” Washington, DC. http://www.unhcr.org/56fc266f4.html
27. Orozco, Manuel and Julia Yansura (2014). “Understanding Central American Migration: The crisis of Central American child migrants
in context.” Washington, DC, Inter-American Dialogue. http://www.thedialogue.org/wp-content/uploads/2015/05/FinalDraft_
ChildMigrants_81314.pdf
28. Partlow, Joshua (2017). “The ‘Trump effect’ has slowed illegal U.S. border crossings. But for how long?” The Washington Post.
https://www.washingtonpost.com/world/the_americas/the-trump-effect-has-slowed-illegal-us-border-crossings-but-for-how-
RULE OF LAW WORKING PAPER | OCTOBER 2017
How Insecurity Shapes Daily Life and Emigration in Central America
15
long/2017/05/21/dfa12a0a-39be-11e7-a59b-26e0451a96fd_story.html?utm_term=.d591388c5113
29. Isacson, Adam (2017). “After Post-Trump Slump, Cross-Border Migration is Increasing.” https://adamisacson.com/after-post-trump-
slump-cross-border-migration-is-increasing/
30. AmericasBarometer (2016/17). LAPOP. IT1.
31. Examples include Kevin Casas-Zamora and Miguel Carter (2017). “Beyond the Scandals: The Changing Context of Corruption in
Latin America.” Washington, DC, Inter-American Dialogue. http://www.thedialogue.org/resources/beyond-the-scandals-the-changing-
context-of-corruption-in-latin-america/; Cardenas, Jose (2015). “It’s Time for the U.S. to Tackle Corruption in Central America.”
Foreign Policy. http://foreignpolicy.com/2015/11/09/its-time-for-the-u-s-to-tackle-corruption-in-central-america/; Raderstorf, Ben
and Michael Shifter (2017). “Can Central America Break the Cycle of Drugs, Corruption and Gang War?” The Cipher Brief. https://
www.thecipherbrief.com/can-central-america-break-the-cycle-of-drugs-corruption-and-gang-war; Maute, Arturo (2017). “Guatemala
Stumbles in Central America’s Anti-corruption Fight.” International Crisis Group. https://www.crisisgroup.org/latin-america-
caribbean/central-america/guatemala/guatemala-stumbles-central-americas-anti-corruption-ght; Lapper, Richard. “Central America
Is As Violent As Ever. What Would it Take to Change?” Americas Quarterly. http://americasquarterly.org/content/central-america-
violent-ever-what-would-it-take-change
32. AmericasBarometer (2016/17). LAPOP. Dependent variable is an index of community participation based on CP6, CP7, and CP8.
Analysis includes Guatemala, Honduras, and El Salvador only. Country xed effects and controls for gender, urban/rural, age,
education, and wealth quintiles are included in the model but not shown in the gure.
33. Casas-Zamora, Kevin (2013). The Besieged Polis Citizen Insecurity and Democracy in Latin America. Washington, DC, Brookings
Institution. https://www.brookings.edu/wp-content/uploads/2016/06/citizen-insecurity-casas-zamora.pdf
34. Ibid. 8.
35. Bateson, Regina (2012). “Crime Victimization and Political Participation,” American Political Science Review, Vol. 106, No. 3, August.
https://www.cambridge.org/core/journals/american-political-science-review/article/crime-victimization-and-political-participation/8
69FE42104FD02272845F47BA09886A4
36. Arias, Enrique Desmond and Corinne Davis Rodrigues (2006). “The Myth of Personal Security: Criminal Gangs, Dispute Resolution,
and Identity in Rio de Janeiro’s Favelas,” Latin American Politics and Society 48, no. 4: 53-81.
37. Ibid. 54.
Cover photo: Samantha Beddoes / Flickr / CC BY 2.0
www.thedialogue.org
