Impact Evaluation of Crime Prevention Programs in 
Ciudad Juárez, Monterrey, and Tijuana 
By 
Pablo Parás, Ph.D. 
The Latin American Public Opinion Project (LAPOP) 
Vanderbilt University 
 
Implemented and Designed by LAPOP
Carried Out by Data OPM
This study was performed with support from the United States Agency for 
International Development. The opinions expressed in this study are those of the 
authors and do not necessarily reflect the point of view of the United States 
Agency for International Development. 
November 2014 

Impact Evaluation Report - Mexico 
3
Contents
List of Figures 4
List of Tables and Maps 7
Executive Summary: Main Findings 9
Background .......................................................................................................................................... 9
Main Findings 10
Section 1: Background of the Study 16
Difference-in-Difference Design 29
Section 2: The Design of the Study and Impact Evaluation Methods 34
Section 3: Results of the Impact Evaluation 39
I. Evaluating the Programs’ Effectiveness .................................................................................... 39
II. Analysis of Survey Data............................................................................................................... 40
Crime Victimization 42
Fear of Crime and Perception of Insecurity 51
Vulnerability of Young People in the Face of Crime and Gang Activity 62
Social Cohesion 73
Crime Prevention Measures 81
Perception of the Police 93
Government and Institutions 102
Conclusions 110
References 117
Appendices 121
Appendix A: Study Design: A Multi-Site “Cluster Randomized Experiment” (CRE) ............ 121
Selection of Study Areas 123
Stages of the Study and Types of Data Collected by LAPOP 125
Second Round Collection 2014 131
Data Collection Sites Mexico 132
Appendix B: Individual Questionnaire ......................................................................................... 134
Appendix C: Systematic Observation Questionnaire .................................................................. 165
Appendix D: Household Questionnaire ........................................................................................ 168
Appendix E. Pre-Visit Questionnaire ............................................................................................ 175
Appendix F: Focus Group Discussion Guide ................................................................................ 180
 
Impact Evaluation Report– Mexico 
4
List of Figures 
Figure 1. Homicide Rates in Mexico 1900 - 2012 ........................................................................... 16
Figure 2. Homicide Rates in Central America and Mexico 2000 - 2012 ........................................ 17
Figure 3. Homicide Rates in the states of interest in Mexico , 2006 – 2012 ................................. 19
Figure 4. Homicide Rates in the municipalities where the research was carried out in Mexico, 
2001 – 2011................................................................................................................................... 19
Figure 5. Fear of Crime in Mexico ............................................................................................... 20
Figure 6. Community Organization Due to Fear of Crime ........................................................... 20
Figure 7. Crime Victimization. ..................................................................................................... 21
Figure 8. Crime Victimization within Household......................................................................... 21
Figure 9. Key Demographic Variables in Treatment and Control Communities, 2012-2014 ...... 32
Figure 10. Illustration of Difference-in-Difference Estimator ...................................................... 41
Figure 11. Respondent-Reported Robberies in the Neighborhoods ............................................. 43
Figure 12. Marginal Change in Respondent-Reported Robberies in the Presence of Schools on 
the Block ....................................................................................................................................... 45
Figure 13. Total Crime Victimization ........................................................................................... 46
Figure 14. Sale of Illegal Drugs in the Community in the Past 12 Months .................................. 47
Figure 15.Marginal Change in Respondent-Reported Sale of Illegal Drugs in the Presence of a 
Catholic Church on the Block ....................................................................................................... 48
Figure 16. Marginal Change in Respondent-Reported Sale of Illegal Drugs in the Presence of a 
Hospital on the Block.................................................................................................................... 49
Figure 17. Fear of Crime in the Neighborhood............................................................................. 52
Figure 18. Marginal Change in Fear of Crime in the Presence of Hospitals on the Block ........... 53
Figure 19. Concern that the Home will be Robbed ...................................................................... 54
Figure 20. Marginal Change in the Concern that One’s Home will be Robbed in the Presence of 
Catholic Churches on the Block ................................................................................................... 55
Figure 21. Marginal Change in Concern that One’s Home will be Robbed in the Presence of 
Hospitals on the Block .................................................................................................................. 55
Figure 22. Perception of Insecurity of Living in the State ............................................................ 56
Figure 23. Marginal Change in the Perception of Insecurity of Living in the State in the Presence 
of Churches on the Block .............................................................................................................. 57
Figure 24. Need to Change Neighborhoods Due to Insecurity ..................................................... 58
Figure 25. Knows of an Institution that has Provided Support Activities for Young People ....... 63
Figure 26. Knows of an Institution that has Taken Steps to Help Young People Use their Free 
Time Better ................................................................................................................................... 64
Figure 27. Marginal Change in the Knowledge of Institutions Working for Young People to Use 
their Free Time in the Presence of Social Disorder on the Block ................................................. 65
Figure 28. Knowledge of a Young Person Who has Consumed Alcohol in Excess in the Past 
Year ............................................................................................................................................... 66
Impact Evaluation Report– Mexico 
5
Figure 29. Knowledge of a Young Person who has Consumed Drugs in the Past Year .............. 67
Figure 30. Marginal Change in the Knowledge of Young People who Consume Drugs in the 
Presence of Hospitals on the Block .............................................................................................. 68
Figure 31. Social Cohesion Index ................................................................................................. 74
Figure 32. Marginal Change in the Social Cohesion Index in the Presence of Social Disorder on 
the Block ....................................................................................................................................... 75
Figure 33. Marginal Change in the Social Cohesion Index in the Presence of Physical Disorder 
on the Block .................................................................................................................................. 76
Figure 34. Marginal Change in the Social Cohesion Index in the Presence of Churches on the 
Block ............................................................................................................................................. 77
Figure 35. Interpersonal Trust Additionally, it is important to note two things. .......................... 78
Figure 36. Interpersonal Trust in the Americas ............................................................................ 79
Figure 37. Basic Crime Prevention Knowledge ........................................................................... 81
Figure 38. Marginal Change in Basic Crime Prevention Knowledge in the Presence of Catholic 
Churches on the Block .................................................................................................................. 82
Figure 39. Government Provides Adequate Information on Crime Prevention ........................... 83
Figure 40. Police Carry Out Prevention Activities in this Neighborhood .................................... 84
Figure 41. Bought a Firearm for Self-Protection from Crime ...................................................... 85
Figure 42. Avoided Dangerous Areas out of Fear of Crime ......................................................... 86
Figure 43. Marginal Change in Avoiding Passing through Dangerous Areas in One’s Community 
in the Presence of Physical Disorder on the Block ....................................................................... 87
Figure 44. Built Walls/Fences to Protect Home from Crime........................................................ 88
Figure 45. The Community Is Organized for Crime Prevention .................................................. 89
Figure 46. Marginal Change in Community Organization for Crime Prevention in the Presence 
of Social Disorder on the Block .................................................................................................... 90
Figure 47. Marginal Change in Community Organization for Crime Prevention in the Presence of 
Catholic Churches on the Block ................................................................................................... 90
Figure 48. Trust in the Police in the Americas ............................................................................. 94
Figure 49. Trust in the Police ........................................................................................................ 95
Figure 50. Marginal Change in Trust in the Police in the Presence of Catholic Churches on the 
Block ............................................................................................................................................. 96
Figure 51. Police Get Involved in Recreational Activities ........................................................... 97
Figure 52. Marginal Change in Reports of the Number of Times that the Police Get Involved in 
Recreational Activities in the Presence of Schools on the Block ................................................. 98
Figure 53. Satisfaction with Police Performance .......................................................................... 99
Figure 54. Police asked for a Bribe in the Last Year .................................................................. 100
Figure 55. Asked for Help from the Police ................................................................................. 101
Figure 56. Public Works Carried Out in the Neighborhood in the Past Year ............................. 102
Figure 57. Perception that the Municipality is Improving Public Safety ................................... 103
Figure 58. Evaluation of Government Management of Citizen Safety in the Americas ............ 104
Impact Evaluation Report– Mexico 
6
Figure 60. Trust in the Judicial System ...................................................................................... 106
Figure 61. Marginal Change in Trust in the Judicial System in the Presence of Social Disorder on 
the Block ..................................................................................................................................... 107
Figure 62. Marginal Change in Trust in the Judicial System in the Presence of a Catholic Church 
on the Block ................................................................................................................................ 108
 
Impact Evaluation Report– Mexico 
7
List of Tables and Maps 
Table 1. Homicide Rates in (per 100,000 population) in Mexico by State 2006 - 2012 .............. 17
Table 2. Type of Implemented Programs by City and Neighborhood .......................................... 23
Table 3. Sample Distribution by Gender ..................................................................................... 28
Table 4. Sample Dates and Sizes for the Mexico Impact Evaluation ........................................... 28
Table 5. Summary of Qualitative Interviews, Rounds 1 and 2 - Mexico .................................... 29

Impact Evaluation Report– Mexico 
9
Executive Summary: Main Findings 
Background 
The USAID Missions in five Central American countries (Guatemala, El Salvador, Honduras, 
Nicaragua, and Panama) have administered and overseen the execution of the Central America 
Regional Security Initiative (CARSI) interventions—a set of programs with the objective of 
“reducing crime rates and improving security in Central America by strengthening community 
capacity to combat crimes and creating educational and employment opportunities for at-risk 
youth”.1
 
USAID/Washington, via its Cooperative Agreement with the Latin American Public Opinion 
Project (LAPOP) at Vanderbilt University, asked LAPOP to design and carry out an impact 
evaluation of the CARSI interventions in Guatemala, El Salvador, Honduras, and Panama. This 
evaluation was part of a broader effort to establish the effectiveness of USAID democracy and 
governance interventions through scientifically rigorous studies such as those recommended in 
the comprehensive study by the National Academy of Sciences (National Research Council 
2008).2
 LAPOP has more than 20 years of experience in carrying out policy-relevant surveys in 
Latin America, having conducted hundreds of country-based surveys, including many 
specialized studies designed to evaluate social programs. 
In the case of Mexico, via the Merida Initiative, the governments of the United States and 
Mexico collaborated to improve citizen security in Mexico. Both countries worked in 
conjunction to reduce the operational ability of organized crime; to strengthen institutional 
capability to maintain the rule of law; to create a 21st century border policy and to construct 
strong and resilient communities that are able to confront the pressures of crime and violence. 
The United States Agency for International Development (USAID) supports the efforts of the 
Mexican government in the construction of strong and resilient communities to foster social 
cohesion and mitigate the factors that contribute directly to violence and crime. The activities 
that are carried out during this effort focus on nine neighborhoods (polígonos) located in three 
cities on the Mexican-US border: Ciudad Juárez, Chihuahua; Monterrey, Nuevo León; and 
Tijuana, Baja California. 
 
1
 See, http://www.whitehouse.gov/ondcp/central-america, http://www.state.gov/documents/organization/148416.pdf. 2
 One of the authors of this report, Mitchell A. Seligson, was a member of the National Academy of Science study 
cited above. The impact evaluation was designed by Abby Córdova, who was also responsible for the 
implementation of the project in the first two rounds in her role as a post-doctoral fellow at Vanderbilt University 
and staff member at LAPOP. Dr. Córdova left the project in July 2012 before the final round of the study was 
completed to take up a position at the University of Kentucky. We also wish to acknowledge the constant and 
dedicated support of Eric Kite, Vanessa Reilly, and Jeremy Biddle at USAID Washington, as well as the extensive 
cooperation we received from USAID personnel in Mexico and Central America. We also much appreciate the 
cooperation we received from the implementing partners, and, most of all, the countless public officials and citizens 
of Mexico who gave freely of their time to make this impact evaluation possible. 
Impact Evaluation Report– Mexico 
10
These neighborhoods constitute large geographical zones (of one or more colonias) pre￾established and classified as high-risk and/or of being in need of attention. They were identified 
based off of specific criteria such as high levels of criminality and local violence prevention and 
social development programs. The programs implemented in each neighborhood may vary in 
theme and coverage and they do not necessarily have a direct impact on the whole population or 
on the entire geographical area that makes up the neighborhood. Said programs were 
implemented in a country that has experienced a significant increase in recent years in the main 
indicators of crime and violence, including both crime victimization and perception of 
insecurity.3
 
Based on the CARSI experience described above, LAPOP launched a study evaluating the 
impact of the programs implemented since 2013 in these nine neighborhoods. 
There are two key differences in the impact evaluation carried out in Mexico compared to those 
evaluations carried out in CARSI: in all of the countries except Honduras, LAPOP carried out 
three rounds of data collection (one base line and two follow-ups) and in Mexico, up to the 
present moment, only two. Furthermore, the design in Mexico is quasi-experimental given that 
the treatment neighborhoods were not randomly selected like in the CARSI countries that have 
an experimental design. 
From the beginning, the plan was to evaluate the programs implemented in the nine 
neighborhoods of the three cities selected for the crime prevention strategy that USAID 
implemented in conjunction with the Mexican Federal Government. Between September and 
December 2012, a “base-line” study was carried out that included a random sampling of 
households in each neighborhood, systematic observations of the infrastructure, social cohesion 
conditions within the colonia, a random sampling of individuals, and a complementary 
qualitative study using focus groups. Between March and June of 2014, we repeated the sample 
of individuals and the qualitative study. 
Main Findings 
To summarize the main findings of the impact evaluation of the community intervention 
programs implemented by USAID under the Merida Initiative in the neighborhoods of the cities 
of Tijuana, Monterrey, and Ciudad Juárez, we conclude that only some outcomes in the 
treatment communities improve more than what they would have improved if the USAID 
programs had not been implemented. Other indicators move in the correct direction but the 
change, compared with the change in the control neighborhoods, is smaller than what we could 
have expected, which may mean that the programs still are not having all the possible impact or 
that it is not going to improve more. However, a group of additional indicators in the treatment 
group move in the opposite direction than expected which implies that they get worse and/or that 
 
3
 For figures on the historical trends of said indicators, see INEGI, ICESI and/or the National System of Public 
Safety. 
Impact Evaluation Report– Mexico 
11
the programs have not been able to favorably impact the indicators or at least manage to contain 
their negative trend. 
Among the main results obtained in the quantitative phase of the study, we find that the USAID 
violence prevention programs have increased the support for the government or its actions in the 
treatment neighborhoods, specifically: 
 59% greater knowledge of the institutions that carry out actions for the youth to use their 
time better than would be expected in the absence of the programs 
 120% greater knowledge of the institutions that support the youth than would be expected 
in the absence of the programs 
 160% greater reported cases of police doing prevention activities than would be expected 
in the absence of the programs 
 17% greater reported cases of police working with children and young people in 
recreational activities than would be expected in the absence of the programs 
Regarding the positive improvements in the treatment group that show a smaller change than 
what would be expected given the implemented programs, we report that: (1) there is a reduction 
in crime victimization; (2) there is a decrease in fear of crime and insecurity perception, and (3) 
certain improvements in the prevention of violence and young people’s level of vulnerability. 
Regarding the changes in the opposite direction from the desired one, we report the following 
findings for the treatment neighborhoods: 
 Decrease in the social cohesion index 
 Decrease in interpersonal trust 
 Decrease in the perception that the community is organized to prevent crime 
 Increase in the reported knowledge of young people consuming drugs or alcohol in 
excess 
 Decrease in confidence in the police, police performance 
 Increase in the incidence of solicitation of bribes 
 Increase in individual fear indicators 
In most cases, the quantitative results of the effectiveness of crime prevention programs in the 
USAID treatment neighborhoods were complemented by qualitative information which helped 
to explain some of the results. It is important to point out that the qualitative component of the 
study is limited given that is not generalizable to the population as a whole and because it was 
carried out at the city level and not at the neighborhood level. For example, the police focus 
group, although it provided information on a key actor, could not be generalized to other 
individuals’ opinions. Furthermore, the police almost always have jurisdiction over a greater 
area than the communities in which we conducted our surveys, generally police precincts are 
assigned to the treated neighborhoods, control neighborhoods, and non-treated neighborhoods; 
Impact Evaluation Report– Mexico 
12
and police, as well, found it understandably difficult, if not impossible, to isolate a neighborhood 
out of all of those that they covered. The schools invariably attract students from various 
neighborhoods, some treated, some in the control groups and some non-treated. For this reason, 
we depend above all on quantitative information from the survey, comparing the treated 
communities and non-treated communities in our impact evaluation given that the surveys were 
conducted at the level for which the study was designed. Also, it is important to point out that in 
the CARSI countries, during the qualitative phase, we carried out, on top of focus groups, in￾depth interviews with key actors which allowed us, in some cases, to focus on particular 
community characteristics, which was not possible in Mexico. 
The qualitative information, however, has a number of advantages that we do not find in the 
quantitative information. One of these advantages is that through the qualitative information we 
see a greater perspective than the individuals’ perspective reported in our surveys which comes 
from their own limited experience in their neighborhoods. In the qualitative research, it is 
possible to have follow-up and in-depth questions to follow interesting lines of research. 
Nevertheless, with all its strengths, in contrast with the quantitative information, the qualitative 
information could not show the magnitude of the change over time nor could it control for 
factors that may be influenced by demographic and socio-economic variables in the same way 
our quantitative analysis could do. To summarize, we use a combination of quantitative and 
qualitative data in order to triangulate our analysis with the objective of presenting the most 
complete, most precise picture of the impact of USAID’s violence prevention program. 
Next, we describe some unique and key findings obtained in the focus groups that we consider 
relevant to USAID. In this executive summary, we highlight the points that we considered most 
important while carrying out the group sessions in Mexico whose objective was to evaluate the 
impact of crime prevention programs in Ciudad Juárez, Monterrey, and Tijuana. 
We remind the reader that the focus groups discussed crime victimization, violence, gangs, and 
organized crime in general and how these issues affect the youth, the priority interest group for 
the programs. The main results are: 
1. Family: both low- and high-risk youth consider that family is an issue that directly 
impacts their lives, especially with respect to the disintegration of family and the loss 
or lack of transmission of values within the family. They perceive a lack of parental 
attention which can create a series of significant problems such as drug addiction and 
juvenile delinquency. Children may grow up with and live with little company and 
adult supervision given that parents have to go to work. This factor can push they 
youth to take to the streets in search of companionship which may take the form of 
joining gangs that end up pushing them to vandalism and drugs. In part, the success 
of organized crime organizations depends on detecting these vulnerable groups, 
infiltrating the groups, and recruiting weak or abandoned youth. 
Impact Evaluation Report– Mexico 
13
2. School and work opportunities for the youth: young peoples’ lack of educational 
opportunities and low expectations for the their future employment possibilities may 
create feelings of being marginalized and of not having much hope of developing a full 
life. In spite of the fact that schools make an effort to retain students and provide them 
with recreational, athletic, and cultural activities, there is a high index of school desertion 
reported which can be explained by: having to support the family, behavioral or drug 
problems, or simply because there is no one to guide them or mentor them in a 
meaningful way. In particular, the youth, above all high-risk youth, are immersed in a 
very violent social atmosphere, with cases of vandalism and gang activity reported inside 
and outside of educational centers. 
3. School violence: in the opinion of many of the participants, school violence seems to 
have increased over the past few years; in part because of the diminished presence or 
total absence of authority figures in the school, be it teachers or school 
administrators, which has led to aggressions within the entire school population, both 
students and teachers alike. It is thought that part of the problem is the downfall of 
controlling structures and the absence of authority figures. It seems that teachers and 
students identify with each other and even maintain a peer-like relationship in which 
they exchange drugs and alcohol. This situation is not generalized; however it could 
be a significant problem in the long run. 
4. Insecurity: in the three cities where we carried out the study, the youth mentioned 
that their main problem is insecurity due to the fact that assaults, robberies and 
vandalism are close to their lives. In the same way, the adults that participated in the 
sessions (business leaders and public officials) mentioned that people live this 
problem on a daily basis because they feel threatened and extorted continually. It is 
reported that extortion is a common practice, above all practiced against business 
owners who have had to close their businesses and have their earnings cut 
significantly. Furthermore, they live terrorized by juvenile delinquents. Although it is 
incipient, there are methods of organization among the population to prevent crime in 
their communities, such as Vecino vigilante (Neighbor Vigilante), private security 
and community police letting people know about prevention strategies. 
5. Economic situation: it is reported as a serious problem for the population in general 
but it affects the youth in a significant way. In the surveyed households, salaries are 
low in the majority of the cases and young people have very few alternatives - and 
many times, it seems that continuing education is one of the least likely alternatives. 
Parents ask their children for help with the daily expenses and do not always get a job 
themselves. This could be the beginning of undesirable behavior such as robbery 
which can develop into assault and escalate from there into the aggressor becoming a 
member of a gang or criminal organization. In high-risk youth, we observe a direct 
relation between this situation and the increase in crime and even more, the 
enrollment of the youngest to organized crime. The younger they are (small 
children), the more vulnerable they are. 
Impact Evaluation Report– Mexico 
14
6. Drugs: for high-risk youth, the issue of drug addiction and drug and weapon 
trafficking is a constant in their lives. Many of these young people have been 
involved in problems with the law in their municipalities, which forces them to 
behave better. However, by being part of this criminal environment, they tend to 
recidivate and it is complicated to break out of this vicious cycle. It has been 
mentioned that programs that have been implemented to avoid the introduction of 
drugs into schools have been somewhat effective although not completely 
satisfactory. Sometimes, high-risk youth after having been in contact with the police, 
public officials or judges, have become less trusting of the authorities and institutions 
because they have been victims of abuse and corruption. 
7. Actions to improve security: a number of actions have been implemented to fight 
crime such as increasing the number of police officers patrolling the street and the 
involvement of marines and soldiers. According to some of the participants, even 
when an improvement in public safety is mentioned, it would seem that organized 
crime has resisted the implementation of actions/programs and has managed to get 
through to entire families that now dedicate themselves to and live off of drug 
trafficking (and cease to be interested in other legal money-making options even 
when the alternative, drug trafficking, runs a great risk). Under this system, there are 
children and youth that grow up and are educated in this environment and under this 
influence. 
8. Knowledge of programs: in general terms, we observe in the two rounds (2012 and 
2014) that people do not know much about the different programs that have been 
implemented and they do not have information about the particular objectives of each 
program or who runs them, what they are about, what their benefits are and even less 
about how to access them. The most frequently mentioned program is the D.A.R.E. 
(Drug Abuse Resistance Education) carried out by the police. This program is the 
most recognized out of other talks offered by the police that are seen as less credible 
just because the police give those talks – an institution that does not enjoy a great 
amount of trust amongst citizens. 
9. Perception of the methods and reach of organized crime: it is perceived that some 
modes of communication have been leaving out some of the activities that organized 
crime carries out. For this reason, it would seem that organized criminal activities 
have been decreasing but there is no information or news reports on this topic. 
However, one difference with respect to the 2012 round of surveys, is that before, 
these organizations were localized in 2 or 3 colonias within the municipalities, which 
were considered the most dangerous. However, in the 2014 round, the perception is 
that organized crime has extended to more colonias which implies that sales of drugs, 
extortions, and kidnappings have spread, and that drug traffickers are operating in 
practically every area of the city. 
Impact Evaluation Report– Mexico 
15
10. Preventative measures: different preventive measures have been identified by 
participants, such as Escuela Segura (Safe School), Mochila Segura (Safe Backpack), 
the presence of police when children enter and exit schools, private security in 
schools and businesses, closed-circuit television (CCTV), among others. These 
measures have tried to create greater security in both education centers and 
businesses. However, these measures are not perceived as key factors in the fight 
against organized crime. In spite of the fact that some public and private institutions 
such as the IMSS, DIF, or the Church carry out talks and recreational activities for 
the youth, their outcomes are not perceived. The reality, as some see, goes far beyond 
supporting young people of a certain profile. The truth is that the majority of children 
do not have a base of values on which to support themselves. 
11. Institutional Trust: there would seem to be a marked difference between insecurity 
and fear of organized crime. Those interviewed in the 2014 round of surveys feel 
more secure when walking in their neighborhoods, including at night (something that 
they have not done since a few years back) due to the fact that the police have 
improved their image, prompting people to feel more trust in them. However, with 
respect to organized crime, people keep being fearful to have to encounter it or report 
it. People know that the reprisals are enormous; reporting a member of a criminal 
organization would mean risking their integrity or the integrity of their family. 
Although now there are more ways to report and a little more security thanks to 
anonymous reporting, these mechanisms need to be more widely used. Reporting 
continues to be an issue that inspires fear and distrust because people not only fear 
reprisals from the criminals but they also do not trust public officials and institutions. 
12. Justice System: In 2014, the police, public prosecutors, and judges seem to do 
slightly better in their public image and performance. Respondents highlight that the 
judges and public prosecutors are now more difficult to corrupt than the police. 
Given that the police is the first institution with whom criminals have contact, the 
possibility of corruption rises since it is in their hands to let the offender go after 
being paid a bribe, neglecting to hand them over to the public prosecutor’s office. 
Although people acknowledge the effort that has been made to better train and 
improve the police, the difference in salaries between police and public prosecutors 
and judges makes the perception of corruption increase. 
This summary does not include the dozens of comments captured in the focus groups, some of 
which are presented going forward in the text of the report. 
Impact Evaluation Report– Mexico 
16
Section 1: Background of the Study 
This study evaluates the impact of USAID’s violence prevention programs in three states in 
Northern Mexico. The broader series of studies will also include the cases of El Salvador, 
Guatemala, Honduras, and Panama. In this section of the report, we present a brief background 
on the violence in Mexico and Central America that spurned the implementation of the anti￾violence programs under examination herein. Given that the literature on regional violence is 
vast, we do not summarize it here; rather our focus is on the impact evaluation itself and on the 
rich results that have emerged from the research. 
Mexico has been one of the countries where violence has grown the most over the last decade in 
Latin America. This growth can be clearly observed by reviewing the homicide rate data. These 
data show that after a slight reduction until 2004, the homicide rate climbed considerably (from 
8.5 homicides per 10,000 inhabitants in 2004 to 22.8 in 2012) as seen in Figure 1. 
Figure 1. Homicide Rates in Mexico 1900 - 2012
 Source: United Nations Office on Drugs and Crime, Homicide Statistics 2000-2011, INEGI 2012. 
In a comparative perspective with Central America, countries like Honduras, El Salvador, and 
Guatemala have significantly higher homicide rates than Mexico as seen in Figure 2. 
0.0
5.0
10.0
15.0
20.0
25.0
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
HOMIDICE RATES IN MEXICO
Impact Evaluation Report– Mexico 
17
Figure 2. Homicide Rates in Central America and Mexico 2000 - 2012
Source: United Nations Office on Drugs and Crime, Homicide Statistics 2000-2012
Even though the growth of violence in Mexico has impacted the entire country, said violence has 
regional components that cannot be ignored. There are states and regions in Mexico where 
expressions of violence such as homicide have a greater presence. In Table 1, we can see the 
evolution of homicide rates for each of the 32 states in Mexico. 
 
Table 1. Homicide Rates in (per 100,000 population) in Mexico by State 2006 - 2012 
State 2006 2007 2008 2009 2010 2011 2012 
Aguascalientes 2 4 5 6 6 7 4 
Baja California 16 12 33 48 47 25 18 
Baja California Sur 5 6 6 5 8 6 5 
Campeche 4 6 7 7 6 6 9 
Coahuila 4 4 7 10 16 26 41 
Colima 7 7 9 9 20 24 39 
Chiapas 12 2 6 11 4 4 8 
Chihuahua 20 15 76 105 182 126 77 
Federal District 9 9 10 11 12 12 12 
Durango 11 11 26 61 66 63 48 
Guanajuato 4 4 5 9 8 11 12 
Guerrero 24 23 30 54 45 70 77 
Honduras
El Salvador
Guatemala
Mexico
Panama
Nicaragua
Costa Rica
0
10
20
30
40
50
60
70
80
90
100
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
HOMICIDE RATES IN CENTRAL AMERICA
Impact Evaluation Report– Mexico 
18
Hidalgo 2 3 3 6 4 8 6 
Jalisco 7 6 8 9 14 20 20 
Mexico 12 8 10 12 14 17 18 
Michoacán 23 13 15 21 16 19 18 
Morelos 8 7 12 15 27 25 36 
Nayarit 10 10 15 18 49 52 24 
Nuevo León 4 6 5 7 20 45 38 
Oaxaca 14 15 16 16 19 17 18 
Puebla 6 5 6 6 6 7 8 
Querétaro 4 3 4 5 4 6 6 
Quintana Roo 6 10 11 11 11 12 11 
San Luis Potosí 6 6 8 8 14 14 17 
Sinaloa 17 14 30 51 85 69 48 
Sonora 10 12 17 21 27 20 19 
Tabasco 7 7 7 8 9 10 8 
Tamaulipas 11 6 8 10 28 32 46 
Tlaxcala 4 3 5 7 5 7 6 
Veracruz 5 5 4 9 6 13 13 
Yucatán 2 3 3 2 2 3 2 
Zacatecas 6 5 7 9 9 19 30 
Source: INEGI, 2012. 
As the data show, the northern region of the country is where we find the highest homicide rates 
and where this indicator has increased the most. This is clearly observed in the states of Sinaloa, 
Chihuahua, and Tamaulipas. In the south, the states of Guerrero and Morelos display the greatest 
increase in murder rates. 
Figure 3 shows the comparative evolution of homicide rates of the states where the three cities in 
our impact evaluation study are: Baja California, Chihuahua, and Nuevo León. Since 2007, the 
three states experienced a considerable increase in their homicide rates. In the graph, we observe 
a smaller incidence in Baja California and Nuevo León since their trends are softened by the 
sharp increase in Chihuahua – an increase especially explained by events of 2009 in Ciudad 
Juárez that led to a homicide rate of 186 killings per 100,000 inhabitants, making this Mexican 
region one of the most violent in the world. From 2010-2011, an increased presence of the 
federal government in these areas of the country decreased significantly the violence at the state 
level, although it should be underscored that this indicator remains above the national average. 
Impact Evaluation Report– Mexico 
19
Figure 3. Homicide Rates in the states of interest in Mexico , 2006 – 2012
Source: INEGI 2012.
In Figure 4, we can see that the homicide rate in the cities where the intervention and fieldwork 
were carried out is very high. In the three cases-Ciudad Juárez, Tijuana, and Monterrey- the 
indicator was higher than the national average in 2011, just one year before the implementation 
of the programs evaluated by LAPOP. In the extreme case, Juarez, the rate is 209, while it is 60 
in Tijuana, and 37 in Monterrey. It worth noting that in 2001, the homicide rate in Ciudad Juárez 
was 18, meaning that the growth from 2001 to 2011 was more than 1000% 
Figure 4. Homicide Rates in the municipalities where the research was carried out in Mexico, 2001 – 2011
Source: INEGI 2012, Secretaría de Salud, 2012.
0
20
40
60
80
100
120
140
160
180
200
2006 2007 2008 2009 2010 2011 2012
HOMICIDE RATES
Baja California Chihuahua Nuevo León Media nacional
0
50
100
150
200
250
2001 2006 2011
HOMICIDE RATES
Tijuana Juárez Monterrey Media nacional
National Average 
National Average 
Impact Evaluation Report– Mexico 
20
The preliminary data that the authorities have provided for 2012 and 2013 show a decrease in 
violence as much as for homicides as for other high impact criminal activities such as 
kidnapping. This appears to be due to diverse factors such as the change of focus in the anti￾violence strategies employed by the federal government, the participation of local government 
and the involvement of different sectors of civil society. This trend is not necessarily reflected at 
the national level in the AmericasBarometer’s indicators given that, as we observe in Figures 5 
and 8, the indicators of perception of fear and crime victimization have remained the same 
between 2012 and 2014. 
Figure 5. Fear of Crime in Mexico
Source: AmericasBarometer2012 2014.
Figure 6. Community Organization Due to Fear of Crime
87% 85%
13% 15%
2012 2014
Have you felt the need to move away from your neighborhood due to fear of 
crime?
No
83% 83%
17% 17%
2012 2014
Out of fear of crime, have you organized with the neighbors in your 
community?
No
Sí Yes
Yes 
Impact Evaluation Report– Mexico 
21
Source: AmericasBarometer2012 2014.
Figure 7. Crime Victimization.
Source: AmericasBarometer2012 2014.
Figure 8. Crime Victimization within Household.
Source: AmericasBarometer2012 2014.
USAID, via its Cooperative Agreement with the Latin American Public Opinion Project 
(LAPOP) at Vanderbilt University, asked LAPOP to design and carry out an impact evaluation 
of the interventions under the Merida initiative in Ciudad Juárez, Monterrey, and Tijuana. This 
evaluation is part of a broader effort to establish the effectiveness of USAID democracy and 
governance interventions through scientifically rigorous studies such as those recommended in 
the comprehensive study by the National Academy of Sciences (National Research Council 
23% 23%
77% 77%
2012 2014
Have you been a victim of any type of crime in the past 12 months?
Sí
No
22% 25%
78% 76%
2012 2014
Has any other person living in your household been a victim of any type 
of crime in the past 12 months?
Sí
No
Yes 
Yes 
Impact Evaluation Report– Mexico 
22
2008). Other assistance organizations and government policymakers have similarly recognized 
the importance of program evaluations like one conducted by LAPOP (Sherman et al. 1998; 
Bamberger et al. 2010). 
While previous studies have shown that USAID assistance programs in favor of strengthening 
democracy and governance (DG) have been very helpful increasing democracy levels in 
recipient countries, there are few rigorous evaluations of how and under what conditions 
assistance is effective (Finkel et al. 2007; Azpuru et al. 2008; Finkel et al. 2008; Seligson et al. 
2009). To be clear, scores of evaluations have been done in the DG arena, but most of them have 
been post-hoc and have not included both baseline and control groups, with the result that there 
is no real way of knowing what the end-of-project conditions would have looked like in the 
absence of the program. The USAID Merida-CARSI program offers a unique opportunity for 
carrying out such a study, and LAPOP is uniquely suited to perform the evaluation. LAPOP has 
more than 20 years of experience carrying out policy-relevant surveys in Latin America and has 
conducted hundreds of country-based surveys, including many specialized studies designed to 
evaluate social programs. 
The implementation of scientifically rigorous evaluations of crime prevention programs is a 
particularly pressing issue for the Latin American region. Violent crime has increased sharply 
during the democratization period in the region. The region has, and has had for some years, the 
highest rates of social violence and crime in the world outside of Sub-Saharan Africa (Casas￾Zamora 2013). Increased insecurity poses high costs for Latin American economies through lost 
human and physical capital (e.g., PNUD 2005; CEPAL 2011; IDB 2012). Furthermore, as 
LAPOP surveys often find, fear of crime and crime victimization erode social cohesion and 
citizens’ confidence in core institutions—both core elements for democratic stability. In the face 
of these problems, there have been numerous efforts to prevent crime among at-risk youth, 
especially in countries with a significant gang presence, violence, and crime. However, as many 
studies point out, despite these valuable efforts, little is known about their true impact (Moser 
and Van Bronkhorst 1999; Moser, Winton, and Moser 2005) or consequently about how to 
prevent crime and violence effectively. In this sense, USAID’s initiative to support a 
scientifically rigorous evaluation of Merida-CARSI crime prevention interventions is of great 
importance for broadening academics’ and policy makers’ understanding of crime prevention, 
and in turn, for promoting effective governance in the region. 
Based on information from USAID’s Democracy and Government mission in Mexico, we 
created a table to show the programs that have been implemented at the city level and in the 
treatment neighborhoods. The programs are grouped in five broad categories: implementation of 
the master plan, youth, public sanitation, communication, and prevention. For some of these 
categories, as Table 2 shows, there is more than one type of activity. The actions that took place 
after August of 2013 are highlighted in yellow to show that many of the programs that we are 
evaluating in this report have been recently implemented. 
Impact Evaluation Report– Mexico 
23
Additionally, we show that for some of the categories such as prevention strategies, the city 
coverage is disparate. These may be factors that could directly impact the results of our 
evaluation. 
Table 2. Type of Implemented Programs by City and Neighborhood 
Evaluation Design and Data Sources 
This impact evaluation study analyzes the effect of USAID’s violence prevention programs in 
nine high-risk neighborhoods. These neighborhoods were selected by the crime prevention 
strategy that USAID implements in three cities of northern states in conjunction with the 
Mexican federal government in order to determine whether or not they have been effective in 
achieving the following strategic objectives aimed at building strong and resilient communities: 
 Strengthen the ability and participation of the community to prevent and reduce crime; 
 Assist the capability of state and local governments in the implementation of crime 
prevention and violence reduction programs; and 
 Foster the participation of at-risk youth in prevention programs and recreational activities 
that make them less vulnerable to crime and violence. 
In order to determine the effectiveness of these programs, LAPOP developed a scientifically 
rigorous evaluation design known as “clustered randomized experimental” (CRE) design. The Improve Community Organizations Capacity Conflict Resolution ,social networks to prevent violence Stregthen NGOS Stop youth involvement in crime Youth model Victims Assistance and Community improvement Comunication Crime prevention and social assistance Planification support to crime prevention strategy
Felipe Angeles X Y X Y X X
Francisco y Madero X Y 1/2 Y X X
Reiveras del Bravo X Y X Y X X
Independencia X Y X X X X
Nuevo Almaguer X Y X X X X
La Alianza X X Y X X X
Granjas Familiares X Y XXX Y X X X
Camino Verdde X Y XX Y X X X
Mariano Matamoros X Y XX Y X X
X= indicates the number of programs
Y= indicates implementation at city level
1/2= indicates program implemented only in some poligonos
Master Plan Implementation Youth Prevention
Poligono
Impact Evaluation Report– Mexico 
24
main features of LAPOP’s research design, following the National Academy of Sciences’ 
recommendations are: 
 (1) Collection of data in the selected units before and during the implementation of the 
programs; ideally, a baseline study and two mid-point measurements, but in the case of 
Mexico, so far, only one measurement has been collected after the baseline study. 
 (2) Random assignment of units to either a “treatment” or “control” group (i.e., units that are 
to be treated by the program and units that in all other respects are eligible for but do not 
participate in or benefit from the program). In the case of Mexico, the assignment of 
treatment neighborhoods was not done randomly since they were chosen under the criteria of 
the federal program known as Convivir Mejor (Live Better Program). 
The study’s point of departure is the cities and communities within the selected cities for the 
crime prevention programs implemented by USAID in conjunction with the Mexican federal 
government.4
 The cities selected for those strategies are Ciudad Juárez, Monterrey, and Tijuana. 
Within each of these three cities, and following the Mexican government’s criteria, three 
geographical areas or neighborhoods were selected for the implementation of specific action 
programs with community involvement. These nine delineated geographical areas constitute the 
treatment neighborhoods where the impact evaluations of action programs were implemented. 
For the purposes of this report, the terms colonias, comunidades, and polígonos will be used 
interchangeably. (All three terms mean “neighborhood” in English). 
Within each of the three cities, we selected three control neighborhoods with the follow general 
selection criteria in mind: (a) that they are not adjacent or very close to the treatment 
neighborhoods; (b) that they have similar characteristics of the treatment communities; and (c) 
that crime prevention programs were not going to be implemented within them. 
Additionally, five communities were selected – two in Ciudad Juárez, one in Monterrey, and two 
in Tijuana– that were classified as “no-risk” whose main characteristic was that their economic 
and social indicators were comparatively better than the treatment and control neighborhoods. 
The former yields a total of 23 neighborhoods: 9 treatment, 9 control and 5 “no-risk” 
communities that make up the entire geographic reach of our research. 
Details on the control and “no-risk” neighborhood selection procedure are described later in this 
chapter. In Section 2, we present a map for each city where we identify the treatment and “no￾risk” neighborhoods. 
Quantitative data consist of: 
(1) Baseline surveys 
 
4
 These programs are independent and/or additional to others that the federal, state, or municipal governments in 
these cities could have had. 
Impact Evaluation Report– Mexico 
25
The baseline data characterize the population of study in both treatment and control groups in 
order to assess the starting condition, or the baseline condition, from which changes over time 
would be measured. Quantitative baseline information was collected in the second half of 2012. 
Phase I: Pre-visit of neighborhoods: As a first step in the survey process, researchers 
determined the eligible control neighborhoods (comparable to the treatment ones). The 
information that we collected in the pre-visits helped us to determine whether or not the 
neighborhoods fulfilled the pre-established criteria to be compared to the treatment 
neighborhoods. A short questionnaire by neighborhood was developed in order to take 
the selection's predetermined criteria into account. 
Phase II: Enumeration of individuals in the selected neighborhoods: The objective of 
the enumeration was to determine (1) the current number of households and their 
characteristics in each neighborhood and (2) the current number of individuals living in 
each household and their characteristics. The enumeration data enabled LAPOP to 
generate a sampling frame for the design of the individual-level surveys. 
 
LAPOP administered a short questionnaire for the enumeration (Appendix C) to 8,177 
households. This enumeration information was only collected in the baseline round to 
serve as a sample framework from which to draw the samples for the individual surveys 
that make up the quantitative base of the impact evaluation. 
Phase III: Systematic observation of selected neighborhood: The systematic 
observation of the selected neighborhoods provided an objective measure of the 
resources and physical and social disorder in the neighborhoods. 
Literature on the problem of crime and violence suggest that the context in which one 
lives is important. A systematic observation of each one of the neighborhoods included 
in the sample was carried out by trained observers using a specialized survey at the 
neighborhood level developed by LAPOP for each of the selected neighborhoods 
(Appendix B). Measurements were taken at the level of the block (portion of the city 
block, typically one of the four sides of a rectangular city block) and include the 
resources (such as schools and churches) and risk factors (such as abandoned houses, 
trash in the street, and other signs of social decay) observed in the neighborhood. Those 
observations at the city block level were added to the observations at the neighborhood 
level so that LAPOP could create a context variable for each one of the neighborhoods. 
The context information was used as part of a “multi-level” statistical analysis in order to 
determine whether or not we could detect some impact of the context on the results. 
 Phase IV: Qualitative neighborhood-level survey of stakeholders: 
The face-to-face questionnaire at the individual level developed by LAPOP (Appendix 
A) contains questions designed to measure results of the programs beyond socio￾economic indicators, demographics and other variables used as controls. The control 
variables allow us to statistically adjust the results for treated neighborhoods and non￾treated neighborhoods, and therefore to make them more similar to each one than what 
Impact Evaluation Report– Mexico 
26
they could have been otherwise. For example, a given community could have more 
young people than another, and since age is associated with many variables that affect 
crime, we need to control for this difference. 
On the other hand, there are unmeasured variables about each individual and 
neighborhood for which we could not control. This is merely a limitation of any social 
science project, not a problem with the design, because it is impossible to collect 
information on the infinite number of variables that define each individual (intelligence, 
life experience, presence of disease, reading ability, etc.). One adult per household 
(defined as a voting age adult) was interviewed in person at his or her home. The survey 
participants within the household were selected by sample quotas based on the 
enumeration data from Phase II to ensure that the sample reflected the demographics of 
the neighborhood. The fieldwork for the first round of individual level surveys in Mexico 
took place between September and November of 2012. (For more detailed information 
on dates of fieldwork, see Table 2.) 
Another source of information that we took into consideration was the research generated by 
LAPOP in their hemispheric survey called the AmericasBarometer. Every two years, LAPOP 
conducts surveys at the national level for 26 countries in Latin America and the Caribbean. For 
Mexico, we have biannual information from 2004 to 2014. These surveys, although they do not 
have the variety and quantity of questions on crime and violence included in the individual 
questionnaire of our impact evaluation, do take a number of important and identical questions 
into consideration on this issue. Therefore, we are able to draw comparisons against the results at 
the national level. 
(2) Final Evaluation: 
The final evaluation whose function was evaluating the results against the baseline was carried 
out for the majority of programs after 12 or more months of being implemented and was 
completed during the months of March and June 2014. In total, we carried out 6,113 interviews 
in the three cities (4,957 in the control neighborhoods, 4,383 in the treatment neighborhoods and 
2,735 in the “no-risk” neighborhoods). In this wave, we repeated research phases III and IV as 
well as the qualitative research using focus groups. For the systematic observations and 
individual interviews we returned to the same neighborhoods and in some cases to the same 
households interviewed in the baseline although we also included adults from other households. 
In both waves, the samples from each neighborhood were representative of and comparable to 
all the households. 
A summary of the waves, disaggregated by respondents’ gender, appears in the tables below. In 
the baseline, we obtained a distribution of adults with a percentage of women around 56%. We 
tried to replicate the same results for the second round, but it was not possible given the greater 
presence of women in the home and the mobility and rejection of males (see Table 3). 
We carried out a total of 2,815 systematic observations (information of the physical 
characteristics of the city blocks where we carried out the surveys). We obtained 1,222 
Impact Evaluation Report– Mexico 
27
interviews in the control neighborhoods, 673 in the “no-risk” neighborhoods and 920 in the 
treatment neighborhoods. In Table 4, we show the dates on which we carried out the interviews 
in both rounds. 
Impact Evaluation Report– Mexico 
28
Table 3. Sample Distribution by Gender 
Individual Level Respondents to the Impact Evaluation 
INCLUDING Non-at-Risk Neighborhoods
Percent Male (N) Percent Female (N) Total (N)
Baseline (2012) 43.3 (2,580) 56.7 (3,382) 100 (5,962) 
Final Evaluation
(2014)
38.0 (2,322) 62.0 (3,791) 100 (6,113) 
TOTALS 40.6 (4,902) 59.4 (7,173) 100 (12,075) 
EXCLUDING Non-at-Risk Neighborhoods 
Percent Male (N) Percent Female (N) Total (N)
Baseline (2012) 41.8 (1,957) 58.2 (2,726) 100 (4,683) 
Final Evaluation
(2014)
36.4 (1,694) 63.6 (2,963) 100 (4,657) 
TOTALS 39.1 (3,651) 60.9 (5,689) 100 (9,340) 
Table 4. Sample Dates and Sizes for the Mexico Impact Evaluation 
 At Risk Non-at-Risk
Treatment Control 
Round Dates Commu
nities 
Inter
views 
Comm
unities 
Inter
views 
Dates Comm
unities 
Inter
views 
Totals 
1 
9/12/2012 – 
11/04/2012 8 2,323 9 2,360 9/20/2012 – 
10/30/2012 5 
1,279 5,962 
2 
3/25/2014 – 
17/5/2014 8 2,060 9 2,597 
4/14/2014 – 
5/28/2014 5 
1,456 6,113 
Totals 16 4,383 18 4,957 10 2,735 12,075 
Beyond the quantitative research, LAPOP carried out qualitative research to have more 
reliable results. The interpretation of the dynamic of the communities under investigation 
and finally, of the impact of the project was reinforced a great deal by obtaining the 
qualitative information. The qualitative part was used in an interactive fashion as part of 
a multi-method strategy: the interviews with residents in the communities under 
investigation and focus groups. The questionnaires utilized are found in Appendix E. 
Impact Evaluation Report– Mexico 
29
Table 5. Summary of Qualitative Interviews, Rounds 1 and 2 - Mexico 
Round 1 Round 2 TOTAL 
Focus Groups 12 12 24
Difference-in-Difference Design 
The core of the quantitative study is an analysis of survey data using a “clustered randomized 
experimental” (CRE) design to allow for the comparison between the control and treatment 
groups over time. As we mentioned before, in the case of Mexico we have a quasi-experimental 
design given that the treatment neighborhoods were not randomly selected. A Difference-In￾Difference (DID) estimator is a widely used econometric technique in impact evaluation that 
dates back thirty years to the work of Ashenfelter and Card (1985)5
 and has become one of the 
“gold standards” in the field. It has the distinct advantage of controlling for what 
econometricians call “omitted variable bias,” since changes in the national economy, 
employment, and overall crime rates can impact the control and treatment groups, whereas in the 
case of a before-and-after study on the treatment group alone, it would be impossible to 
distinguish if such factors were responsible for changes in the treatment group rather than the 
treatment alone. It is, of course, not the only approach to measuring impact, but in the study we 
carried out, it was the most obvious and logical. For example, another widely used approach, 
that of regression discontinuity, works in situations in which detailed information about the 
population is already available (e.g., all have taken an entrance examination and some have 
scored just above the pass level and others have scored just below), but that was not the case in 
the violence prevention study in Mexico. 
The DID estimator was utilized to estimate both the difference between the treatment and 
control groups and the difference between each group and itself at a later time. The treatment is 
considered effective at the point of evaluation if outcomes in the treatment group improve 
significantly more (or decline less) than they would have if the treatment had not been 
administered. One of the most important features of this model is that it does not require that the 
treatment and control groups have identical starting values, since what is of interest is the trend
in each group and the difference between these trends.6
 We estimate a counterfactual case with 
 
5
 Orley Ashenfelter and David Card, “Using the Longitudinal Structure of Earnings to Estimate the Effect of 
Training Programs,” The Review of Economics and Statistics, Vol. 67, No. 4. (Nov., 1985), pp. 648-660. 6
 One challenge to the DID estimator is when one of the two groups is affected by something other than the 
treatment, but that factor is not present in the other group. Since our study relies on many neighborhoods and is 
spread across several municipalities, the chances of the occurrence of a situation like that are very low. As far as we 
know from the qualitative reports from the field, no such factor was present. We hasten to note that when a change 
occurs that affects both treatment and control, our results remain robust. The truce with the gangs, discussed in this 
report, is exactly one such change that was nationwide and thus affected both of our groups and did not undermine 
the results of the study. Another challenge to the DID estimator is when the composition of treatment or control 
changes over time. We checked for this, and found that on the key observables (education, wealth, age, etc.) the 
groups did not change over the course of the experiment. In short, the challenges to causal inference in this study are 
minimal but, as in all social science work, cannot be ruled out entirely, with the greatest challenge being that of 
external validity. That is, the treatment may be effective in the municipalities studied but nowhere else. However, 
Impact Evaluation Report– Mexico 
30
the model parameters to simulate what the outcome would be if no treatment had been 
administered. The counterfactual case is the change in the control group over time but at the 
same levels of the treatment group. Thus, the difference between the estimated counterfactual 
case and the predicted level7
 of the treatment group is considered the “treatment effect.” The 
“treatment effect” is the impact that can be said to have been caused by the intervention 
The multi-level nature of the data gathered by LAPOP and the design of the study made it 
possible for the outcome to be evaluated at the individual level while taking into account 
features of the neighborhood to determine the way in which the effectiveness of the programs 
varies across environments. Thus, it is possible to identify the features of neighborhoods (or 
even countries in cross-national studies) that contribute to the success of the program. As the 
number of countries in our study was very small, country characteristics could not be a major 
focus of our analysis since each country differs in innumerable ways from each other (GNP, 
income distribution, education levels, ethnicity, population density, etc.). In effect, we had too 
many variables and not enough countries. But variation at the community level, once all of the 
countries in the LAPOP CARSI impact evaluation are included, may well give us some 
analytical power and significant results. Using multi-level modeling techniques, the study draws 
statistical inferences about the effectiveness of the programs and the effect of the context in 
which the programs were implemented. 
The parallel paths assumption: threat to causal inference. The experimental design and DID 
estimator allow us to draw inferences from the collected data about the overall effectiveness of 
the programs. Like all methodologies, the DID estimator is based on assumptions that, if 
violated, can pose threats to these inferences. The most problematic assumption inherent in the 
DID estimator is that of parallel paths.
8 The DID methodology assumes that the trend observed 
over time in the control communities is the same as that which would have been observed in the 
treatment community if the treatment programs had not been implemented. For example, if there 
is an increase in perceptions of insecurity in the control communities over time, we would 
expect to see the same trend in the treatment communities if the treatment were not implemented 
(or if it were ineffective). The parallel path assumption refers to the changes in the outcome over 
time, not the starting or ending levels. This assumption is made explicit in the construction of the 
counterfactual case used to derive the treatment effect. The counterfactual case rests on an 
assumption of “parallel paths”—that the counterfactual experiences the same trend as that of the 
control group. 
 
Studies most vulnerable to this threat are those where assignment to treatment and control is not 
random and/or where assignment results in treatment and control groups that differed on factors 
that affected the outcomes of interest; that is the case in Mexico. For example, if treatment 
communities are clustered in gang territories but control communities are not, we might expect 
 since we replicated the study in several CARSI countries, external validity questions are of less concern than they 
would otherwise have been had the study been restricted to a single country. 
7
 The predicted level is the model estimate of the dependent variable, which controls for covariates, such as age, sex, 
education household wealth, years of residence, and children in the household, as well as contextual effects. 8
 See Michael Lechner. “The Estimation of Causal Effects by Difference-in-Difference Methods.” Discussion Paper 
no. 2010-28. October 2011. Department of Economics, University of St. Gallen. 
Impact Evaluation Report– Mexico 
31
different trends over time in outcomes related to crime in the two groups. Those were precisely 
the errors our study made every effort to avoid through the careful pre-selection process in phase 
I when we selected the control neighborhoods. 
While the CARSI study employed random assignment of treatment and control groups among 
the eligible communities in the municipalities selected by USAID, random assignment is not 
sufficient to guarantee that the assumption will hold, so the parallel paths assumption could still 
be violated. There was no way to directly test this assumption in this study (or virtually any 
other study), but we were able to largely rule it out through the use a variety of robustness 
checks to see if the parallel path assumption was likely to have been violated. These checks 
showed that the treatment and control groups were sufficiently similar to each other and to the 
nation as a whole, which gives us confidence that the assumption of parallel paths very likely 
held. Analyses of the treatment and control communities, as we detail below (Figure 9), show 
that the respondents did not significantly differ on key socioeconomic variables. This is an 
important finding since it suggests that the communities in both treatment and control were quite 
similar to each other in their makeup. Since the treatment and control communities were similar 
in composition, this suggests that the trends in these groups are also likely to have been the same 
over time. If, instead, we had observed significant differences between treatment and control 
groups on these variables, we might have expected the trends in these groups to diverge over 
time if the outcomes were correlated with these variables. 
A second—and we believe more telling—robustness check that gets to the core of our dependent 
variables is that trends in the control communities match those in national-level data. Several 
questions in the CARSI study are identical or similar to questions in the 2010 and 2012 
AmericasBarometer surveys that were also carried out by LAPOP using the same survey 
methods (face-to-face interviews conducted by trained interviewers using handheld electronic 
data collection instruments).9
 We were able to use these data to explore the trend across a similar 
time period for the nation as a whole. In the majority of these cases, we found that the trend in 
the national data was indeed parallel to that observed in the control communities (the national 
trend line is shown in our graphs along with the other results, where available). These results 
repeatedly demonstrate that the control communities were not an anomalous sample and that 
they reflected the same trends observed in the nation at large. Thus, since the treatment 
communities were selected by a matching process and have a similar socio-economic and 
demographic composition to the control group, and since the control group’s trends parallel 
national trends, we would expect to see the same trends in the treatment communities if the 
treatment had not been applied. 
However, it is worth noting that the sample selection procedures for the national study and those 
utilized for households and individuals within the treatment and control neighborhoods are not 
identical. Additionally, we did not have precise information on criminality and violence at the 
neighborhood level so that the matching procedure to identify control neighborhoods are based 
on socio-economic information and direct observation of the physical conditions of the 
 
9
 The cellular devices used to administer the questionnaire were used only in the 2014 round; on the baseline round, 
the administration of the survey was carried out with paper questionnaires. 
Impact Evaluation Report– Mexico 
32
neighborhoods and not necessarily of the levels of criminality and violence of each one. These 
two facts may have a significant impact on the parallel paths assumption. 
 2012 2014 
Figure 9. Key Demographic Variables in Treatment and Control Communities, 2012-2014
Results 
The quantitative results are directly related to the goals described earlier. As we have mentioned, 
they are based on the difference-in-difference estimator and a multi-level model that takes strata 
of the levels of city and community into consideration. The control variables of the models are 
age, gender, years of residence in the neighborhood, household wealth, education, and presence 
of young people in the home. The contextual models also include the percentage of people 
between 10-18 years of age in the neighborhood. 
As we mentioned in our discussion of the DID estimator above, a potential challenge for the 
validation of the type of design that we carried out are the possible changes in the composition 
of treatment and control groups over the passage of time. If, for example, in our study, the 
number of young people in the households would have decreased significantly, then whatever 
positive effect of the treatment could have in fact been the result of this decrease in place of the 
implemented treatment. The data in Figure 1 (above) show that a group of key variable did not 
significantly change in any of the treatment or control groups. 
Impact Evaluation Report– Mexico 
33
Detailed information on the design specifications of the study can be consulted in Appendix A of 
this report that contains the following sections: the selection of the areas of the study, the phases 
of the study and kinds of information collected by LAPOP, the methodological description of the 
second round of surveys in 2014 and information on data collection sites. 
Section 2: The Design of the Study and Impact Evaluation Methods 
USAID-Merida Interventions and their Anticipated Impacts 
The USAID Merida Interventions in the nine treatment neighborhoods include a variety of 
interventions. Many of these programs’ components could overlap strategically, which is a 
characteristic of the USAID Merida programs that are designed to create a programmatic 
continuity and to raise the probability of success for the entire crime and gang prevention effort. 
However, the coincidences and differences make it difficult if not impossible, to untangle and to 
evaluate the effect of each individual program, and even less the effects of each program’s 
specific components. In fact, it could be that each of the programs’ individual components 
would not be effective in isolation but only as part of a whole array of programs. Therefore, even 
a study designed to evaluate each component separately could not yield satisfactory results. For 
that reason, LAPOP has evaluated the global performance of the crime prevention programs 
implemented in the nine treatment neighborhoods. 
Given the ambitious goals of the USAID Merida programs, such as the increase in community 
participation in crime prevention, promotion of safe communities, and crime prevention among 
at-risk youth, the interventions may have direct and indirect effects. The direct effects include 
outcome indicators associated with the program’s final goal of producing safer and more secure 
environments. As we have already described, LAPOP evaluates those direct effects 
quantitatively relying on the survey information from the treatment and control communities 
instead of on external information that could have been obtained from other sources. In reality, 
we could have not relied on any other source of information at the community level. Municipal 
data are the most disaggregate available data, but the treatment does not cover entire 
municipalities. Therefore, only the data we collected can help us measure the programs’ impact. 
These direct effects include: fewer respondents’ reports of robberies, extortions and murders in 
the treated neighborhoods than would be expected in the absence of the program; lower levels of 
insecurity perception; and fewer youth in gangs and loitering. 
The indirect effects are related to the programs’ second-order consequences that may also be 
instrumental in the creation of safer, more secure environments and in the promotion of good 
governance. The indirect effects include the increase of interpersonal trust, satisfaction with 
democracy, and an increase in trust in local and national institutions such as favorable 
evaluations of the government handling citizen security or greater satisfaction with the 
performance of the police. All of these variables have been related with good governance in 
LAPOP’s studies results – an even with democratic stability. Therefore, all these indirect effects 
are potentially much broader and important in the long run. We should highlight that many of 
these indicators of democratic governance are measured by LAPOP in national samples in the 
AmericasBarometer, and therefore we have additional and extensive evidence at the national 
level and city level for the nations under that are being examined. 
Impact Evaluation Report– Mexico 
35
On the pages that follow, we document the key indicators that have been monitored over the 
course of the study. After controlling for a number of factors such as age, education, gender, 
household wealth, number of years of residence in the neighborhood and the presence of young 
people in the household, the observed improvement in these indicators could be interpreted as 
improvements attributable to the programs implemented in the treatment neighborhoods. The 
selection of the outcome indicators was based on: (1) an extensive review of the academic 
literature on the causes and consequences of crime and violence at the neighborhood level, (2) 
on-site observations by LAPOP of at-risk neighborhoods during the fieldwork in El Salvador, 
Panama, and Guatemala, and (3) interviews with experts and community leaders. 
The information was collected using quantitative and qualitative methodologies. The mixed￾method approach allowed us to evaluate the program’s impact as well as a broader 
understanding of the most effective aspects of the program (Bamberger et al. 2010). The 
quantitative data include household surveys, interviews with individuals and systematic 
observations of the neighborhoods. The qualitative part includes focus groups in the three cities 
that are being examined. A detailed explanation of the difference-in-difference design is 
provided later on in this report, but first we will describe the outcome indicators that are 
analyzed with the qualitative and quantitative methods. 
The indicators that we describe in this section represent only a segment of the whole list of 
indicators collected in the study, but they reflect the ones we think are the most relevant. Other 
questions in the survey serve as control variables. The questionnaires administered in Mexico 
are included in this report and they include all the variables in the study (see appendices). The 
questionnaires for the individual interviews were developed based on the literature review and 
they also included batteries of questions that had already been included in LAPOP’s 
AmericasBarometer survey. These questions were revised, pre-tested, and adapted to the local 
language. 
Outcome Indicators10
Fear of Crime 
By attenuating risk factors associated with crime and violence at the neighborhood level 
and thereby creating safer environments, USAID Merida programs were also expected to reduce 
residents’ fear of crime victimization. Fear of crime is a multidimensional concept associated 
with three main related dimensions: cognitive, affective, and behavioral (Fattah and Sacco, 
1989). As summarized by Franklin et al. (2008), “the cognitive dimension involves a rational 
thought process whereby perceptions of risk are developed; the affective dimension recognizes 
emotions associated with fear; and the behavioral dimension captures physical responses to the 
situation at hand.” Each of these dimensions is measured throughout the study. The attached 
individual-level questionnaire (Appendix A) includes specific measures for each of the three fear 
of crime dimensions. 
 
10 This section was originally drafted by Abby Córdova. 
Impact Evaluation Report– Mexico 
36
 Crime Victimization and Violence 
Changes in crime victimization and violence at the individual and neighborhood levels are 
assessed in the impact study. Measurements of crime victimization at the individual level identify 
the types of crimes and the place where they take place (at home, in the neighborhood, in the 
municipality, or somewhere else). Additionally, for a more extensive characterization and 
accounting of crime incidence, items asking whether non-respondent household members had 
been victims of crime were included in the LAPOP survey. In addition to personal/household 
member crime victimization, the survey asked respondents about their knowledge of the 
incidence of crime in the neighborhood. Worldwide, household surveys are considered to be the 
most reliable data on crime victimization, as other sources, such as police reports, are notoriously 
unreliable because of underreporting by crime victims. The LAPOP survey instruments used in 
the evaluation have been tested extensively and maximize the validity of reporting, yet, 
ultimately, our data only included reported incidents of crime since there was no practical way 
for the study team to do direct observations of criminal events. Survey items used to measure 
violent behavior at the neighborhood level (i.e., a violent argument between neighbors) were also 
included in survey questionnaires in order to capture the broader context. In order to understand 
the dynamics of vulnerable neighborhoods, both levels of crime and violence are measured. 
Retrospective Perceptions of Insecurity in the Neighborhood 
One expected result from the USAID Merida interventions is the creation of a safer environment. 
We anticipate that if this condition is fulfilled, the residents of the neighborhoods in which the 
programs are implemented will see improvements. Therefore, apart from measuring objective 
outcomes related to the safety and security of a neighborhood (for example, crime victimization), 
we also include perceptions or subjective indicators. 
Social Cohesion 
Evidence from studies of U.S. neighborhoods demonstrates that higher levels of social cohesion 
are associated with reduced crime and violence (Sampson, Raudenbush et al. 1997; Brown, 
Perkins et al. 2004). Social cohesion is typically defined as mutual trust and value sharing among 
neighbors. Given that the crime prevention interventions in the nine neighborhoods were 
expected to encourage community participation, social cohesion was also expected to improve. 
Evidence suggests that social cohesion increases the likelihood that residents will regularly take 
actions for the common good, particularly for preventing crime (Sampson, Raudenbush et al. 
1997; Sampson and Raudenbush 1999). The LAPOP AmericasBarometer survey shows that 
social trust is generally quite low in much of Latin America. However, there is considerable 
variation in social trust within countries, with some communities exhibiting far higher or lower 
trust than national averages alone would suggest (Córdova Guillén 2008). Therefore, the study 
measures trust as an outcome of the implemented programs. 
Impact Evaluation Report– Mexico 
37
Besides social cohesion, the impact study evaluated changes in the level of neighborhood 
involvement in crime-prevention activities or informal social control (see below). Other types of 
neighborhood activities oriented to improve well-being in general were also measured. 
Social Control 
Social control is defined in the literature as “the capacity of a group to regulate its members 
according to desired principles” (Sampson, Raudenbush et al. 1997). The literature differentiates 
between two types of social control: informal and formal. Informal social control refers to “the 
willingness of neighborhood residents to actively engage in behaviors aimed at preventing 
criminal and deviant behavior in the local area” (Miller and Silver 2004). Examples of informal 
social control are neighbors’ interventions aimed at discouraging young people from using 
and/or trafficking drugs and joining gangs. Formal social control refers to externally induced 
actions that seek to combat or prevent crime, such as police patrol. By strengthening community 
capacity to combat crimes and encourage community policing, we expected the programs to 
increase informal social control. 
Neighborhood Disorder
Neighborhood disorder is linked to both increased fear of crime and crime rates (Skogan 1990; 
Perkins, Wandersman et al. 1993; Kelling and Coles 1996; Perkins and Taylor 1996; Brown, 
Perkins et al. 2004). The well-known “broken windows” theory—developed by political scientist 
James Q. Wilson in 1982 and popularized by Police Commissioner Bratton in New York City 
which led to community police actions across the U.S.—states that both physical and social 
disorder lead to increased crime rates as these factors serve as cues to crime perpetrators, 
indicating that residents in the area do not care or are indifferent about what occurs in the 
neighborhood, and therefore result in provoking further and more serious crime. Thus, 
neighborhood disorder has been regarded as a risk factor associated with crime and violence. 
Physical disorder refers to “the deterioration of urban landscapes, for example, graffiti on 
buildings, abandoned cars, broken windows, and garbage in the streets,” and social disorder is 
understood as “behavior usually involving strangers and considered threatening, such as verbal 
harassment on the streets, open solicitation for prostitution, public intoxication, and rowdy 
groups of young males in public” (Sampson and Raudenbush 1999). 
Crime prevention programs are targeted to reduce neighborhood disorder and to mitigate 
physical risk factors through various infrastructure improvements, after school activities for at￾risk youth, and other services for the community as a whole which can reduce the signs of social 
disorder. Thus, reduced neighborhood disorder where the programs were implemented could be 
treated as an indicator of program success. 
Following the suggestions of recent literature, in addition to gathering survey data on residents’ 
perception of neighborhood disorder, data from a systematic observation of public places by 
trained observers was collected using a specialized neighborhood-level survey. Studies show a 
correlation between observed neighborhood disorder and crime in the U.S. (Perkins, 
Wandersman et al. 1993; Sampson and Raudenbush 1999). Consequently, in order to most 
effectively assess the impact of CARSI programs on neighborhood disorder, LAPOP 
Impact Evaluation Report– Mexico 
38
implemented a neighborhood-level survey using the Systematic Social Observation (SSO) 
method, as carried out in multiple studies on crime victimization and fear of crime in the U.S. 
(e.g., Baltimore, Chicago, and Salt Lake city studies; Sampson and Raudenbush, 1999). 
Perceptions of the Police 
By fostering strong ties between citizens and the police in neighborhoods, citizens’ trust in and 
satisfaction with the police were expected to increase in neighborhoods where USAID 
implements crime prevention programs. Thus, USAID interventions were expected to improve 
both perceptions of the police and perceptions of community policing. From AmericasBarometer 
LAPOP studies over the years in the region, we have a very good idea of the national and 
regional images of the police. Items on citizens’ perceptions of the police are included in the 
LAPOP AmericasBarometer questionnaire and in the questionnaire used in this impact 
evaluation. 
Other Outcome Indicators 
We have a variety of other outcome indicators, some of which have been measured in LAPOP 
AmericasBarometer surveys. Examples of additional outcome indicators include the following: 
 Social interactions and participation in local civic organizations 
 Interpersonal trust 
 Quality of life: life satisfaction and satisfaction with the place of residence 
 Intention to move to other neighborhood/country due to fear of crime 
 Trust in the local government 
 Participation in meetings convened by the local government or NGOs 
 Satisfaction with the services provided by the local government 
 Satisfaction and support for democracy 
The attached individual-level questionnaire (Appendix A) also included a series of “control 
variables” (variables that were not expected to change over time due to the interventions but 
were important to account for so that observed changes in outcome indicators could be 
differentiated as products of the USAID crime prevention interventions and not as a result of 
other intervening factors). Control variables included standard socio-economic and demographic 
characteristics (i.e., sex, age, employment status, income level) and other variables related to 
poverty. 
Section 3: Results of the Impact Evaluation 
Some of the main objectives of the USAID programs in the treated communities are the 
reduction of crime and the improvement of safety and security by strengthening the capabilities 
of the community and promoting citizen participation, as well as improving municipalities’ 
ability to prevent crime. Crime prevention activities include municipal crime prevention 
committees, community police, outreach centers for at-risk youth, vocational training, and safety 
and security infrastructure. 
Some of the hypotheses related to the direct effects of neighborhood security are: 
 Reduction in the perception of insecurity 
 Reduction in crime victimization 
 Decrease in the vulnerability of youth to participate in criminal activities and 
join gangs 
 More crime prevention measures 
 Better disposition to participate in collective neighborhood activities 
Some of the hypotheses regarding the indirect effects of the programs related to the promotion of 
good governance are: 
 Greater support for the police 
 Higher levels of interpersonal trust 
 Greater satisfaction with the functioning of democracy 
 Improved perception of the government’s safety/security efforts 
I. Evaluating the Programs’ Effectiveness 
The following sections show the results of the analysis of the survey data at the individual level 
as well as the qualitative and quantitative data at the community level. In this report, we do not 
provide the results of each variable in the data. Instead, we focus on all the variables that we 
believe will illuminate policy makers and those who implement the policies in the field.11
 
11 Given that the individual variables can produce statistically significant results simply by chance, we only report 
those results that are robust and consistent across multiple indicators. In spite of the fact that our models are fairly 
sophisticated, they are not capable of capturing the complexity of all of the contextual and individual factors that 
may impact the results. Some statistically significant results that seem unexplainable could be the result of a 
complex interrelationship between systems. We place more emphasis on the findings that are backed by the 
literature. 
Impact Evaluation Report– Mexico 
40
II. Analysis of Survey Data 
The clustered randomized experimental (CRE) design allows for the comparison between the 
control and treatment groups over time. A Difference-In-Difference (DID) estimator is a widely 
used econometric technique in impact evaluation that dates back thirty years ago to the work of 
Ashenfelter and Card (1985) and has become one of the “gold standards” in the field. It has the 
distinct advantage of controlling for what econometricians call “omitted variable bias,” since 
changes in the national economy, employment, and overall crime rates can impact the control 
and treatment groups, whereas in the case of a before-and-after study on the treatment group 
alone, it would be impossible to distinguish if such factors were responsible for changes in the 
treatment group rather than the treatment alone. It is, of course, not the only approach to 
measuring impact, but in the study we carried out, it was the most obvious and logical. For 
example, another widely used approach, that of regression discontinuity, works in situations in 
which detailed information about the population is already available (e.g., all have taken an 
entrance examination and some have scored just above the pass level and others have scored just 
below), but that was not the case in the neighborhoods we were studying. 
The DID estimator was utilized to estimate both the difference between the treatment and control 
groups and the difference between each group and itself at a later time. The treatment is 
considered effective at the point of evaluation if outcomes in the treatment group improve 
significantly more (or decline less) than they would have if the treatment had not been 
administered. One of the most important features of this model is that it does not require that the 
treatment and control groups have identical starting values, since what is of interest is the trend
in each group and the difference between these trends. We estimate a counterfactual case with 
the model parameters to simulate what the outcome would be if no treatment had been 
administered. The counterfactual case is the change in the control group over time but at the 
same levels of the treatment group. 
Thus, the difference between the estimated counterfactual case and the predicted level of the 
treatment group is considered the “treatment effect.” The “treatment effect” is the impact that 
can be said to have been caused by the intervention. Figure 10 illustrates the DID estimator. 
In Figure 10, the solid red and green lines represent the control and treatment groups, 
respectively. In this example, the levels of the two groups are different before starting the 
treatment. This kind of difference is common and, unless the number of selected communities 
have been very large (in the hundreds), it would have been unusual to expect that treatment and 
control had identical averages. Fortunately, the difference at the starting point is not a problem 
for the DID design. Differences at starting points are taken into account in the model by 
comparing each group with itself over time. The dotted green line represents the counterfactual 
case. The difference between the counterfactual case and the treatment group is the effect of the 
intervention. 
Impact Evaluation Report– Mexico 
41
Figure 10. Illustration of Difference-in-Difference Estimator 
Another advantage of the research design used by LAPOP in the impact evaluation is that the 
multilevel nature of the information obtained in the study makes it possible for the result to be 
evaluated individually and continue to take into account neighborhood characteristics (context) 
to determine the way in which the program effectiveness varies across different environments. 
Therefore, it is possible to identify the characteristics of neighborhoods or even countries (in 
cross-national studies) that contribute to the success of the program. Using multilevel modeling 
techniques, the study generates inferences about the program effectiveness and the impact of the 
context in which the programs were implemented. 
All statistical analyses in the project take into account “clustering” (the similarities shared by 
individuals living in the same areas such as neighborhoods within municipalities within 
countries, Hayes and Moulton (2009)). Standard errors and hypothesis tests are adjusted for this 
clustering, which is referred to as the “design effect.” Such studies must take into account that 
individuals belong to specific clusters (in this case, neighborhoods within municipalities in each 
country), and therefore corrected standard errors rather than the conventional t-test p-values need 
to be estimated and reported. Specifically, individual-level models include random intercepts for 
neighborhoods and municipalities. Models that include neighborhood or block-level contextual 
variables also include random intercepts for blocks within neighborhoods. 
In total, difference-in-difference model estimators were generated and analyzed for 51 variables / 
indices measured in surveys to individuals and systematic observations of the context of the 
Impact Evaluation Report– Mexico 
42
block. The results of these analyses are described further on, grouping the variables in seven 
thematic categories: crime victimization, fear of crime and perceptions of insecurity, 
vulnerability of youth to crime and gang activity, social cohesion, prevention measures, 
perception of the police, and the role of institutions. Within each category, the results of the 
models that were statistically significant are presented and discussed. As mentioned above, the 
models control for the following respondents’ characteristic: education, age, gender, household 
wealth, years of residence in their community, and the presence of youth at home. 
Additionally, within each category, we present some of the results about the impact of context on 
certain variables as well as the results of the qualitative research in order to complement the 
quantitative results. 
Crime Victimization 
In this section and the next, we present a long list of measurements designed to assess the 
effectiveness of USAID crime prevention programs in terms of (1) the reduction of crime 
victimization and (2) the increase in perception of security over the two periods of measurement 
for this impact evaluation conducted in the three municipalities in the north of the country. 
In this section we discuss four indicators: robberies in the community, overall crime 
victimization rate, damage to private property, and the sale of illegal drugs in the community, the 
latter being an indirect indicator of the incidence of crime. 
Robberies in the Community 
We observe that said probability decreases from one period to another, both in the treatment 
neighborhoods and in the control neighborhoods, but the decrease in the control neighborhoods is 
much greater than in the treatment ones. A little over a year into the implementation of the first 
programs in the treatment neighborhoods, we observe that the incidence of robberies reported in 
the community decreases six points, falling from 50% to 40%. This is a positive change and in 
the expected direction. However, the change in control neighborhoods is much more marked 
since the incidence of the occurrence of robberies is halved, going from 58% to 35% , i.e., a 
decline of 23 points which is almost four times higher than the observed decrease in treatment 
neighborhoods. 
The dotted green line in Figure 11 shows the counterfactual, which as explained above, presents 
the results expected for treatment neighborhoods assuming a similar trend to that observed in the 
control neighborhoods. As the decline was much more pronounced in the control neighborhoods, 
than in the treatment neighborhoods, we conclude for the treatment neighborhoods that although 
the trend of this indicator is considerable and in the desired direction, this is 63.5 % less than the 
expected minimum given the behavior observed in the communities where no programs were 
implemented. 
Impact Evaluation Report– Mexico 
43
It is noteworthy that only in some of the indicators analyzed (mainly in this one and in the 
overall rate of victimization) we observed a marked change in magnitude in the trends in the 
control group and large disproportionate differences between the treatment and control 
neighborhoods. In most of the models that we discuss, we observed less pronounced and less 
distinct changes. 
 
Figure 11. Respondent-Reported Robberies in the Neighborhoods 
Does this mean that the USAID crime prevention programs implemented in the treatment 
neighborhoods are not having the desired impact on this indicator of robberies in the 
community? This is a difficult question to answer. On the one hand, and based on the analyses of 
the quantitative evidence of this impact evaluation, under the difference-in-difference method, it 
appears that the impact is less than expected. To be precise, the estimate of the impact evaluation 
in the treatment neighborhoods is negative 63.5% [(26.9 - 44) /26.9]. Although theoretically 
speaking we could say that for this indicator, the treatment neighborhoods would be better off 
without the programs, i.e. in the absence of such programs, it would be expected to experience a 
decrease in the treatment neighborhoods in comparison to the control neighborhoods, it is 
difficult to conclude that programs lack impact. There are several reasons that make it possible 
even when lower than expected results are observed that the programs implemented have a 
favorable impact or that they have not yet been able to produce the desired impact. These reasons 
are discussed in the conclusions at the end of this report and are grouped into four types: design 
58.3
35.2
50.0
44.0
26.9
Treatment
Control
Counterfactual
20.0
30.0
40.0
50.0
60.0
Robberies in the Neighborhoods in the Past 12 months
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
44
reasons, specifics of the case of Mexico, time-related reasons, and the composition of the 
neighborhoods. 
Apart from the difference-in-difference models discussed herein, we also ran multi-level 
statistical analyses. They are multi-level, because they include the observations at the individual 
level of the questionnaire administered to the representative samples of adults in each 
neighborhood and systematic block-level observations of neighborhoods and municipalities. In 
the latter, the observations that we made about the blocks of the households where we conducted 
interviews with individuals constitute additional contextual information and provide an objective 
measurement of physical resources and of the social order/disorder in the neighborhoods. 
The analysis of the context in which the treatment was applied reveals that the effectiveness of 
the treatment varies according to the resources and conditions of the neighborhoods. 
Observations include resources such as schools, hospitals, and churches, as well as risk factors 
and other signs of social decay observed in every neighborhood. These block-level observations 
were added to the neighborhood level so that LAPOP could create a context variable for each of 
the neighborhoods. The context information is used as part of this “multi-level” statistical 
analysis to determine whether or not we can detect any impact of the context on the results. 
For each of the difference-in-difference models presented in the results section, we ran multi￾level statistical analyses. After each difference-in-difference model, we show the multilevel 
analyses that were statistically significant in the treatment group. These multi-level analyses 
control for the year of data gathering, age, gender, presence of young people in the household, 
education, income level, and type of neighborhood. 
Regarding the respondents’ reports on the incidence of robberies in the community we can see in 
Figure 12, that it is inversely related to the presence of schools on the block. The green line in 
the treatment group shows the predicted change in the reported incidence of robberies in the 
community in the presence of schools, an inverse relationship was observed because where there 
is the presence of schools, the incidence of robberies in the community decreases. The same 
behavior is observed in the control group (red line), but with a smaller slope. This finding is 
important and suggests that schools may be holding off or preventing criminal activity as far as 
robbery is concerned to some extent. 
Impact Evaluation Report– Mexico 
45
Figure 12. Marginal Change in Respondent-Reported Robberies in the Presence of Schools on the Block 
Total Crime Victimization 
This indicator is constructed using two items from the individual questionnaire: (1) whether the 
respondent was a victim of a crime such as robbery, assault, fraud, blackmail, extortion, threat, 
or some other type of crime over the course of the last year, and (2) whether any household 
member of the respondent was a victim. With these two questions, we build the overall rate of 
crime victimization which, unlike the previous indicator that is based on the occurrence of 
robberies in the community, is based on direct experience of victimization at the individual and 
household level. 
The results shown in Figure 13 show a similar reading to those discussed above: we observe a 
positive trend in the indicator seen since the overall rate of victimization decreases, and this trend 
is observed in the treatment group and the control group but the trend is much more marked in 
the second (3.8 and 14.2 points respectively). From the foregoing, we may interpret that the 
impact of the decline in the rate of victimization in the treatment neighborhoods is going in the 
right direction, but it is at least 88% lower than expected. Again, it would fit the interpretation 
that for this indicator, the treatment neighborhoods would be better off in the absence of the 
programs. 
-8 -6 -4 -2
0
Marginal Change
Reported Robberies in the
 Presence of Schools on the Block
Absent Present
School
Treated Non-Treated
Impact Evaluation Report– Mexico 
46
Figure 13. Total Crime Victimization 
Other indicators of crime victimization 
In addition to perception of robberies at the community level and the overall rate of crime 
victimization, we ran models for three other direct indicators of victimization: damage to private 
property, extortion, and kidnapping. For the first one, we get a similar result to the one reported 
in the case of robberies in the community and the overall rate of victimization: improvements in 
the trends in both the treatment and control groups, but more marked in the control group which 
implies that the result in the treatment neighborhoods is again lower than expected. In the case of 
extortion and kidnappings, we do not report any significant differences and their incidence is 
very low in both the control and treatment groups. 
Regarding the incidence of damage to private property in the last year, a positive trend, i.e. a 
decrease in both the control and the treatment groups is obtained. For the latter, decrease is 5.4 
points down from 23% at baseline to 17.6% in the post- treatment. 
Sales of Illegal Drugs 
Finally in this section of crime victimization, we present the results for the question on the sale 
of illegal drugs in the community which is seen as an indirect indicator of the incidence of crime. 
32.4
18.2
25.9
22.1
11.7
Treatment
Control
Counterfactual
10.0
15.0
20.0
25.0
30.0
35.0
Crime Victimization (%)
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
47
In Figure 14, we observe the results in the treatment neighborhoods, whose trend runs contrary to 
expectations (it increases by 11 percentage points, going from 23.4% to 34.4%) and a slight 
improvement (5.4 points) in the control neighborhoods. This is a less favorable result than the 
previous ones as the indicator moves in the opposite direction than anticipated; that is to say, we 
do not observe an improvement that is less than expected but instead, there is a marked 
deterioration in the treatment group regarding the percentage of individuals reporting sales of 
illegal drugs in their community. In terms of the model, we could say that in the treatment 
neighborhoods the observed outcome is of minus 95.4% compared to the expected results if a 
similar behavior in the control neighborhoods had been presented. 
Figure 14. Sale of Illegal Drugs in the Community in the Past 12 Months 
In spite of the fact that the research design is at the treatment group level, we observe some 
interesting differences in the treatment group at the city level – mainly that there seems to be an 
increased drop or smaller improvement in the Tijuana neighborhoods than in the other two cities. 
In the multi-level models we can see that the sale of illegal drugs over the course of the last year 
is inversely related to the presence of Catholic Churches and hospitals. In Figure 15, the 
treatment group’s green line shows the predicted change in the reported incidence of drug sales 
in the community in the presence of Catholic Churches and we can see that they are inversely 
related because where there is a presence of this type of Church, the reported incidence of illegal 
30.2
23.1 24.8
34.4
17.6
Treatment
Control
Counterfactual
15.0
20.0
25.0
30.0
35.0
Sale of Illegal Drugs in the Community
 in the Past 12 Months
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
48
drug sales decreases. The opposite behavior, but weaker, is observed in the control group (red 
line). 
Regarding the presence of hospitals, we can observe a similar relationship for both groups but 
more pronounced in the treatment group (see Figure 16). In both groups, the predicted change in 
the reported incidence of drug sales in the community decreases as the presence of hospitals in 
the block increases. Both institutions seem to be playing an important role as far as drug use is 
concerned, and according to the data, the hospitals’ role is more effective in creating a more 
pronounced marginal change and affecting both the treatment and control groups. 
Figure 15.Marginal Change in Respondent-Reported Sale of Illegal Drugs in the Presence of a Catholic 
Church on the Block 
 
-10 -5
0
5
Marginal Change
Sale of Illegal Drugs in the Community 
in the last 12 months
Absent Present
Catholic Church
Treated Non-Treated
Impact Evaluation Report– Mexico 
49
Figure 16. Marginal Change in Respondent-Reported Sale of Illegal Drugs in the Presence of a Hospital on 
the Block 
Below, we present some findings from the qualitative research that complements the results from 
the quantitative indicators in this section. Searching for clues in the qualitative research about the 
improvement of the indicators of victimization that help us to understand the trends is not an 
easy task given that for the vast majority of participants in the focus groups, it is easier to 
remember criminal acts in their community and judging by the tone of their comments, the 
events seem to occur daily and hit close to home: 
“ahí en el barrio donde yo vivía pasó, gente rateros, ahí siempre había por decir en cualquier 
barrio que fulanito es ratero y al tiro con él” 
“Here in the neighborhood where I used to live it happened- pickpockets and thieves- people 
used to say in any neighborhood that whoever is a thief, to hell with him” 
“No cambia nada es lo mismo robos y robos, pos ahorita en las escuelas están poniendo muchos 
climas y todo eso y pues se lo roban, ahorita es lo mismo” 
"Nothing ever changes. It’s always the same. Robbery after robbery… Right now schools are 
putting a lot of AC units and all that, and even still, they rob them…. Right now it is the same.” 
“en lo personal a mí me robaron una camioneta...afuera de mi casa con armas largas, eran seis 
muchachos jóvenes de 15 y 17 años” -15 -10 -5
0
Marginal Change
Sale of Illegal Drugs in the Community
 in the past 12 months
Absent Present
Hospital
Treated Non-Treated
Impact Evaluation Report– Mexico 
50
“Personally, they stole my van… right outside my house... with rifles; they were six young 
boys…15 and 17 years old.” 
“Robo o sea a mí me tocó, va uno con sus bolsas, va uno y compra ropa, tenis, viene con sus 
bolsas y el arrebatón y a correr” 
“Robbery, that happened to me, you go down the street with your bags, you go and buy clothes, 
sneakers, you go with your bags and they snatch your stuff and take off” 
“A las diez de la mañana fue el robo y yo creo que echan algo para que la gente se quede 
dormida porque vilmente están las personas ahí dormidas y no sienten que alguien entró y le 
pasó por un lado y se robó las cosas y no nada más fue en la casa de al lado, fueron también 
otras tres casas de mi cuadra” 
”At ten in the morning, we got robbed and I think that they spread something around so that 
people stay asleep because the people were sound asleep and couldn’t tell that someone had 
come in and gone around the whole house stealing stuff…. They also went into the house next 
door and to three other houses on my block.” 
In environments so full of victimization as suggested by these comments, it may be encouraging 
that there are improvements in the indicators trends, even if they are uneven between 
neighborhoods and lower than expected in the treatment groups. 
The sale of drugs could turn out to be a phenomenon of greater frequency and proximity to 
citizens according to some of the comments from the focus group participants. Additionally, they 
suggest a worsening of the situation and the additional concern of drug sales directly affecting 
youth and children. The following comments illustrate this: 
“Es que por decir en la esquina de mi casa pues la casa que está ahí tiene un árbol muy grande, 
entonces si cubre todo lo que viene siendo la banqueta y ahí se juntan, ahí como para que no se 
vean, ahorita ya son más descarados, ahorita ya en plena luz, en el pleno día ya están ahí 
vendiendo y que triste porque ves que van puros huerquitos, ya no es como que ya señores a 
comprar la droga no, ya son puros niños de 16, 15 años…” 
If I had to say, on the corner of my house, the house that’s right over there has a really big tree, 
then if it covers all the bench and that’s where they come together, there so that you can’t be 
seen. Now they have no shame, they go out in the light, in the middle of the day and they are 
selling and it’s so sad because you see that they are just little kids, it’s not even like old people 
buy drugs, no, they’re all kids… 15 and 16 years old…. 
“Es que anteriormente ahí en la Col. Independencia era donde se concentraban mas porque ahí 
ya no hay caminos, nada más son puras escaleras y los militares bueno ni uno sube ahí, ni los 
que viven ahí tienen miedo de subir y pos bueno como ya todo está igual…” 
Back in the day in Colonia Independencia (Independence Neighborhood) this was the place 
where there were a ton of them because there, there are no roads… there are just stairs and the 
Impact Evaluation Report– Mexico 
51
soldiers do not come up there… even the people who live there are afraid to go up the stairs…. 
Anyways… everything stays the same. 
“Yo creo que ya en todas las colonias está igual, antes no verdad?, pero ahorita yo creo que ya 
en todos está igual, en todos” 
I think that in all the colonias (neighborhoods) it’s the same but before not, right? But now I think 
that everything’s the same everywhere. 
“los pandilleros y los que venden drogas se acercan mucho a las escuelas, obligan a los niños a 
vender droga si no los golpean o igual les dicen que les van a hacer algo a sus familias” 
The gang members and drug dealers go right up to the schools and they force kids to sell drugs 
and if they don’t they beat them up or they tell them that they are going to do something to their 
families. 
“ahora con las bandas organizadas se metieron ya con las pandillas, ahora entraron a las 
colonias de las periferias y agarraron a los pandilleros para que vendieran inclusive lo que en 
mi colonia pasó que había 13, había 18 y con la venta de droga se unieron, dejaron de pelearse” 
Now the organized bands are getting together with the gangs… they entered into the colonias 
(neighborhoods) from the outlands and scooped up the gang bangers so that they sell together… 
what happened in my neighborhood was that there were 13, there were18…. And they all got 
together to sell drugs… they stopped fighting. 
Fear of Crime and Perception of Insecurity 
In this section, we analyze five indicators in which the difference between the treatment and the 
counterfactual turned out to be statistically significant: fear of crime in the community, worry 
that someone would come into your home to commit robbery, perceived safety of living in your 
municipality and state, and the necessity to change neighborhoods due to insecurity. 
Additionally, we ran other four models for four indicators that are not discussed in this report for 
not complying with the requirement of being statistically significant. The indicators of this topic 
that did not turn out to be statistically significant are: perception of insecurity in the community, 
concern that one would be threatened or attacked on the street, perceived threat from crime and 
violence, and perception that there is a higher level violence in one’s community than in others. 
Fear of Crime in the Neighborhood 
Using five items of the questionnaire applied to individuals, we build a fear of crime index 
whose scale goes from 0-100 that captures the cognitive dimension of individuals’ fear. These 
items are: how safe / unsafe one feels being at home with his family during the day, when alone 
Impact Evaluation Report– Mexico 
52
at home during the day, when alone at home at night, walking alone through the neighborhood 
during the day, and walking alone at night. 
The difference-in-difference model’s results are shown in Figure 17. We observe that there is a 
very slight decrease in the index in the treatment group that goes from a score of 43 in the 
baseline to 42 in the second round of the study. Regarding the control group, we see a change in 
the same direction but of greater magnitude because it decreases by almost four points, going 
from 42.6 to 38.7 (on a 0-100 scale). As some figures in the previous section show, the results 
are that there was a change in the right direction, but it was lower than expected as shown by the 
counterfactual line and where we see that given the beginning of the indicator in the treatment 
group and expecting a behavior at least equal to the control group, it was expected that in the 
second round the indicator in the treatment group would obtain a score of at least 39.1; that is to 
say 7.4% or more below what we observed in the second round. 
Figure 17. Fear of Crime in the Neighborhood 
In the multi-level models, we observe that fear of crime is inversely related to the presence of 
hospitals. In Figure 18, the green line representing the treatment group shows the predicted 
change in the reported incidence of fear of crime in the community in the presence of hospitals. 
We can see that there is an inverse relationship between them since where there are hospitals, 
there is a smaller presence of fear of crime in the community. The control group shows almost no 
change in the presence or absence of hospitals (red line). It is likely that the presence of hospitals 
42.6
38.7
43.0
42.0
39.1
Treatment
Counterfactual
Control 39.0
40.0
41.0
42.0
43.0
Fear of Crime
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
53
is indicative of far more favorable conditions than in the treatment groups’ context because it has 
an impact on perceptions of fear of crime in the community. 
Figure 18. Marginal Change in Fear of Crime in the Presence of Hospitals on the Block 
 
Concern with Household Burglary 
Regarding the concern that someone will break into your home to steal, we observed similar 
results to the previous indicator: i.e. slight changes in both groups in the right direction but less 
change than expected in the treatment group. It is worth noting that the change in the control 
group is even more pronounced when these neighborhoods reported a larger baseline (see Figure 
19). The change in the treatment group is 3.5 points - falling to 69.0 from 72.5 (0-100 scale), 
which is still a relatively high score of perceived insecurity as measured by this indicator. The 
control group decreased by 9.1 points and based on this trend, we would have expected a change 
in the control neighborhoods of at least 8.8% higher than of what we observed. 
This behavior, which is apparently paradoxical, has three characteristics: (1) in the baseline, the 
result in the control groups is worse than in the treatment group; (2) there is a greater magnitude 
of improvement in the trend in the control group than in the treatment group; and (3) the post￾treatment observation in the control group shows better results than in the treatment group. This -8 -6 -4 -2
0
Marginal Change
Fear of Crime in the Community 
 (Cognitive Dimension)
Absent Present
Hospital
Treated Non-Treated
Impact Evaluation Report– Mexico 
54
pattern is repeated in many of the models outlined in this report and suggests the possibility that 
the elasticity or sensitivity rate between groups is different, the control group being more 
sensitive / elastic than the treatment group. 
Figure 19. Concern that the Home will be Robbed 
In the multi-level models, we get the same results as those observed above with respect to the 
incidence of drug sales. Figures 20 and 21 show that the concern that your home will be robbed 
is inversely related to the presence of Catholic Churches and hospitals. In the first figure, we 
observe that for the treatment group, the predicted change in the concern that one’s home will be 
robbed decreases in the presence of Catholic Churches. We observe a similar but almost 
imperceptible change in the control group (red line) behavior. 
Regarding the presence of hospitals, we observe a similar relationship in both groups, but again 
the result is more pronounced in the treatment group (see Figure 21). In both groups, the 
predicted change in the concerns that one’s home will be robbed decreases in the presence of 
hospitals on the block. Again, for this indicator, both institutions appear to play a prominent role. 
73.9
64.8
72.5
69.0
63.4
Treatment
Control
Counterfactual 64.0
66.0
68.0
70.0
72.0
74.0
Concern that the Home will be Robbed
(Predictors)
Pre-treatment Post-treatment
Time
*Resultados significativamente diferentes de 0 en p menor .05
Fuente: Estudio CARSI por LAPOP
Impact Evaluation Report– Mexico 
55
Figure 20. Marginal Change in the Concern that One’s Home will be Robbed in the Presence of Catholic 
Churches on the Block 
Figure 21. Marginal Change in Concern that One’s Home will be Robbed in the Presence of Hospitals on the 
Block 
-8 -6 -4 -2
0
Marginal Change
Concern that One's Home will be Robbed
Absent Present
Catholic Church
Treated Non-Treated
-6 -4 -2
0
Marginal Change
Concern that One's Home will be Robbed
Absent Present
Hospital
Treated Non-Treated
Impact Evaluation Report– Mexico 
56
Perception of Insecurity in Living in the Municipality/State 
These two indicators show very similar results and therefore we only present data on the 
perception of the security of living in the state of residence. In Figure 22, we observe that the 
treatment group shows a significant decrease in the expected direction of those who consider that 
living in their state is unsafe as shown by the solid green line. This indicator fell by just over 8 
points from 53.3 to 45.1 (on a 0-100 scale). The change at the municipal level is similar, but 
slightly smaller, as this indicator decreased 7.5 points. These two positive developments for the 
treatment group are of greater magnitude in this impact evaluation. But once again, for the 
treatment group, results are lower than expected due to the slope of the change in the control 
group. According to the model, the change in the treatment group should have decreased by at 
least 13% more than we observed. 
Figure 22. Perception of Insecurity of Living in the State 
In the multi-level models, we see that in both cases, i.e. the perception of insecurity of living in 
the state or the municipality, that there is a decrease of insecurity in the presence of a church on 
the block. As shown in Figure 23, again we observe that the change is significant for the 
treatment group (green line) but practically nonexistent for the control group (red line). 
56.8
43.5
53.3
45.1
39.9
Treatment
Control
Counterfactual 40.0
45.0
50.0
55.0
60.0
Insecurity of Living in the State
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
57
Figure 23. Marginal Change in the Perception of Insecurity of Living in the State in the Presence of Churches 
on the Block 
 -4 -3 -2 -1
0
Marginal Change
Perception of Insecurity of Living in the State
Absent Present
Chuch
Treated Non-Treated
Impact Evaluation Report– Mexico 
58
Need to Change Neighborhoods Due to Insecurity 
The last indicator of this section presents different behavior than others given that it worsens 
slightly in the treatment group, rather than improves. Additionally, we observe that there is an 
improvement for the indicator in the control group which highlights the negative result given that 
the implementation of the programs in the treatment groups would create the expectation of 
behavior equal to or better than what we observed in the control group. 
Figure 24 illustrates this behavior and shows that while the solid green line (treatment group) 
shows a slight increase of 1.5 points in the negative direction, the red line moves in greater 
magnitude and moves in the opposite direction. The counterfactual line (dotted green) represents 
where we would expect the treatment group to be given the tendency of the control group. 
 
In terms of the model, we observe in the treatment neighborhoods, a behavior of minus 86% than 
would be expected if those groups would have followed a similar behavior than that of the 
control group. 
Figure 24. Need to Change Neighborhoods Due to Insecurity 
As in the previous section, here we present some of the qualitative research’s findings that 
complement the quantitative indicators shown in this section. Several things stand out. First, it is 
worth noting that we perceive differences by city which suggests different conditions or 
situations with respect to crime and violence in both the baseline and post-treatment survey. 
17.3
11.3
14.5
16.0
8.6
Treatment
Control
Counterfactual
8.0
10.0
12.0
14.0
16.0
18.0
Need to Change Neighborhoods Due to Insecurity
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
59
Second, some participants expressed a perception of improvement in the situation which seems 
to be multifactorial and different by municipality. And thirdly, the data show that individuals are 
either starting to feel that insecurity is the normal state of affairs and/or they are growing 
extremely tired of it. Here are some comments that illustrate these observations. The first ones 
come from youth in Monterrey showing perceptions of recent improvement: 
“bares que se llenaba completamente, empezó la inseguridad y bajó el nivel de las personas que 
iban y ahora ya se está llenando otra vez, ya nada más es una cuadra pero son como cuatro 
bares y ya la gente está empezando a salir otra vez, ya están agarrando confianza para ir a 
Barrio Antiguo porque hace como dos años era como que no, no quiero ir a Barrio Antiguo" 
“Bars would fill up completely… insecurity started and then the number of people that went 
dropped and now they’re starting to fill up again…it’s just a block but there are like four bars 
and the people are starting to go out again… they are gaining confidence to go out the Barrio 
Antiguo (Old Neighborhood) because like two years ago, it was like “No I don’t want to go the 
Barrio Antiguo.” 
“Yo creo que en principio la gente tenía mucho temor y a final de cuentas al paso del tiempo ha 
tenido repugnancia ante esos grupos también, a final de cuentas ha cambiado el miedo a ser un 
poquito más como que aislada la gente" 
“I think that in the beginning people were really afraid and at the end of the day with the passage 
of time, they started to hate those groups, also when you add it all up, now fear make people 
isolate themselves.” 
“No sé si es porque lo oculten o porque en realidad sea pero yo ya no he escuchado tantas 
matanzas como antes, antes era día tras día, día tras día y a lo mejor ahorita también pero ya no 
lo hacen tan público” 
“I don’t know if it’s because they hide it or because in reality it’s because I haven’t heard of 
many killings like before… before it was day in, day out but right now it might be that they don’t 
make it so public.” 
“Yo creo que la gente tiene un poquito más de confianza de los tiempos que se vivieron de hace 
tres o cuatro años para atrás, ya tiene un poquito más de confianza a lo mejor no tanto en la 
policía de creer en ellos pero a lo mejor ahorita ya salen, antes no salían o muchas, muchas 
veces era nada más por simples rumores pero pues uno lo que no quería era arriesgarse a que 
fueran de verdad las cosas" 
“I think that people have a little more trust than they did 3 or 4 years ago… now they have a little 
more trust not really in the police.. they don’t believe in the police… but now they go out, they did 
not go out before and many, many times it was only because of simple rumors but you weren’t 
going to risk that the rumors were true.” 
The perception of change and optimistic and positive attitude is not exclusive to the young 
people of Monterrey. The following comments from two different municipalities relate this 
change from the perspective of leaders / officials and reveal a positive attitude: 
Impact Evaluation Report– Mexico 
60
“Pues yo lo que trato de brindar como principal objetivo es traer gente, que venga confiada para 
que se hospede en mi hotel ¿si?, la gente entra con tanta inseguridad o vienen tanto del 
extranjero o del interior de la república en donde hay la misma situación pero creen que todavía 
estamos en tiempos de hace dos años, matanzas, salir, que te agarren y te den un balazo o 
robarte o extorsionarte o secuestrarte, cosa que ya no está sucediendo…Y para solucionar eso 
pues tienes que poner de tu parte para decir ¿sabes qué?, Cd. Juárez está renaciendo, está 
volviendo a tener economía, vuelve a haber movimiento de dinero para que exista más fluidez de 
turismo.” 
“Well, what I try to offer as the main objective is to bring in the people – that they come with 
confidence to be a guest in my hotel, right? People show up with so much insecurity or they come 
from abroad or from the middle of country where the same situation is going on, but they think 
that we are still in the middle of those bad times from two years ago – killings, going out, they 
snatch you up and put a bullet in you or rob you or blackmail you or kidnap you – things that are 
not happening anymore… And to solve all this, you have to do your part to say ‘You know what? 
Ciudad Juárez is going through a renaissance, the economy is growing, money is moving so there 
is more dynamic tourism.” 
“Tijuana es una ciudad que tiene mucho potencial y el potencial nosotros lo debemos explotar 
los jóvenes, entonces por ejemplo ahorita en el centro ya hay negocios nuevos que anteriormente 
estaban en picada" 
“Tijuana is a city that has a lot of potential and we young people have to exploit that potential so, 
for instance, right now in the downtown there are new businesses that were going downhill 
before.” 
In the city of Monterrey, part of the recent change is largely attributed to the armed forces’ 
presence of security as noted by one of the participants which could mean that the containment 
of more severe crimes such as kidnapping and extortion may cause an increase in the rates of 
other crimes such as home break-ins: 
“Pos metieron a los soldados y fue como se fue calmando, desde que entraron los soldados fue 
que se fue calmando, se fueron calmando y hasta ahorita están calmados, pero cuando no 
estaban los soldados ni nada y que la misma policía era un desacato todo, muertos por aquí y 
muertos por allá, por donde sea.” 
“Well, they rolled out the soldiers and all of the sudden, things calmed down since soldiers 
showed up, things have calmed down and up to now things are calm but when the soldiers 
weren’t around or anything and when everyone was disrespecting the cops, there were dead 
people here, dead people there… wherever….” 
“Hay Secuestros, extorsiones… Bueno el año pasado a mediados de otoño hubo muchas redadas 
de Marina, de Militares etc., y ahora como ya vieron que muchos o fallecen porque los balean los 
militares y así, ya ven como que más peligro, pero lo que hacen mucho últimamente es el robo a 
Impact Evaluation Report– Mexico 
61
casa habitación, bueno por mi colonia es lo que más se sabe… Últimamente lo están haciendo 
mucho, en mi colonia también la casa de al lado la robaron a las diez de la mañana." 
“There are kidnappings, blackmail…. Well last year in the middle of fall, there were a ton of 
Marine and Military raids, etc… and now as they already saw that many people died because the 
soldiers shot them and so they see that things are more dangerous but they’ve been doing a ton of 
robberies of houses… well in my neighborhood anyway this is what we know… Lately they’ve 
been doing it a lot.. in my neighborhood too… they robbed the house next door at ten in the 
morning.” 
Another interesting point is the recognition (and suspicion) that the media now report far fewer 
incidents, and this has to do with the improvement in public perception of the security situation: 
“No sé si es porque lo oculten o porque en realidad sea pero yo ya no he escuchado tantas 
matanzas como antes, antes era día tras día, día tras día y a lo mejor ahorita también pero ya no 
lo hacen tan público.” 
“I don’t know if it’s because they hide it or because in reality it’s because I haven’t heard of 
many killings like before… before it was day in, day out but right now it might be that they don’t 
make it so public.” 
“Y la diferencia en las noticias verdad?, que por decir 20 muertos el día de hoy, tantos en tal 
parte, tantos en tal parte hay una diferencia que a lo mejor ahora no, las noticias no reportan ni 
un muerto." 
“The difference in the news, right?... saying 20 deaths today, so many wherever, so many 
wherever, there’s a difference that now maybe not… the news doesn’t even report a single 
death.” 
However it is important to underscore that while there are signs of improvement, the dominant 
perception is that this is still a complicated situation that has consequences for families and the 
economic activity of these cities. The following comments illustrate this situation: 
“Pues miedo a que te maten, a tu familia, que te maten y ya la familia se queda sola. No queda 
de otra más que cambiarse. Es que también mucha [gente] en su casa ya tengan miedo y por eso 
a veces se ponen así que ven algo y cierran candados y todo y se ponen hasta a llorar, se ponen 
mal...” 
“Well, fear that they’ll kill you, your family and they’ll kill you and your family will end up all 
alone. There’s nothing left to do but move. Also, a lot of people are afraid in their homes and 
that’s why they get like this and they see something and bolt their doors and all and they get all 
worked up and cry… they feel awful…” 
“Las personas que quieren iniciar negocios propios ya no lo quieren hacer porque cuando están 
iniciando su negocio llegan las personas que están metidas en eso y piden piso (extorsión), 
entonces ya no les sale el negocio o sea tienen miedo a emprender algo…” 
“People that want to start their own businesses don’t want to do that anymore because when they 
start up their business, people who are involved in that ask for a kick-back (blackmail), so their 
business doesn’t work out and they feel really afraid to start something up….” 
Impact Evaluation Report– Mexico 
62
“… los jóvenes lo ven y piensan que a todos igual no saben a quién sí y quién no, entonces les 
tienen miedo, les tienen temor, los miran y ya van caminando y se van a un lado y ni siquiera los 
voltean a ver por miedo a que les vayan a hacer algo, prefieren cambiarse a otra zona porque 
viven siempre con temor…” 
“…the young people see that and think that everyone is the same and the don’t know who’s who, 
so they get scared… they are afraid of then and they see them walking down the Street and get 
out of the way and they don’t even turn around out of fear that they are going to do something to 
them, they prefer to move to another area because they are always afraid….” 
Vulnerability of Young People in the Face of Crime and Gang Activity 
In this section, we analyze four indicators in which the difference between the treatment and the 
counterfactual is statistically significant. The first two have to do with knowledge of the 
implementation of recreational activities for young people or job training and the second two 
have to do with the consumption of drugs or alcohol by young people. 
Additionally, we ran models for six other indicators in this category that we do not present here 
because they did not show significant results. Those were: perception that there are young people 
in the community doing nothing; the perception that youth gang involvement is a problem and 
that there gang fights in their neighborhood, reported incidence of gang presence, perception that 
youth joining gangs is an option, and whether or not people are aware of a young person who has 
participated in youth activities in the last year. 
Support Activities Directed at Young People 
We evaluate the following question: “Have you seen or heard that an institution has provided 
support activities aimed at the young people of this neighborhood in the past 12 months?”; and in 
particular, regarding activities also aimed at young people to better their use of free time. In 
Figures 25 and 26, we show the models of these two assessments whose results are very similar. 
In both figures, we observe that while the results in the treatment neighborhoods have increased 
by 3.2 percentage points, in the control groups, the percentage of adults who have knowledge 
about actions taken to support young people has decreased. The green dashed line shows the 
expected behavior for the treatment group in the absence of the programs (assuming a similar 
control group behavior). Given that the result of the observation in the treatment group (solid 
green line) is significantly higher than the counterfactual (dotted green line), it turns out that the 
treatment effect is the distance between the two green lines; the results of Figure 25 show that the 
effect of the treatment for this indicator is 120.9% better than expected without the 
implementation of the programs. 
Impact Evaluation Report– Mexico 
63
Figure 25. Knows of an Institution that has Provided Support Activities for Young People 
Regarding the evaluation of the knowledge of institutions that have taken action to support young 
people and to help them better use their free time, we observe the same behavior as seen in the 
model above: an increase in the treatment group and a decrease in the control group. The increase 
in the treatment group compared to the baseline is 2.4 points (going from 11% to 13.4%) so that 
the treatment effect is 59% better than expected without implementing programs. 
11.9
7.6
10.5
13.7
6.2
Treatment
Control
Counterfactual
6.0
8.0
10.0
12.0
14.0
Knows of an Institution that has Provided
 Support Activities for Young People
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
64
Figure 26. Knows of an Institution that has Taken Steps to Help Young People Use their Free Time Better 
In the multi-level models, we see that in the treatment group, the knowledge of institutions that 
have provided activities for youth (leisure or other support measures) is inversely associated with 
the observed social disorder. This means that as there is a decrease in the knowledge of these 
institutions the observed social disorder increases (see Figure 27). The behavior in the control 
group (red line) is the opposite, that is to say that as social disorder increases, we observed a 
greater awareness of institutions carrying out activities aimed at young people. We only present 
the graph with the results for the question recreational activities for young people but we obtain 
the same result when looking at the question about other kinds of institutions that provide other 
activities aimed at young people. 
9.6
7.1
11.0
13.4
8.4
Treatment
Counterfactual
Control
6.0
8.0
10.0
12.0
14.0
Knows of an Institution that has Taken Steps to Help
 Young People Use their Free Time Better 
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
65
Figure 27. Marginal Change in the Knowledge of Institutions Working for Young People to Use their Free 
Time in the Presence of Social Disorder on the Block 
Youth Consumption of Alcohol and Drugs 
We also assessed whether the respondent knew or had heard of a young person in the 
neighborhood who has used drugs or alcohol within the past 12 months. While the changes 
between the baseline and post-treatment rounds are slight and inconsistent for the treatment 
group, in the control group, there are no significant decreases over time for both indicators. 
In relation to alcohol use in the treatment group, we show in Figure 28 that there is a slight 
increase in the indicator from 40.5% to 41.2%. With regard to drug use (Figure 29), in the same 
group, a small decrease of 33.5% to 33.6% was recorded. Both alcohol and drug use decreases 
significantly in the control group. The reported alcohol consumption goes from 42.2% to 37.7% 
and 37.7% to 30.4% for drugs. As in some previous models, and as the change in the control 
group was higher, the results obtained in the treatment group for these two indicators is worse 
than expected: 16.4% for alcohol consumption and 19.1% for consumption drugs. -4 -2
0
2
4
Marginal Change
Institution Working for Young
 People to Use their Free Time Better
0 1 2 3 4
Social Disorder
Treated Non-Treated
Impact Evaluation Report– Mexico 
66
 Figure 28. Knowledge of a Young Person Who has Consumed Alcohol in Excess in the Past Year 
42.2
37.1
40.5
41.2
35.4
Treatment
Control
36.0 Counterfactual
38.0
40.0
42.0
Knowledge of a Young Person
 Who has Consumed Alcohol in Excess
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
67
Figure 29. Knowledge of a Young Person who has Consumed Drugs in the Past Year 
Regarding the impact of context on the knowledge of young people using alcohol or drugs, we 
observe that both items are inversely related to the presence of hospitals. Although the results are 
similar, we only present the results of this interaction in the case of drug use (Figure 30) and not 
for excessive alcohol consumption. For both groups, the control and treatment, we predict a 
decline in the knowledge of drug use in the presence of hospitals. Again this variable context, i.e. 
the presence of a hospital, has a favorable impact on relevant indicators of this impact 
evaluation. The individual questionnaire items do not provide many clues to help us understand 
why the presence of hospitals is still an important factor. It is likely that the presence of a 
hospital means on the one hand, more robust conditions in terms of community resources and, on 
the other hand, hospitals are important and nearby institutions that provide support, especially in 
areas of public health and drugs and alcohol abuse. 
37.7
30.4
35.5
33.6
28.2
Treatment
Control
Counterfactual
28.0
30.0
32.0
34.0
36.0
38.0
Knowledge of a Young Person
 who has Consumed Drugs in the Past Year
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
68
Figure 30. Marginal Change in the Knowledge of Young People who Consume Drugs in the Presence of 
Hospitals on the Block 
To conclude this section, we present the qualitative research findings that complement the 
discussion of quantitative indicators. Young people are a fundamental part of the country's 
population, so it is strategic to support them through educational, cultural, and/or recreational 
programs that reduce their vulnerability to criminal networks. Different public and private actors 
have invested in social programs in at-risk areas that focus on young people given the insecurity 
that exists in the three cities in this study. However, a significant number of people do not know 
about these programs and the people who know about them see them as insufficient to achieve 
their stated goals: 
“Pues igual le digo que como hace rato le digo que si existen los programas pero no son lo 
suficientemente grandes para satisfacer la capacidad de demanda de la ciudad.” 
“Well, I tell you that just a little while back there were programs, but now they’re not big enough 
to satisfy the high demand in the city.” 
From the perspective of young people, the main problems they face are: drug addiction, family 
breakdown, lack of educational opportunities, and unemployment. These problems have several 
consequences, such as, the ease with which young people get involved in gangs which leads to 
getting involved in vandalism which in turn creates a vicious cycle of theft, robbery, mob 
violence, and eventually criminal organizations step in and introduce them to guns and drugs. -8 -6 -4 -2 0 Marginal Change Knowledge of Young People who Consume Drugs
Absent Present
Hospital
Treated Non-Treated
Impact Evaluation Report– Mexico 
69
The perception is that the vulnerability of young people to crime and gangs has been increasing 
steadily recently. The government has created programs, along with some civic and religious 
organizations. They provide young people with entertainment, sports and recreational activities 
so that they are not out on the streets all day and thereby these activities reduce their 
vulnerability. However, some feel that the problem lies in early childhood, where children do not 
have a family emotional stability, and this increases when economic problems are greater. 
The most vulnerable young people are those who live on the street, those who do not belong to a 
family, those who not taken care of or are abandoned by their parents, and those who receive no 
guidance: 
“Muchos padres no tienen comunicación con los hijos y ahí está el problema, lo que falta es 
comunicación con los padres, para que la familia esté bien comunicada sobre lo que está 
pasando. No piensan en lo que van a decir los jóvenes, los hijos, piensan en sí mismos más que 
nada y por eso hay muchos divorcios.” 
“Many parents don’t communicate with their children and that’s where the problem is. What’s 
missing is communication with the parents so that family members know what is going on. They 
don’t think about what their kids are going to say – the children – they think of themselves above 
anything else and that’s why there are a lot of divorces.” 
Schools are a common place where young people gather to study but is also where groups with 
certain preferences are created. To a great extent, gangs originate in schools, and they move to 
the street and finally end up joining criminal and drug trafficking organizations. One of the 
reasons why criminal organizations have very young children in their ranks (even younger than 
15 years) is because it works better. Mexican laws do not prosecute minors severely, so a boy 
who has been subjected to any criminal proceedings, may only spend two years in prison and 
later return to society. The easy money and the fact that they do not place a high value on their 
own lives as well as not having positive expectations in life, cause young people from a very 
young age to decide to join gangs / crime. This solves three of the main perceived missing 
elements to their lives: (a) belonging, i.e. feeling like they are part of a group or family; (B) 
having financial security; and (c) feeling valued. 
“Porque son chavos que saben que nomás están en la calle… No, y aparte porque ellos saben 
que si los agarran…Salen, Póngale tiene 16, salen a los 18… Más pronto que los mayores de 
edad. No los pueden procesar… Y aparte hay otra cosa… A los 18 años al tribunal y no le dan 
más pena que a los mayores de edad… Y es que como tienen también la mente fresca y ellos los 
hacen como quieren. Pues eso sucedió hace como cuatro años, aquí en Padres, que estaban 
agarrando a puro chavalito y los metían a la correccional y eran como de 13 o 14.” 
“Because their just little guys, that they know they’re just hanging out in the Street… No, and 
besides this they know that if they get caught… they get out… Let’s say they’re 16… they get out 
at 18. Sooner than older people. They can’t process them and besides that, there’s another 
thing… A the age of 18, the courts can’t give them a stiffer sentence than adults… And it’s like 
they have a fresh mindset and they make them do what they want. Well, this happened like 4 years 
ago, here in Padres, they were picking young kids up and sticking them in juvenile hall and they 
were just like 13 or 14.” 
Impact Evaluation Report– Mexico 
70
In all three cities, people perceive that drug trafficking activities are commonplace in many of 
the communities and that is where young people interact with criminal groups. In many cases, 
young people live under the uncertainty of being at risk or of being subject to violence by drug 
traffickers as well as blackmail and abuse by the police. Many perceive that the simple fact of 
being young makes the police believe that they are dangerous. The custom is to stop young 
people on the street asking for money - a “bite” (or bribe) in exchange for not arresting or 
detaining them. Often, they are beaten and accused of a crime: 
“Pues muchas de las veces yo pienso que los del crimen organizado agarran a los de las 
pandillas para la distribución… Pues eso es como trabaja aquí eso.” 
“Well, many times I think that the leaders of organized crime grab kids from gangs for 
distribution…. Well, that’s how it works around here…” 
“Puros jóvenes, andan en las drogas yo pienso que vamos pa allá con las muchachas y todo y la 
verdad que por eso se hace la pandilla, por eso hay muchos crímenes y no es que vamos con las 
muchachas y vamos a pistear y a tomar y ahí se empiezan pues…” 
“Just young kids… they mess with drugs I think.. We go out with the young girls and all and the 
truth is that’s why they became a gang, that’s why there are so many crimes and it’s not that we 
are hanging out with girls and going to get drunk or get wasted and that’s how they start…” 
“Y aparte a estas alturas los policías tampoco ya no se prestan, los policías también siempre 
están agrediendo a los jóvenes, antiguamente la policía estatal antes no era … obviamente la 
fuerza civil era la estatal, había una caseta, era la 46 en mi colonia, estaba arriba de un kínder, 
en esa caseta siempre había dos elementos estatales, esos elementos con los chavitos jugaban 
policías y ladrones y ahí los encerraban en la caseta y todo y los chavillos se la pasaban o sea, se 
acercaban al elemento ¿si me explico? ahorita ya los ven como sus enemigos, ya los ven con odio 
y temor.” 
“And aside from that, by now the police just stand by and do nothing, the police are also always 
beating up the kids… before the state police, it wasn’t that way… obviously the civil force was the 
state police, there was a guard house, it was 46 in my colonia (neighborhood), it was above a 
kindergarten, in that guard house there were always two state police officers, who played ‘cops 
n’ robbers’ with the kids and there was where they locked them in – in the guard house and all 
the little boys had a great time… that is to say they went up to the cops…. Do you know what I 
mean? But now, kids see them as their enemies… they look at them with hatred and fear.” 
The school environment should make young people feel safe and confident and have a healthy 
living environment, but in many cases it has become hostile and violent to their development. 
Some schools have now become places where kids are abused, discriminated against, get into 
fights, and deal drugs. The school is a microcosm of how people live in cities with respect to 
issues of insecurity, crime, drug trafficking, drug addiction and alcoholism, and especially the 
loss of values. Violence has become a daily part of life among young people which may even 
represent a certain status for them and also be a strategy for group survival. Gangs are in and out 
of schools, they started as groups with certain similar characteristics within the school, and later 
they go out to the streets and become a problem for society, because they are “emboldened” and 
Impact Evaluation Report– Mexico 
71
get into things they should not be a part of like taking drugs and drinking alcohol on the street, 
thefts and armed robberies. 
“… no es tanto el dinero ¿eh?, no es tanto el dinero… Por sentirse mucho más que otras 
personas, por querer ser alguien más importante.” 
“… It’s not so much the money, right? It’s not so much the money…. It’s about feeling better 
than other people and wanting to be someone more important.” 
“Pues muchas de las veces yo pienso que los del crimen organizado agarran a los de las 
pandillas para la distribución… Pues eso es como trabaja aquí eso.” 
“Well, a lot of time I think that organized crime leaders snatch up gang members for 
distribution… Well, that’s how it works around here.” 
“Porque se sienten solos a veces… Sí, para sentirse importantes, dicen: “aquí me siento 
cobijado, aquí me quedo”… Sienten que es su familia… Protegidos, se sienten protegidos… 
Cuando no tienen la responsabilidad de unos padres de familia que estén al pendiente de ellos, 
así es, osea, casi siempre son de padres separados… Claro, podría ser como una especie de 
compensación del abandono…” 
“It is because they feel lonely sometimes… Yes, in order to feel that they are important, they say: 
‘I feel safe, so I’ll stay here’… They feel like they are their family… Protected, they feel 
protected… When they don’t have responsible parents who take care of them, that is to say… 
ummm… they almost always come from broken homes… Of course this could be a kind of 
compensation for being abandoned…” 
“En la secundaria el subdirector nos correteaba con un pedazo de llanta para ponernos a mitad 
de patio con las mochilas cargadas y con las libretas y ahí nos paraba en todo el sol, hasta como 
dos o tres horas de clase… Pero nos daban así en los pies, nos pegaba en los pies, era un pedazo 
de llanta así como de hule y nos correteaba.” 
“In high school, the vice-principal ran after us with a piece of a tire and chased us out onto the 
middle of the playground with our backpacks loaded with notebooks and they left us there in the 
sun for 2 or 3 (class) periods… But they hit us on the feet, they hit us on the feet, it was a piece of 
rubber tire and they chased us around with it.” 
“…no cambia nada es lo mismo robos y robos … Los mismos alumnos si saben que dejan abierto 
un salón o algo ellos les avisan a otros y ya hacen… Pues ahorita están metiendo muchas 
unidades a las escuelas, hacen rondines para que no haiga nada de problema, es lo que está 
habiendo ahorita… Si ven un sospechoso o algo lo paran, si ven que anda haciendo el bien pues 
lo dejan, si ven que anda haciendo el mal pues se lo llevan ¿verdad?, es como todo.” 
“…nothing changes… it’s the same – robberies and more robberies… The same students know 
that if a classroom is going to be left open, the let the other ones know and they do it… Well, now 
they’re placing lots of police in the schools and they patrol around so that there are no 
problems.. that’s what happening now… If they see a suspicious kid or anything, they stop him, if 
they see that he is doing nothing wrong, they leave him alone, if they see that he is up to 
something he shouldn’t be doing, they take him away, right?, it’s like everything else.” 
Impact Evaluation Report– Mexico 
72
Within schools, gang members go unnoticed to some extent, because many schools require 
uniforms, but there are certain personality traits that give them away: attitudes, speech, and 
activities. It is common among gangs to graffiti school property to leave their mark that they 
were there. Previously they distinguished themselves by physical characteristics such as haircut, 
tattoos, and dress, however they are no longer distinguishable because they try not to draw 
attention to themselves. 
“Como cuatro o tres por cada salón también a veces, tres o cuatro por cada salón. Había cuatro 
pandillas por grupo y ya cuando salían al recreo pues ahí se hacía… Pero decían hey tiro en el 
campo y ahí iban todos corriendo y de ahí se peleaban de un grupo y de otro.” 
“Like four or three in each classroom, three or four in each classroom. There were four gangs by 
group and when they went out to recess, that’s where they did it… But they said “Hey… “tiro en 
el campo” (a call for gang members to join a fight. It literally means “A shot on the field”) and 
they all came running and that’s where one group fought the other. 
“A veces la forma de hablar con la persona, hablan muy grosero o te hablan muy fuerte, como de 
que tienen mucho poder y son más que tú, te humillan. Por ejemplo veces están así de que están 
dos camionetas y están abajo muchachos y vas pasando y te dicen voltéese que chingados anda 
viendo, váyase para allá, a usted no le importa, algo así o te sacan así una pistola y te dicen 
váyase de aquí, que anda haciendo aquí, así…” 
“Sometimes a person’s way of speaking… they speak really rudely or sometimes they talk to you 
really loudly, like they have a lot of power and they’re better than you, they humilate you. For 
example, sometimes there are two trucks and there are boys hanging around the trucks and they 
tell you ‘Turn around! What the hell are you looking at? Go away! It’s none of your business!’ or 
they take out a gun and they tell you to get out of there, what are you doing here?, like that…” 
Organized crime looks for vulnerable people, which generally means under-aged kids (mostly 
children) or people who are members of gangs (because they are often under-aged). Knowing 
that organized crime can solve much of their economic problems, and also give emotional 
benefits (such as being part of a group and belonging to a "family"), can make such a life 
attractive for young people the profile of those searching for solutions to their problems (both 
familial and economic). 
“Te amenazan… Si, si pues te sacan el cuete… A veces hasta son más menores que uno y pos ya 
trayendo una pistola en la mano…Es lo que se sienten, ya trayendo una pistola ya cualquiera te 
dice… De repente lo ves sin dinero y ya de repente lo ves con carro y así te das cuenta de volón 
que anda mal… O lo ves parado en unas esquinas ahí…Con los radios, con celulares, con drogas 
vendiendo.” 
“They threaten you… “Yes, yes well they take you out to get loaded… Sometimes they are 
younger than you and already carrying a gun in their hand… This is what they feel, already 
carrying a gun anybody will tell you… Suddenly you see them with money and suddenly they have 
a car and then you realize that they’re going down a bad path… Or you see them standing there 
on the corner… with radios, with cell phones and with the drugs they are selling…” 
Impact Evaluation Report– Mexico 
73
“Es atractivo para los jóvenes por el dinero, porque no quieren estudiar, quieren ganar dinero 
fácil, no quieren hacer nada… Te damos dinero y les dan una moto, tu ve checándome los 
puntos… Y lo más importante es ser violento y eso hace que ellos digan ay soy mejor, soy 
superior… Tiene ventajas para su familia porque si trabajas para ellos puedes cuidar y proteger 
a tu familia… Pero también los pones en riesgo, porque también es un riesgo para la familia… Y 
no es estable, podrás trabajar una semana y a la otra semana ya te mataron y ellos mismos están 
en contra.” 
“It’s attractive for young people because of the money, because they don’t want to study, they 
want to earn easy money, they don’t want to do anything… We’ll give you money and someone 
else gives them a motorcycle, you go around checking out the girls… And the most important 
thing is to be violent and this makes them say ‘Ay, I am better, I am superior… It has its 
advantages for their family because if you work for them you can take care of and protect your 
family… But you also put them at risk because it’s also it’s a risk for the family… And it’s not 
stable, you can work one week and next week, they’ll kill you and they’ll all be fighting each 
other.” 
Also, young people can be an easy target for organized crime as well as for the police. The police 
abuse young people in an authoritarian way, they ask them for money and even steal their 
belongings, taking advantage of the fact that the young people do not have any social standing 
and most of them are unaware of their legal rights. That is to say, young people are a part of the 
population that is abused and unprotected by all. 
 
“A mí me subieron a la patrulla y me dijeron… súbase, ah pues me subí, una policía me quitó mi 
visa y otro policía me quitó mi celular y luego ya le dije a una señora hey háblele a mi mamá y ya 
llegó y le dije mamá ella trae mi visa y él el celular, no, nosotros no traemos nada, si como no, se 
fue mi mamá a la casa y luego me dijeron porque estás hablando y me agarraron a bachones los 
policías.” 
“They took me inside the police cruiser and they said to me… get in, so I got in. A policewoman 
took my visa and a guy cop took away my cell phone and later I told a woman ‘Hey, go and tell 
my mother’ and she came and I told her ‘Mom, she (the policewoman) took my visa and he took 
my cellphone’ and then the police officers said “No we don’t have anything,” I said, ‘Yes, of 
course you do’ and then my Mom went home and they said to me ‘Why are you talking?’ and then 
took me and slapped me around. 
Social Cohesion 
Regarding the issue of Social Cohesion, we analyzed the results of three variables. The first is a 
social cohesion index built from four questions. The second is an interpersonal trust indicator 
which is a key component of social cohesion. The third is a social disorder index measured 
during the systematic observation phase and is a correlate of social cohesion. The following are 
the results from the first two because for the last model there was no significant difference 
between the treatment group and the counterfactual therefore that is not reported here. 
Impact Evaluation Report– Mexico 
74
Social Cohesion Index 
The social cohesion index is constructed with four questions about individual perceptions of the 
organization, cooperation, trust, and union of neighbors within the community that are different 
dimensions of social cohesion. The first measures the perception of the extent to which 
neighbors organize to face a common problem in their neighborhood. The second relates to the 
perception that the neighborhood is united. The third is about the people’s willingness to support 
each other and the fourth is about trust among neighbors. Out of these questions, we built a 
social cohesion index that ranges from 0-100. 
Figure 31 shows the results of the difference-in-difference analysis for the social cohesion index. 
We see that in the control group, this indicator improved slightly as the score rises from 50.6 to 
52 points (on a 0-100 scale). The results in the treatment group are the opposite with a slight 
decrease of two points from 53 to 51.1. Given the improvement in the control group, it was 
expected that the treatment group would report a score of at least 54.4 points (dotted green line 
of the counterfactual) which is 6% better than the obtained result. 
Figure 31. Social Cohesion Index 
The following three figures show results that we consider key with respect to the impact that 
context has on social cohesion. In the first one (Figure 32) we see, as expected, the impact of 
social disorder on social cohesion. We see that for the treatment group (green line), the model 
predicts an inverse relationship given that as the social disorder increases, the level of social 
50.6
52.0
53.0
51.1
54.4
Counterfactual
Treatment
Control
51.0
52.0
53.0
54.0
55.0
Social Cohesion Index
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
75
cohesion in the community decreases. For the control group (red line), no significant changes 
were observed as social disorder increased. 
Also as expected, we get the same results when we analyze the relationship between physical 
disorder and social cohesion. As in the previous case, Figure 33 shows that as the physical 
disorder increases, social cohesion decreases. Again we observed the same relationship in the 
treatment group but not for the control group. 
For the treatment groups, it could be a perverse relationship or vicious cycle where the worst 
physical and social conditions at the context level harm the social fabric (i.e. social cohesion) 
and a weakening of the latter has, in turn, additional consequences for the social and physical 
situation of the neighborhood. If that were the case, the situation is getting worse because of the 
fact that social cohesion (and levels of interpersonal trust discussed below), present for both the 
control group and the treatment group opposite trends, i.e. no improvement in the control group 
and a worsening in the treatment group. 
These two findings in relationship with social cohesion (opposite trends and the non-impact of 
the physical / social disorder over the control group) should be considered as relevant and we 
need to ask ourselves: what makes the control group improve and / or become “vulnerable” to 
changes in the physical and social disorder? Perhaps part of the answer has to do with the ability 
or internal strength that each one has, i.e. the need for external agents to improve (or rebuild) the 
social fabric. 
Figure 32. Marginal Change in the Social Cohesion Index in the Presence of Social Disorder on the Block -6 -4 -2 0 Marginal Change Social Cohesion Index
0 1 2 3 4
Social Disorder
Treated Non-Treated
Impact Evaluation Report– Mexico 
76
Figure 33. Marginal Change in the Social Cohesion Index in the Presence of Physical Disorder on the Block 
In Figure 34, we observe an example that illustrates this hypothesis. We see that for the control 
group, there is a direct association between the presence of churches and higher levels of social 
cohesion. This result suggests that agents like churches could be working as effective tools 
maintaining or improving levels of social cohesion. The intriguing thing is that for the control 
group, the presence of churches has the opposite effect, but with less intensity. -8 -6 -4 -2 0 Cambio marginal Social Cohesion Index
0 2 4 6 8
Physical Disorder
Tratada Non-Treated
Impact Evaluation Report– Mexico 
77
Figure 34. Marginal Change in the Social Cohesion Index in the Presence of Churches on the Block 
Interpersonal Trust 
With regard to interpersonal trust, a key component of social cohesion (Figure 35), we observe a 
behavior similar to and consistent with the social cohesion index described above. The control 
group improved slightly (52.1 to 53.7 points) and the treatment group worsened because 
interpersonal trust decreased, going from 55.5 to 51.1 points (on a scale of 0-100). In this 
indicator, the result for the treatment group is 10.5% less than the minimum expected given the 
improvement in the control group. -4 -2
0
2
4
6
Marginal Change
Social Cohesion Index
Absent Present
Church
Treated Non-Treated
Impact Evaluation Report– Mexico 
78
Figure 35. Interpersonal Trust Additionally, it is important to note two things. 
First, that in comparative perspective, Mexico is a country with low interpersonal trust, and 
second, levels of interpersonal trust reported in both the control and treatment neighborhoods are 
similar, although slightly lower than the national data. In Figure 36, the results of the 2012 
AmericasBarometer for 26 countries in the hemisphere are presented. They show that 
comparatively, Mexico occupies one of the last places in levels of interpersonal trust and is 
below most Central American countries. The national figure is slightly higher than those 
obtained for the neighborhoods studied in this impact evaluation. 
52.1
53.7
55.5
51.1
57.1
Counterfactual
Treatment
Control
50.0
52.0
54.0
56.0
58.0
Interpersonal Trust
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
79
Figure 36. Interpersonal Trust in the Americas 
Source: AmericasBarometer 2012 
As in the previous section, we present some qualitative research findings that complement these 
quantitative indicators. The following comments from focus groups with community leaders 
show the tremendous impact crime and violence can have on the social fabric, as well as on 
interpersonal trust between people as well as on the individual. These comments briefly illustrate 
how it possible to lose cohesion at community level and the need of working with young people 
on issues as basic and necessary as trust: 
 “Durante la crisis, hubo muchas familias que perdieron sus casas, sus trabajos, que hubo 
divorcios a raíz de una pérdida de una casa, a raíz de una pérdida de trabajo, entonces, esta 
gente es víctima también, víctima de una situación social macro a nivel país, y a lo mejor, 
40.3
50.3
54.4
55.7
56.4
56.9
58.0
59.0
59.9
60.2
60.5
60.9
61.5
62.0
62.4
63.0
63.1
63.9
65.2
65.7
65.9
66.4
66.6
67.1
67.5
67.6
Haiti
Peru
Bolivia
Belize
Mexico
Trinidad & Tobago
Brazil
El Salvador
Ecuador
Venezuela
Jamaica
Dom. Rep.
Nicaragua
Guatemala
Colombia
United States
Chile
Honduras
Suriname
Canada
Guyana
Uruguay
Costa Rica
Paraguay
Panama
Argentina
0.0 20.0 40.0 60.0 80.0
Interpersonal Trust
 95 % Intervalo de confianza 
 (Efecto de diseño incorporado)
Fuente:  Barómetro de las Américas por LAPOP
Impact Evaluation Report– Mexico 
80
incluso, con raíces mundiales, pero que ahora la estamos viviendo y hace que ya no haya unidad 
social" 
“During the crisis, there were a lot of families that lost their home, their jobs, there were a lot of 
divorces at a result of losing the house, as a result of losing a job, so, these people are victims as 
well, victims of a macro social situation on the level of country and, maybe, including, the origin 
comes from what is going on with the global economic crisis, but now that we are living through 
this, it makes it so there isn’t any social unity.” 
“Debe de haber más seguridad de parte de la misma persona, recuperar la confianza para poder 
empezar y volver a ser la misma gente que era antes…” 
“There should be more security for the individual, to regain trust in order to start over and be the 
same people that we were before…” 
“Orientación, aconsejar, hablar con ellos, dedicarles tiempo, dedicarles un minuto o dos, y hace 
la diferencia, claro que hace la diferencia, porque son muchachos que no tienen confianza en 
ellos mismos…” 
“Orientation, giving advice, talking with them, making time for them, giving them a minute or 
two, that makes the difference, of course it makes the difference because they are kids who don’t 
have confidence in themselves…” 
But still there are signs of optimism and positive attitudes which allow people to think about 
reaching better horizons in a short period of time. These comments serve to illustrate the hope 
and optimism among young people: 
“Pues es como le digo como ahorita, pos ahorita uno se siente feliz, alegre porque pos como le 
digo ya no hay tanta matanza, tanta delincuencia, tanta cosa que la verdad pues se llega uno 
hasta a trastornar de tanta cosa, ahorita ya no pasa eso, ahorita ya sales con gusto con tu 
familia a la calle a la tienda onde quieras, aunque no tan a gusto porque de repente anda un loco 
por ahí, pero si hay más unión en la sociedad” 
“Well, it’s like I’m saying like now, well, now you feel happy, happy because well, like I’m 
saying, there aren’t so many killings, so much crime, and so on – that the truth is, well, that you 
can go crazy from all this stuff, and now that doesn’t happen anymore. Now, you look forward to 
going out with your family to the street, to the store, wherever you want, although you’re not as 
carefree because you cross paths with some crazy guy, but there is more unity in society.” 
“Qué bien si no fue prevención como de un trabajo preventivo o social como debe de ser, el unir 
a la sociedad, pero creo que fue el primer paso para darnos cuenta como sociedad que teníamos 
que trabajar desde las escuelas…” 
“It wasn’t prevention from a preventive work or social as it should be, uniting society, but I think 
that it was the first step for us to realize as a society that we have to work from the schools….” 
Impact Evaluation Report– Mexico 
81
Crime Prevention Measures 
The questionnaire at the individual level contains several indicators on crime prevention. Below 
are those indicators that yielded significant results in the difference-in-difference models and are 
grouped into three categories: knowledge of crime prevention measures, individual actions to 
protect yourself from crime and collective actions to combat crime. Within these categories, we 
present the results for a total of seven indicators. 
Knowledge of Crime Prevention Measures 
Regarding knowledge on crime prevention, three indicators of our analysis show significant 
results. The first is a question about whether the respondent considers that he/she has basic 
knowledge of how to prevent crime and the results are shown in Figure 37. We observe that in 
both groups, there are slight changes but in opposite directions. While for the treatment group, 
we see a decrease in its score from 55.4 to 52.5 points, in the control group, we observe the 
opposite tendency to grow 50.6 to 53.8 (on a scale of 0-100). Again, the behavior in the 
treatment group is lower (10.3 %) than expected and in this particular case, opposite from the 
desired direction. 
Figure 37. Basic Crime Prevention Knowledge 
50.7
53.8
55.4
52.5
Counterfactual 58.5
Treatment
Control
50.0
52.0
54.0
56.0
58.0
Basic Crime Prevention Knowledge
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
82
The multi-level analysis shows that there is a significant relationship between the knowledge of 
crime prevention and the presence of the Catholic Church (Figure 38). In systematic observations 
of the blocks, we identified whether or not there was a presence of these institutions and 
analyzed this information. We observed no changes in the control group, but in the treatment 
group, there is a direct relationship between the presence of the Catholic Church and the 
increased levels of knowledge of crime prevention. That may suggest, as mentioned above, that 
churches play a relevant role as “agents” of change in treatment communities. 
Figure 38. Marginal Change in Basic Crime Prevention Knowledge in the Presence of Catholic Churches on 
the Block 
Regarding the perception of whether or not the government provides adequate information on 
how to prevent crimes, we see a small decrease in the treatment group from 57 to 52.4 points (on 
a 0-100 scale). That is to say, the score of citizen perception of the government providing 
adequate information decreases (Figure 39). In the control group, we observe no change over 
time as scores remain around 51 points. 
The minimum expectation for the treatment group, is that as it happened in the control group, it 
would have stayed around 57 points, but for some reason that is difficult to understand, the score 
dropped. However, the score for the treatment is 8% higher than expected. What makes adults in 
the treatment group in the second round perceive a decline in basic crime prevention knowledge 
and a lower perception that the government provides adequate information? Perhaps it has to do 
0 1 2
3
4
Marginal Change
Basic Crime Prevention Knowledge
Absent Present
Catholic Church
Treated Non-Treated
Impact Evaluation Report– Mexico 
83
with decreased campaigns or saturation (and less attention to the issue) or even it may be 
confusing information. 
Figure 39. Government Provides Adequate Information on Crime Prevention 
We observe a positive and higher than expected result on the question of whether or not the 
police have been helping with prevention activities in the neighborhood. Figure 40 shows that 
the percentage of adults who reported having seen police doing this type of activity grows in 
both groups, but with greater magnitude in the treatment group. The change in the treatment 
group is significant because it goes from 4% at the baseline to 12.2% in the post-treatment round. 
Those who were in charge of the evaluated programs are better qualified to know to what extent 
this marked increase is directly attributable to neighborhoods treatment programs. 
The control group also shows an increase in the percentage measured by this item as it goes from 
5.6% at baseline to 7.5% in the second round. The vertical black line that runs from the 
counterfactual to the result of the treatment group in the second round is considered the treatment 
effect and is equivalent to an increase of 107% higher than expected in the absence of the 
programs implemented in treatment neighborhoods. Later on, we present additional results on the 
perception and evaluation of the police, some favorable and some unfavorable, that complement 
the analysis of this important institution of the justice system. 
51.5 51.6
57.0
52.4
57.0
Counterfactual
Treatment
52.0 Control
54.0
56.0
58.0
Government Provides Adequate
 Information on Crime Prevention 
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
84
Figure 40. Police Carry Out Prevention Activities in this Neighborhood 
Individual Actions to Protect Oneself from Crime 
The questionnaire at the individual level has a considerable number of questions about actions 
that people can take to protect themselves from crime. In this section, we discuss the results of 
three of those items for which we got statistically significant results in the difference-in￾difference models. The first one is about the percentage of adults who reported having acquired a 
firearm to defend themselves from crime. As expected, the incidence of this indicator is low in 
both groups and the interesting thing is that it decreases in both groups in relation to the baseline; 
however the decline is more pronounced in the treatment group. 
Figure 41 shows that the percentage of adults who said they bought a gun for protection 
decreases in the control group from 1.8% to 1.3% and in the treatment group, goes from 3.1% to 
1.3%. Once again, the vertical black line that runs from the counterfactual to the result of the 
treatment group in the second round represents the effect of treatment and is equivalent to a 
reduction of 51.9% to the expected incidence in the absence of the programs implemented in 
treatment neighborhoods. 
5.6
7.5
4.0
12.2
5.9
Treatment
Control
Counterfactual
4.0
6.0
8.0
10.0
12.0
Police Carry Out Prevention
 Activities in this Neighborhood
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
85
Figure 41. Bought a Firearm for Self-Protection from Crime 
Regarding other actions that individuals may take to protect themselves from crime, we present 
the results of two additional indicators. First, we observe an improvement in the control group 
and a worsening in the treatment group; in the second indicator, the outcome worsens in both 
groups. 
Regarding the percentage of individuals who reported avoiding going through places that they 
consider to be dangerous as a prevention measure, we observe that the control group showed a 
significant decrease, going from 56.8% to 49% (Figure 42). Conversely, in the treatment group, 
we see that the indicator does not improve by having an opposite tendency, rising slightly from 
54% to 57.4%, a difference of 24% above that which was expected based off the counterfactual 
case. As noted throughout this report, the data that matters is the difference between the 
treatment group and the predicted value of the counterfactual (dotted green line). As already 
discussed, a positive evaluation consisting of (a) a change in the treatment group that represents 
an improvement (i.e. increase or decrease to something positive from something negative), and 
(2) that the result is greater than the counterfactual. In this case none of these conditions are met. 
1.8
1.3
3.1
1.3
2.7
Counterfactual
Treatment
Control
1.5
2.0
2.5
3.0
3.5
Bought a Firearm for Self-Protection from Crime
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
86
Figure 42. Avoided Dangerous Areas out of Fear of Crime 
As expected, in both the control group and the treatment group we observe higher levels of 
physical disorder. As there are higher levels of physical disorder, there are more people who 
avoid going through dangerous places as a crime prevention measure. We present the results in 
Figure 43. 
56.8
49.0
54.0
57.4
46.2
Treatment
Control
Counterfactual
45.0
50.0
55.0
60.0
Avoided Dangerous Areas out of Fear of Crime
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
87
Figure 43. Marginal Change in Avoiding Passing through Dangerous Areas in One’s Community in the 
Presence of Physical Disorder on the Block 
Finally, in Figure 44 we present the percentage of individuals who claimed to have built walls or 
fences at home as a preventative measure. We observe that in both groups this indicator is 
higher, i.e., the percentage of adults who respond that they “fortified” their home against crime 
by building walls, fences, or additional exterior walls is higher. The increase is significant in 
both groups but higher in the control group, growing from 21.9% to 34.1%. The treatment group 
has a result of 14.6% lower than expected in the absence of treatment. That this change has been 
less pronounced in the treatment group may be seen as positive, but the ideal is that the change 
over time be in the opposite direction. Yet, it is likely that through the programs implemented in 
the treatment group, there has been a minor increase in the number of households that considered 
it necessary to increase their security by building walls or fences. Another possibility is that 
changes are due to an improvement on their economic reality or their expectations or that 
because of this, individuals have been able to afford this type of expenditure. 
0
5 10 15 20
Marginal Change
Avoided Walking through Dangerous Areas in Community
0 2 4 6 8
Physical Disorder
Treated Non-Treated
Impact Evaluation Report– Mexico 
88
Figure 44. Built Walls/Fences to Protect Home from Crime 
Collective Action in the Prevention of Crime 
The last indicator in the category of prevention and protection against crime is the perception of 
the extent to which residents of the neighborhood are organized to prevent crime and violence. 
We observe a small shift in the perception of the control group that goes from 44.6 to 45.7 points 
(on a scale of 0-100); however in treated communities, we observe a marked decrease of this 
perception passing points 49.4 to 41.3 (Figure 45). The treatment group presents a result in their 
perception of community organization, 18.2% lower than expected in the absence of treatment. 
This movement in the opposite direction to the desirable one is consistent with decreased 
interpersonal trust and social cohesion index as shown in previous section and may be indicative 
of a widespread disenchantment of perceptions about the capabilities and strengths of the 
community. 
21.9
34.1
21.4
28.7
33.6 Control
Counterfactual
Treatment
20.0
25.0
30.0
35.0
 Built Walls/Fences to Protect Home 
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
89
Figure 45. The Community Is Organized for Crime Prevention 
Regarding the analysis of the impact of context, we observe something similar to what we 
reported above when discussing the social cohesion indicator: social disorder negatively impacts 
the community organization and the Catholic Church makes it better. In Figure 46, we can see 
that there is no significant effect on the control group but there is a significant effect for the 
treatment group where as social disorder levels rise, levels of community organization to prevent 
crime fall. 
In Figure 47, we again observe a change, which is very slightly pronounced in the control group 
but a significant change in the treatment group; in the latter, we observe that the presence of the 
Catholic Church improves levels of community organization for crime prevention. It seems again 
that the Church (Catholic) is an important agent in explaining this community indicator. 
44.6
45.7
49.4
41.3
50.5
Treatment
Control
Counterfactual
42.0
44.0
46.0
48.0
50.0
The Community Is Organized for Crime Prevention 
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
90
Figure 46. Marginal Change in Community Organization for Crime Prevention in the Presence of Social 
Disorder on the Block 
Figure 47. Marginal Change in Community Organization for Crime Prevention in the Presence of Catholic 
Churches on the Block 
-2
0
2
4
6
Marginal Change
Community Organization for Crime Prevention 
Absent Present
Catholic Church
Treated Non-treated
Impact Evaluation Report– Mexico 
91
Here we present some findings from the qualitative phase that complements the results of this 
section. For many of the youth who participated in the focus groups, it was obvious that there 
was an increased police presence in different activities and several young people and officials 
could identify specific programs supporting prevention. The following comments illustrate this: 
"Sí, porque ahí por la cuadra están sobres ahí los policías, porque antes ahí por la cuadra pues 
casi no había policías y ahorita andan mas ahí" 
“Yes, because there by the block, the police are there because before here on the block there 
were hardly any cop and nowadays, they are all over the place” 
"Pues es un programa que tienen en donde llevan a adolescentes que a lo mejor tienen problemas 
en su casa o problemas de drogadicción y ahí los educan los fines de semana ¿no?" 
“Well, it’s a program that they have where they take teenagers that might have problems in the 
home or drug problems and they educate them on the weekends, right?” 
"El grupo de D. A. R. E., surgió de Estados Unidos y luego lo trajeron acá a Mexico y los que 
imparten ese tipo de cursos son policías, son policías municipales que van a las escuelas, incluso 
ellos tienen algo que se llama Policía Juvenil, invitan a los jóvenes a que formen parte de ese 
grupo para prevenir las drogas y la violencia." 
“The group D.A.R.E originated in the United States and later on they brought it here to Mexico 
and the people that deliver these kinds of courses are police officers, they are city police that go 
to the schools, and they even have something called Youth Police where they invite young people 
to join their group to prevent drugs and violence.” 
"… en esa dependencia que se llama prevención del delito hay programas donde van a las 
escuelas y les dan explicación de que es lo que hacen y para que no se droguen o algo, pero 
depende…" 
“….on that thing that’s called crime prevention, there are programs where they go to the school 
and they give the kids explanations about what they do and so that they don’t take drugs or 
anything like that, but it depends…” 
"...como actuar por ejemplo está siendo uno asaltado o hay gente que a veces se mete a los 
locales a robar y pos como actuar, a veces nos dicen que es mejor quedarse callados que hablar, 
ahorita estaba un programa ahí en el centro comercial que van a poner alarmas de pánico 
silenciosas según, en cuanto haya un movimiento cada negocio, va a apretar su botón de 
pánico." 
“…how to act, for example, when you are being attacked or there are people that sometimes 
break in to the stores and steal and, well, how do you act, sometimes they tell us that it’s better to 
stay quiet than to talk, how there was a program there in the mall where they are going to put in 
silent panic alarms, whenever there’s any movement, every store is going to push its panic 
button.” 
Impact Evaluation Report– Mexico 
92
"...Seguridad Escolar que se enfoca a los conflictos de vía internet, lo que deben de tener cuidado 
los niños por decir con las páginas para adultos, páginas de delincuencia… Ha crecido bastante, 
ha ayudado a que muchos niños ya no estén…bueno últimamente que usan las redes sociales con 
su computadora, laptop, Iphone etc., que tengan más cuidado en seguridad de la familia para 
evitar secuestros, robos, etc." 
“…. School Safety that focuses on Internet conflicts, about how kids should be careful what they 
say on adult pages, crime pages… It has grown a lot, it helped many kids get off the Internet… 
well lately they are using social network with their computer, laptop, iPhone, etc., that they 
should be more careful with the safety of their family in order to avoid kidnappings, burglary, etc. 
"Y hay programas de gobierno también… El de Si Jala y Si Trabaja [SISI]… Va enfocado a las 
personas que quieren superarse… Específicamente los jóvenes… Te dan beneficios de seguir 
estudiando y trabajando, ya sea estas trabajando y te apoyan con una beca para tu estudio, 
también te informan de cómo no caer en las drogas y en las pandillas" 
“And there are government programs as well… The Si Jala y Si Trabaja [SISI]…. It focuses on 
people who want to improve themselves… Specifically young people… They give you incentives 
to keep studying and working or if you are working, they support you with a fellowship for your 
studies, and also inform you on how not to get involved with drugs and gangs.” 
And additionally, there are some comments that show the presence and identification of 
community measures to prevent crime in coordination with the police: 
"Pues en la clase alta yo me he dado cuenta que ellos se organizan, Los llamados vecinos 
vigilando o algo así… Si, vecinos vigilando exactamente son los que ellos se organizan y están al 
pendiente de lo que sucede alrededor, a sus alrededores y si ven algo extraño… Hay algunos 
privados o lugares como colonias donde tienen sistemas de alarmas, incluso sirenas que si algún 
vecino nota algún delito puede ir a un botón de pánico o un teléfono que va a tener conexión 
directa con la policía y si, pero son sistemas" 
“Well, I’ve realized that in the upper class, they get organized, the so called “neighborhood 
watch” or something like that. Yes, neighbors watching… that is exactly what they are.. they are 
organized and they are keeping an eye on what happens in their surroundings and if they see 
something strange… There are some private neighborhoods or places like colonias 
(neighborhoods) that have their own alarm systems, even sirens so that if a neighbor sees a crime 
being committed, he can go to a panic button or phone that is directly connected to the police and 
yes…. There are systems.” 
"...programación comunitaria ¿Verdad? Como los comités, yo estoy más cerca de las personas, 
ellos son policías, los ven a lo mejor con miedo o con respeto pero ya uno lo ven con más 
confianza y ¿sabes qué? Va a venir fulanita de tal, ella representa a la policía y va a venir echar 
unas platicas con ustedes y ya las personas acceden a que vengan y se juntan y empiezan a 
caminar por toda la colonia con la policía" 
“…community programs, right? Like those committees, I am closer to the people, they are police, 
they look upon them fear or with respect but now they’re looking at them (the police) with more 
trust and you know what? Miss So-and-So is going to show up, she represents the police and 
Impact Evaluation Report– Mexico 
93
she’s going to come to talk to you and already people allow them to come and they get together 
and start to walk around the neighborhood with the police.” 
Regarding the presence of firearms, we find comments that on the one hand show that young 
people can have access to them but, on the other hand, there have been efforts in the schools for 
instance that try to the make the possession and use of guns more difficult: 
 
"Bueno anteriormente, ahorita ya no hay muchos conflictos a lo mejor como dice aquí el 
compañero que no hay mucha difusión de las bandas que están operando pero hace unos años 
como en el 2008, 2009, pues era de que balaceras a cada rato, entonces obviamente todas esas 
armas son introducidas al país por el crimen organizado y se las venden muy fácilmente a los 
jóvenes... " 
“OK, before, now there aren’t as many conflicts maybe because as our friend here says there are 
not a lot of the groups around that are operating but a few years back like in 2008, 2009, well 
there were bullets flying all over the place all the time back then, so obviously, all these guns are 
brought into the country by organized crime and they sell them very easily to the young 
people….” 
"Bueno...desde la prepa, lo principal era desde riñas así simples hasta el uso de armas de fuego, 
se compraban en cualquier esquina casi, para poder asustar no sé a vivales que se creaban 
bandas ahí entre la preparatoria…" 
“Well, since high school, the main thing was from simple fights to the use of guns, we bought 
them on any corner almost, to be able to scare… I don’t know, your rivals that were making 
bands (gangs) there in the high school…” 
"… en revisarle a los jóvenes su material de la escuela, ahí logramos encontrar en ocasiones 
drogas, marihuana, logramos encontrarles…Armas blancas…incluso pistola se han encontrado, 
nos dicen que la compraron en la calle..." 
“By checking kids’ school stuff, that’s when we occasionally find drugs, marijuana, we manage 
to find knives…all the way up to a gun we found… they tell us that they bought it in the street….” 
Perception of the Police 
In this section, we present five indicators on the police with significant results in difference-in￾difference models: perception that the police are involved in educational and recreational 
activities, incidence of requesting police support, trust in the police, satisfaction of police 
performance, and incidence of bribes (“mordidas”) requested by the police. 
Do USAID crime prevention programs increase institutional support for institutions like the 
police? The expectation is that the implementation of programs in the treatment neighborhoods 
has an impact on trust and perception of police performance. We analyzed the results of Mexico 
in a comparative perspective to the other countries in the region. Figure 48 shows that trust in the 
police in Mexico is relatively low in comparative terms with a confidence level of 39.9 points on 
Impact Evaluation Report– Mexico 
94
a scale of 0-100, which places Mexico in the lower levels of the graph, with levels comparable to 
Paraguay, Peru, Bolivia, and Venezuela and below most Central American countries (except 
Guatemala and Honduras). 
Figure 48. Trust in the Police in the Americas 
Source: AmericasBarometer2012 
Figure 49 shows the results of the difference - in - difference models about the trust in the police. 
For the control group, we observe comparable levels to the national data reported above and a 
slight decrease is observed over time as it passes from 42.5 to 41.9 points on a scale of 0 to 100. 
With respect to the treatment group, we observe an initial value, that is to say the baseline that is 
considerably larger at 47.6 points and experiences a marked drop over time, falling to 40 points, 
a score similar to the national average in the post-treatment round. It is not good news that trust 
28.7
32.8
34.9
34.9
37.6
38.9
39.9
40.1
41.6
43.3
45.5
45.8
48.0
51.4
51.6
52.3
53.5
53.7
54.0
55.0
59.8
60.6
61.8
62.9
63.5
65.7
Honduras
Trinidad & Tobago
Guatemala
Dom. Rep.
Venezuela
Bolivia
Mexico
Peru
Paraguay
Argentina
Costa Rica
Guyana
Jamaica
Belize
Brazil
Uruguay
Colombia
Ecuador
El Salvador
United States
Panama
Nicaragua
Haiti
Suriname
Canada
Chile
0.0 20.0 40.0 60.0 80.0
Trust in Police
 95 % Intervalo de confianza 
 (Efecto de diseño incorporado)
Fuente:  Barómetro de las Américas por LAPOP
Impact Evaluation Report– Mexico 
95
in the police has fallen in both groups and even worse that the treatment group shows a 
considerable drop and a 15 % lower score than expected given the behavior of the control group. 
Figure 49. Trust in the Police 
There are two elements about context that has a favorable impact on trust in the police: the 
presence of a Catholic Church and the presence of a hospital. In both cases, the impact affects 
both the control group and the treatment group, but the marginal change is significantly higher in 
the latter. Figure 50 shows the direct and favorable relationship that the presence of Catholic 
Churches on trust in the police (We do not show the model on the presence of hospitals but the 
effect is the same). 
42.5
41.9
47.6
40.0
47.1
40.0
42.0
44.0
46.0
48.0
Trust in the Police
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
96
Figure 50. Marginal Change in Trust in the Police in the Presence of Catholic Churches on the Block 
0
2
4
6
8 10
Marginal Change
Trust in the Police
Absent Present
Catholic Church
Treated Non-Treated
Impact Evaluation Report– Mexico 
97
The paradoxical decline in the decrease of trust in the police is the fact that individuals reported 
greater activity and involvement of the police in potentially beneficial community activities. On 
the one hand, the percentage of individuals in the treatment group who said the police help carry 
out prevention activities increases and something similar happens with the incidence of police 
involvement in recreational and educational activities with children and young people reported in 
Figure 51 below. 
It is noteworthy that in the baseline in both groups, the incidence of police involvement in these 
activities was low (5%) and increases significantly in both groups but more sharply in the 
treatment group. For this indicator, the difference-in-difference model shows a favorable impact 
of treatment effect as the treatment group is 17% above the expected value in the absence of 
treatment. 
Why does trust in the police decrease in the face of a greater presence of this institution in 
prevention activities and recreational activities for children and youth? The following indicators 
give additional clues to answer this question and suggest that more is not necessarily better. 
Figure 51. Police Get Involved in Recreational Activities 
As expected, the presence of schools has a direct impact on police involvement in recreation and 
prevention activities as shown in the multi-level model in Figure 52. This effect occurs in the 
treatment group but not in the control group which shows the opposite behavior. Perhaps this is 
further evidence that there are greater coordinated and synchronized institutional efforts in the 
treatment neighborhoods than in the control neighborhoods. 
5.0
10.8
5.0
13.6
Treatment
Control
Counterfactual
4.0
6.0
8.0
10.0
12.0
14.0
Police Get Involved in Recreational Activities 
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
98
Figure 52. Marginal Change in Reports of the Number of Times that the Police Get Involved in Recreational 
Activities in the Presence of Schools on the Block 
The questionnaire at the individual level asks respondents about the perception of police 
performance; the results are shown in Figure 53. We observe opposing trends between treatment 
and control groups. While we observe a small absolute increase in the perception of the police 
performance going from 43.3 to 44.6 in the control group, the treatment group reported a slight 
decrease going from 46.6 to 45.5 (on a scale of 0-100). The dotted green line that represents the 
counterfactual reveals that the outcome in the treatment group is 5% lower than expected. 
It is likely that in the control group, the decrease in trust in this institution is explained in part by 
the fact that the performance rating shows no improvement. -2
0
2
4
Marginal Change
Police Involved in Recreational Activities
Absent Present
Schools
Treated Non-Treated
Impact Evaluation Report– Mexico 
99
Figure 53. Satisfaction with Police Performance 
An additional indicator about the police, that can help us to understand the decrease in trust in 
this institution, has to do with the number of reports by individuals of bribes. Figure 54 presents 
the results to the question of whether a police officer asked the respondent for a bribe in the last 
twelve months. We observe, as in the chart above, a similar pattern: decrease in the control group 
and an increase in the treatment group. The decrease in the control group is 3.3 points from 
19.9% to 16.6% as shown by the red line. The treatment group presents an opposite trends as it 
slightly increases from 16.6% to 17.8% in the frequency of bribes requested by the police. 
Again, in contrast to the counterfactual, the result in the treated communities is 33% worse than 
expected. 
It is unclear whether the police shows different performance depending on the area it serves, i.e. 
whether a treatment or a control group; or if the differences have more to do with the presence of 
agents where greater presence accounts for a greater chance of incidence of bribes. 
43.3
44.6
46.6
45.5
Counterfactual 47.9
Treatment
Control
43.0
44.0
45.0
46.0
47.0
48.0
Satisfaction with Police Performance 
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
100
Figure 54. Police asked for a Bribe in the Last Year 
Finally, within this category of results, we present the incidence of police support requested. This 
indicator may mean at least two things. On the one hand, an increase in the amount of requests 
for help from the police can be seen as something negative if it is related to an increase in the 
reasons (for example insecurity) that make those requests necessary. On the other hand it may 
be that the increase indicates a greater presence (closeness, availability) of officers in the 
community. What is not consistent for the treatment group, is the decrease in the trust levels 
reported earlier (see Figure 49 above) when there is a greater presence and contact with the 
police via requests for support as shown in Figure 55. 
Regarding this indicator about the frequency of requests for help from the police, we find, once 
more, opposing trends. While the control group shows a decrease in the number of requests to 
the police for support, from 19.3% to 15.9%, in the treatment group, the frequency increases 
significantly from 18.1% to 24.1%, i.e. an increase of 6 points. 
Increases in the indicator will be interpreted as a sign of improvement because it seems to be 
clear that there has been an increased police presence in the treatment group shown by the 
increase discussed above, and a large segment of the population relies on police or is satisfied 
with their performance. Additionally, as we reported in the first category of results, there is a 
decrease in the treatment group in the incidence of robberies and overall crime victimization rate 
suggesting that the request for support to the police is not explained by an increase in the most 
frequent types of crime and it could mean more police willingness to support the communities. 
19.9
16.6
16.6
17.8
13.4
12.0
14.0
16.0
18.0
20.0
Police asked for a Bribe in the Last Year
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
101
Figure 55 shows the effect of the treatment that is equal to a 52% improvement above what is 
expected regarding the solicitation of support from the police in the absence of the implemented 
programs. 
Figure 55. Asked for Help from the Police 
Qualitative research shows many comments about the police that have already been presented in 
the categories above. The following are some comments that illustrate the still present distrust 
and distance of many citizens from the police and reveal some possible factors explaining why 
the institution gets poor performance evaluation results: 
" Les tienes más miedo a los policías que a los delincuentes." 
“You are more afraid of the police than of the criminals.” 
"Son gandallas los polis." 
“The cops are such bullies.” 
"...desde que entraron y están haciendo sus vueltas es cuando siento más inseguridad que cuando 
no están" 
19.3
17.1
18.1
24.1
15.9
Treatment
Control
Counterfactual
16.0
18.0
20.0
22.0
24.0
Asked for Help from the Police
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
102
“…since the came in, when they are walking their rounds is when I feel the most unsafe 
compared to when they are not…” 
Government and Institutions 
Finally, in this section we discuss four additional indicators that have to do with other institutions 
and / or the government and their performance in the area of public safety. The first has to do 
with the frequency of community institutions that carry out public works such as improving 
lighting, cleaning or construction / repair of streets, fields or parks. In Figure 56, we observe a 
considerable increase in the frequency of such institutions for the treatment group that goes from 
21.2% in the baseline to 28.1% in the post-treatment. By contrast, in the control group, we see a 
marked decrease of about half, going from 27.3% to only 15.6% in the post-treatment. It is 
unclear whether or not this outcome is directly attributable to the USAID prevention programs 
and if that is the case, the treatment effect is equivalent to an improvement of 192% more than 
expected in the absence of programs implemented. 
This result is consistent with other indicators shown in this impact evaluation suggesting that the 
treatment group is getting a lot more attention and / or social investment given the increase in 
indicators such as increased police activities and involvement with youth, an increase in the 
number of institutions that carry out support, and recreation for the young and public works in 
their neighborhood. In this regard, it seems to be a “synchronization of attention” focused 
primarily on these treatment groups. 
Figure 56. Public Works Carried Out in the Neighborhood in the Past Year 
27.3
15.6
21.2
28.1
9.6
10.0
15.0
20.0
25.0
30.0
Public Works Carried Out in the Neighborhood 
in the Past Year
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
103
The presence of multiple support programs and beyond the ones exclusive to USAID, could 
mean several things: (a) real cooperation / synchronization between different institutions; (b) we 
expect much more favorable results in the treatment group; (c) programs could be 
complementing each other (although it is also possible that they could contradict each other, or 
cancel each other out); and/or (d) it could be a sign that many actors identify the treatment 
neighborhoods as the most needy and unsafe. 
The next indicator shows the perception that the nearest level of government, i.e. the municipal 
level, is improving public safety. Figure 57 shows that this perception decreased in both groups 
but the decrease is slightly more pronounced in the treatment group. The results are not in the 
desired direction and for the treatment group they are 11.5% worse than those results predicted 
by the counterfactual. 
Figure 57. Perception that the Municipality is Improving Public Safety 
Using data from the AmericasBarometer, in Figure 58 we present the national results from 2012 
for this indicator. Interestingly, we see that in the baseline and perception for both groups in the 
neighborhoods of the three cities studied were better than the national indicator. The decrease 
over time makes the results in the post-treatment round more closely resemble the national 
results from 2012. On the other hand, we observe that Mexico ranks high comparatively 
54.8
51.4
58.9
49.2
55.6
Counterfactual
Treatment
Control
50.0
52.0
54.0
56.0
58.0
60.0
Municipality is Improving Public Safety
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
104
speaking, but some countries with more violence as measured by the homicide rates, such as 
Nicaragua and El Salvador, receive a better rating than Mexico. 
Figure 58. Evaluation of Government Management of Citizen Safety in the Americas 
Source: AmericasBarometer 
In a related indicator about whether or not the government provides adequate information on 
public safety, the results of Figure 59, on a scale of 0-100, show a slight increase for the control 
group (50.5 to 51.6) and a small decrease in the treatment group (55.8 to 52.4). Given the 
opposite direction of trends and the fact that in the treatment group there is no improvement, we 
could say that the programs are not having the expected impact regarding this indicator. In terms 
of the model, we could say that for the treatment neighborhoods, we observe a behavior of 8% 
31.3
36.2
38.5
39.8
39.9
40.2
41.1
41.8
41.9
43.7
45.2
45.8
46.0
46.2
46.6
47.3
47.5
47.8
49.6
52.4
53.3
55.2
56.3
61.5
63.4
Honduras
Paraguay
Argentina
Brazil
Bolivia
Costa Rica
Dom. Rep.
Trinidad & Tobago
United States
Peru
Panama
Jamaica
Venezuela
Uruguay
Guatemala
Guyana
Haiti
Chile
Mexico
Belize
Colombia
El Salvador
Ecuador
Suriname
Nicaragua
0.0 20.0 40.0 60.0 80.0
Administration Improves Security
 95 % Intervalo de confianza 
 (Efecto de diseño incorporado)
Fuente:  Barómetro de las Américas por LAPOP
Impact Evaluation Report– Mexico 
105
less than expected if a similar behavior had occurred in these groups rather than in the control 
groups. 
Figure 59. The Government Provides Adequate Information on Public Safety 
Finally, we present the perceptions of trust in the justice system, a variable that is expected to 
improve with the implementation of the programs. The results of Figure 60, however, report that 
for both groups, there were decreases over time (greater in the treatment group and 9% lower 
than expected). 
TREATMENT 
Impact Evaluation Report– Mexico 
106
Figure 59. Trust in the Judicial System 
Regarding the analysis of the impact of context on trust in the justice system, we see something 
similar to that which was reported above when discussing the indicators of social cohesion and 
community organization: social disorder negatively impacts the trust in the justice system and 
the presence of a Catholic Church improves it. In Figure 61, we can see that for the control 
group, there is a very slight effect but the effect in treatment group is significant and shows that 
as levels of social disorder increase, we see a decrease in trust in the judicial system. 
55.5
51.2
56.8
47.9
52.6
Counterfactual
Control
Treatment
48.0
50.0
52.0
54.0
56.0
Trust in the Judicial System
(Predictors)
Pre-treatment Post-treatment
Time
*Significantly different from 0 at p< .05
Source: CARSI Study by LAPOP
Impact Evaluation Report– Mexico 
107
Figure 60. Marginal Change in Trust in the Judicial System in the Presence of Social Disorder on the Block 
In Figure 62, we observe another change, this time less pronounced in the control group, but a 
significant change in treatment group. In the latter, we observe that in the presence of the 
Catholic Church, we note better levels of trust in the justice system. It seems again that the 
Church (Catholic) becomes an important agent impacting meaningful indicators like this one. -10 -8 -6 -4 -2
0
Marginal Change
Trust in the Judicial System
0 1 2 3 4
Social Disorder
Treated Non-Treated
Impact Evaluation Report– Mexico 
108
Figure 61. Marginal Change in Trust in the Judicial System in the Presence of a Catholic Church on the 
Block 
The following are some comments from the qualitative research that show clearly that many 
citizens perceive multiple efforts to improve conditions in the city and crime prevention 
programs: 
"Las carreteras que están haciendo, si han mejorado mucho” 
“The roads they’re making… they’ve gotten a lot better…” 
"Pos yo si he visto que ponen parques” 
“Well, yeah I’ve see that they’ve put in parks.” 
"Hay más centros comunitarios ya." 
“There are more community centers now.” 
"...Pos los sistemas de vialidad que ha implementado el gobierno y que ha estado haciendo, vías 
rápidas, fuentes de trabajo, maquiladoras." 
“…We the road system that the government’s put in and they’ve been building fast highways, 
sources of jobs, factories (maquiladoras).” 
0
2
4
6
Marginal Change
Trust in the Judicial System
Absent Present
Catholic Church
Treated Non-Treated
Impact Evaluation Report– Mexico 
109
"Están rehabilitando también las canchas y los parques, me ha tocado ver, una colonia que en su 
tiempo fue muy marginada La Cañada Verde y por medio del Gobierno Federal hicieron muchas 
canchas" 
“They are rehabbing the (athletic/sports) courts and the parks, I’ve seen it… a neighborhood that 
its time was very marginalized, La Cañada Verde and through the Federal Government, they 
made a lot of courts.” 
",...en las colonias se vio en muchas colonias por parte del municipio o del gobierno no sé 
muchas brigadas limpiando de pintura, pintando fachadas y ese fue un programa del gobierno, 
del municipio de pintar fachadas, iban cuadrillas cuadra por cuadra y preguntaban ¿quieres que 
te pinte tu fachada?, órale pues, de color fulano, tengo estos colores, entonces entraba uno a la 
colonia y ya la veía más diferente, ya pintadito aquí" 
“… and in the colonias (neighborhoods), we saw lots of neighborhoods… from the government or 
the city… I don’t know… there were brigades of people cleaning with paint, painting walls and 
this was a government program, from the city to Paint the walls, they went in squads block by 
block and they were asking “Do you want us to Paint your Wall? Yes, go ahead, use that color, I 
have these colors, then you went into the colonia and you saw a big difference, all painted.” 
"Y lo que está haciendo el municipio en las colonias son por parte de seguridad pública se 
llaman marchas exploratorias, en estas marchas exploratorias va el representante de la 
seguridad pública, un promotor de cada colonia y va y le presenta a los vecinos ¿Verdad? A los 
colonos al comándate, al fulanito que les va a dar el rondín y detectan puntos o usted con toda 
confianza." 
“And what the municipality is doing in the colonias (neighborhoods) as part of public safety is 
what is called exploratory walks. On these exploratory walks, there is the public safety 
representative, a promoter from each colonia and they go and introduces them to the neighbors. 
Right? They introduce the neighbors to the representative, to the guy who is going to show them 
around and they identify spots or you with all trust.
Impact Evaluation Report - Mexico 
110
Conclusions 
During the course of the evaluation, we conclude now that only some of the outcomes in the 
treatment communities improve more (or decline less) than they would have if the USAID 
programs had not been implemented. Several other outcome are moving in the right direction, 
but the change is less than expected which may mean that programs still do not have all the 
possible impact. However, a set of additional indicators in the treatment group moves in the 
opposite direction than what is desired, which means that they worsen and/or the programs have 
not been able to make a favorable impact or at least stop the negative trend. 
Among the main results of the quantitative phase of the study, we observe that the USAID 
prevention programs in the treatment neighborhoods have an impact on: 
 (1) Increased presence of support programs and government actions 
 59% greater knowledge of the institutions that carry out actions for the youth to use their 
time in a better way than would be expected in the absence of the programs 
 120% greater knowledge of the institutions that support the youth than would be expected 
in the absence of the programs 
 160% greater reported cases of police doing prevention activities than would be expected 
in the absence of the programs 
 17% greater reported cases of police working with children and young people in 
recreational activities than would be expected in the absence of the programs 
Regarding the positive improvements in the treatment group but that show changes that were 
lower than expected given the programs implemented, we can report that: 
(2) There is a reduction of crime victimization 
 Decrease of six points in the percentage of occurrences of theft in their community (63% 
less than expected) 
 Slight reduction in the overall rate of victimization (88% less than expected) 
(3) Decreased fear of crime and feelings of insecurity 
 Very slight reduction in the feeling of fear of crime in their community (7% lower than 
expected) 
 Slight reduction of concerns that the home would be broken into (9% lower than expected) 
 Decrease of eight points (on a scale of 0-100) in the perception of insecurity of living in 
their state (13% less than expected) 
Impact Evaluation Report– Mexico 
111
(4) Prevention and vulnerability of young people 
• Slight decrease in the reports of knowledge of a young person consuming drugs over the course 
of the last year (33.5% to 33.6%) 
As for changes in the opposite direction of the desired result given the expected impact of the 
programs that USAID implemented, we report that: 
In the Community Context 
 Slight decrease in the social cohesion index (55.5 to 51.1 points on a 0-100 scale) 
 Decreased interpersonal trust (53 to 51.1 points on a 0-100 scale) 
 Considerable reduction in the perception that the community is organized to prevent crime 
(49.4 to 41.3 points on a 0-100 scale) 
 Slight increase in the reported knowledge of young people who abuse alcohol (40.5 to 
41.2 on a 0-100 scale) 
 Considerable increase (23.4% to 34.4%) in the percentage of respondents who report the 
sale of illegal drugs in their community 
On Police 
 Decreased trust in the police (47.6 to 40 points on a 0-100 scale) 
 Slight reduction in the perception of police performance (46.6 to 45.5 points on a 0-100 
scale) 
 Slight increase in the incidence of frequency of the solicitation of bribes (16.6 to 17.8 
points on a 0-100 scale) 
Fear of Crime and Feelings of Insecurity 
 Slight increase in the percentage of individuals who avoiding walking through dangerous 
places 
 Slight increase in the percentage of individuals who built walls / fences to protect 
themselves from crime 
 Slight increase in the percentage of individuals who express a desire to move to a different 
neighborhood due to insecurity (14.5% to 16%) 
 Slight reduction in the basic knowledge of crime prevention measures (55.4 to 52.5 range 
on a 0 -100 scale) 
The context in which treatment was administered was related in some cases with its 
effectiveness. We see specifically that the presence of the Catholic Church and hospitals in the 
community have a major and significant impact. 
Impact Evaluation Report - Mexico 
112
In particular, we observe a lower rate of drug sales, less concern that someone may come to rob 
your home, more basic knowledge of crime prevention, a better organized community to fight 
crime, greater trust in the police, and judicial system in places where there is a presence of the 
Catholic Church in the community. We also observe an improved perception of safety and better 
levels of social cohesion in places where there is the presence of Catholic Churches and/or 
Protestant churches. 
The presence of hospitals has a significant impact on some variables that have to do with public 
health issues such as the lower rates of the sale of illegal drugs and less knowledge of young 
people using drugs; but it also impacts other variables such as lower levels of fear of crime in the 
community, and greater trust in the police. 
The treatment also may counteract some deleterious effects that “poor” contexts have in terms of 
resources and physical and social characteristics. Where we see social disorder on the block, the 
treatment managed to contain the marked tendency that disorder has to reduce social cohesion 
and community organization, and to increase crime prevention and confidence in the judicial 
system. Similarly, in the context of physical disorder, the treatment increases social cohesion and 
lowers the rate of people who report avoiding walking through dangerous places. 
However, there are also counter-intuitive behaviors such as decreased theft, sale of drugs, or fear 
of crime in context with greater social disorder; or greater knowledge of youth consuming 
alcohol and decreased fear of crime in contexts with large physical disorder. 
In many cases, the quantitative results were supported and complemented by qualitative data. 
Below are some quotes that illustrate perceptions and/or both positive and negative attitudes in 
relation to the context, mood, and security situation in the three cities we examined. We start 
with positive comments that illustrate the positive impact that recent actions have had, and 
differences by city: 
Regarding crime prevention: “Pues para prevenir la delincuencia las administraciones pasadas fue 
pues hacer de nuevo los parques y crear así áreas para que los muchachos fueran a jugar, a hacer 
deporte. De hecho ese es uno de los grandes logros, ahí si los felicito en lo personal porque se han 
visto parques por allá afuera de donde yo vivo han remodelado todo, han puesto todo nuevo, canchas 
de futbol pa’ los niños y ya se ve, ya se ve un cambio muy satisfactorio en cuanto al deporte, se ven a 
los niños jugando.” (Ciudad Juárez) 
“Well, in order to prevent crime, the past administrations did the parks over and that way created 
áreas for children to play and to do sports. In fact, this is one of the greatest achievements, 
Personally, I have to congratulate them in that area because there have been parks around where I 
live where they have re-done everything and they have made everything new – soccer fields for kids 
and you can already see a very sastisfactory change in terms of sports, you see kids playing.” 
(Ciudad Juárez) 
Regarding the presence of the Marines or the Army: “Porque antes, antes haga de cuenta para las 
tres o dos de la tarde…Ya no se podía salir. Se escuchaban las balaceras, al ratito mataron a uno 
allá y acá a otro. Ya no podías salir a la calle. Antes entraban y salían y ahora no porque a veces 
anda la marina o los soldados o la fuerza civil y no entran las camionetas, ahora ya se puede salir 
sin tanto miedo.” (Monterrey) 
Impact Evaluation Report– Mexico 
113
“Because before, before three or two in the afternoon, you couldn’t go out. You Heard gunshots or 
Heard them killing people… one here, one there and you couldn’t go out on the Street. Before they 
came and left and now they don’t because sometimes the Marines or the soldiers or the civil force are 
patrolling around and trucks don’t come in, now you can go out without being so scared.” 
(Monterrey) 
Regarding the police: “Si, yo veo mucha mejoría porque antes si se veía que llegaban y subían a todo 
el que estuviera, no preguntaban si tú estabas o no estabas, nada más los subían, nomás los subían y 
lo peor del caso es que les ponían que hicieron esto, hasta lo que no, e mí en esa cuestión me tocó 
muchas veces ver con los vecinos que los ponían ahí como los peores delincuentes y ahora no.” 
(Ciudad Juárez) 
“Yes, I see a bug improvement because before if you saw that they were coming and everybody who 
was there got it, they didn’t ask it you were in or not, they just picked you up, they just were picking 
people up and the worst thing is that accused them of doing stuff, even things that they had not done, 
and me – I lived that situation many times. I saw neighbors that were accused of being criminals and 
that doesn’t happen anymore.” (Ciudad Juárez) 
Regarding the insecurity situation: “No, de hace dos años a la fecha no, a lo mejor de hace un año 
para acá si mejoró un poco. Sí, porque por ejemplo en las noticias ya no vemos este… Tanta 
violencia, Ya no hay tantas balaceras como antes. Sí, así tan agresiva aquí en Nuevo León que llego 
a ser un estado donde decíamos si tuviera la oportunidad de erradicar en otro lugar que estuviera 
tranquilo pues me iría pero este todo igual en todas partes entonces de un año para acá como que si 
ya nos hemos mejorado.” (Monterrey) 
“No, two years ago to the day, or maybe since one year ago, things have got a little bit better. For 
instance, in the news, you don’t see so much violence. There aren’t as many gunshots as before. Yes, 
it became dangerous here in Nuevo León that it became a state where we said that if we had the 
opportunity to move somewhere else to a peaceful place, then we would go but then again, it’s the 
same everywhere but from one year ago, we see that things are getting better.” (Monterrey) 
“Si, de hecho yo hace cinco años que llegué, llegué a vivir a la Zona Norte, yo vengo de Morelos y 
cuando yo llegué todavía encontrabas mucha gente inyectándose en la calle tirada y ahorita la 
verdad es que no ves eso la verdad, lo que es la calle…Primera… La Coahuila y La Primera la 
cambiaron muchísimo. … El cambio es en todo, está más limpio, ya no hay gente que se esté 
inyectando ahí a la vista, continuamente pasan las patrullas y a veces hasta los llevan, llevan muchos 
en la patrulla y se los llevan si ven que andan ahí…Estamos hablando de un 75% de cambio, pero si 
ha cambiado mucho la Zona Norte, la Zona Centro donde está la Catedral, la Calle Segunda, la 
Calle Tercera, también había mucho malandro como dicen, ahorita ya no, ya disminuyó mucho eso.” 
(Tijuana) 
“Yes, in fact I arrived five years ago, I arrived to live in the Zona Norte (Northern Area), I come from 
Morelos and when I arrived, you still saw a lot of people shooting up, sprawled out all over the street 
and now the truth is that you don’t see this so often, talking about the streets, First…La Coahuila y 
La Primera (on Coahuila and First Street), things have changed a lot… You can see the changes 
everywhere… it’s cleaner, there aren’t people shooting up there in plain sight, the patrols go by all 
the time and sometimes they take them away, the patrols take a lot of people… they take them if they 
see that they are around… We are talking about a 75% change, but yes the Zona Norte has changed a 
lot, la Zona Centro (Central Area) where the Cathedral is, Second Street, Third Street, there used to 
be a lot of theft as they say, but not now… a lot of that has gone away.” (Tijuana) 
Impact Evaluation Report - Mexico 
114
Regarding perceptions and negative, or less favorable attitudes, here are four comments 
that show that many participants still perceive and/or face adverse situations in their 
everyday lives and suggest that there are still many challenges: 
“.. entonces hay muchas cosas, la delincuencia ahorita está por donde quiera casi al 100%, entonces 
sí muchas veces nos vamos a ocupar por ejemplo los directores se ocupan muchas veces dentro de lo 
que es la escuela correcto, pero también debe de ver que si no se enfoca a raíz de los alumnos que 
porque como en muchos casos está pasando que dejan las escuelas vacías, se llevan las 
computadoras, se llevan todos los valores y eso es un daño para los jóvenes, para los escolares, los 
dejan sin protección, si armas, entonces lo que tiene que hacer ahí el director es exigirle al gobierno 
que le pongan guardias al trayecto de las 12 horas de la noche hasta las 6 de la mañana.” (Tijuana) 
“… so there are a lot of things, crime now is around 100%, so yes, a lot of times we take care of 
things, for example the principals take care of what’s going on in the schools, right, but also they also 
should see that if they don’t attack the problem at its root, because in many cases, students leave the 
schools empty, they take the computers, they take anything valuable and this causes damage to the 
kids, to the students, it leaves them without protection, yes, weapons, so what the principal has to do 
there is demand of the governement that they place guards on patrol at the schools from 12AM – 
6AM. (Tijuana) 
“Como yo lo veo es el hecho de que Tijuana en si está aumentando ¿no? o sea está creciendo en 
tamaño y eso pues está como implícito que va a crecer todo así ¿no?, los empleos, la educación, la 
industria y también el crimen organizado o sea ahora los que tenemos chance de vivir aquí más en 
una zona céntrica nada más lee lo que dice el periódico pero digo en las afueras matan gente todos 
los días y no sabe la gente, nomás te dicen las cosas bonitas.” (Joven Tijuana) 
“As I see it, the fact is that in Tijuana, it’s getting worse, right? I mean that it’s increasing in size and 
well, it seems that everything is just going to keep getting worse, right?, jobs, education, industry as 
well as organized crime – meaning that now we have the chance to live here in downtown…all you 
have to do is read what the newspapers say, but I say that in the suburbs they are killing people every 
day and no one knows, they just tell you the good things.” (Joven Tijuana) 
“A lo mejor muchos se van a identificar conmigo, cuando estaban pequeños a veces hasta la familia, 
eran muy unidos los vecinos y si había alguna controversia o diferencia había a fuerzas se ponían de 
acuerdo y se les quitaba el enojo ¿en cuestión de qué?, de que antes estábamos chavitos y todos en la 
calle jugando soccer un ejemplo, soccer y ponía uno así las piedritas como portería, las niñas 
jugando a la muñeca con la vecina, los papás conviviendo en mecedoras en el porche y ahorita ya 
no, ya está todo solo, vacío y a las diez de la noche ya es como si fueran las dos de la mañana.” 
(Monterrey) 
“Maybe a lot of people are going to identify with me, when they were little, even the family was really 
tight with the neighbors and if there was any kind of problems or big difference, people would get 
together and find a way to get along, they’d drop their anger, why?, because before we were kids and 
we were all in the streets playing soccer for example and you laid out stones to make a goal, the girls 
used to play with their dolls with the neighbors and parents got together on porch swings and now 
none of this happens. Everything is empty and at 10PM it’s as if it was 2AM.” (Monterrey) 
Impact Evaluation Report– Mexico 
115
 “Comparada a otros años si se ha acentuado mucho…Yo digo que se llega al extremo de ya estar 
viviendo en una situación de terrorismo, porque mientras antes se te presentaban con una navaja o 
con una pistola ahora te amenazan vía telefónica o simplemente personalmente, yo algunos 
compañeros les platiqué porque traía la situación de impotencia, yo sufrí un robo en el cual no se me 
presentó ninguna arma, simplemente me para un Federal de Caminos en la carretera, me detiene 
pero yo vi a dos a un kilómetro de distancia había visto que estaba una camioneta que al minuto que 
está el oficial Federal abordándome de la documentación de la camioneta en una bajada donde te 
excedes de la velocidad que también hay un letrero ahí después lo vi de 110 Km por hora, entonces 
llega la camioneta y el de la camioneta y me dice si valoras tu vida súbete a la patrulla y en el otro 
asiento dejas todo el dinero que traes si valoras tu vida. En la patrulla y con el silencio del Federal 
de Caminos, el Federal de Caminos con su pistola de radar abordándome bien él pero era una 
actuación prácticamente porque el otro ya estaba en comunicación con él, estaban coludidos.” 
(Monterrey) 
“Compared to recent years, it’s gotten worse… I am telling you that it’s going to an extreme… 
people are living in a terrorism situation because while before they just had a knife or a gun but now 
they threaten you over the phone or simply in person, I told some of my friends because I was in a 
powerless position, I was robbed and the guy didn’t have a weapon, I get pulled over by the Highway 
Police, he detains me but I saw two people about a kilometer away, I had seen that they were in a 
truck and one minute later the Highway Police officer was there asking me for the papers for my 
truck on a hill that makes you go really fast and there is also a sign there that I saw that said 110 
kilometers per hour, then the other truck shows up and the guy in the truck ‘If you value your life, get 
into the police car and in the other seat, leave all the money you have, if you value your life.” I get in 
the police car and the Highway Patrol officer just sits there in silence with his radar gun pointed at 
me, but it was just an act practically because the other one had already been in touch with him… they 
were in cahoots.” (Monterrey) 
Beyond this impact evaluation of the programs in general, we are limited by the scope of the 
information and we cannot go deeper to give clues on exactly what types of programs were most 
effective. This is because the interviews at the neighborhood level (level or area of influence of 
specific programs) are not sufficient to perform statistical inferences with the level of reliability 
required at the city level. 
We believe that in general, in the case of Mexico, the results are weak and modest. However, for 
a significant number of indicators the results are positive regarding the direction of the impact, 
but unfortunately (still) not very encouraging in terms of the magnitude of the impact. Unlike 
CARSI’s impact evaluations in four Central American countries, where the results were more 
favorable, we believe that there are four possible reasons that may explain the results obtained so 
far for the evaluation of Mexico. These are: 
a) Design Reasons: As mentioned in the earlier sections of this report, where we describe in 
detail the methodology used, we see that Mexico has two major variations from most 
CARSI countries that have to do with the selection of the neighborhoods we studied. The 
first has to do with the fact that the treatment neighborhoods were selected for 
convenience and according to the criteria of the Living Better program. The second has to 
do with the selection procedure of control neighborhoods, that even though it was 
rigorous, extensive and in-depth vis-a-vis its and comparison (i.e. matching) of socio-
Impact Evaluation Report - Mexico 
116
economic indicators, it lacked the ability to include additional comparisons of indicators 
and specific data on crime and violence. 
b) Reasons of Time: Here we refer to the relatively short time since the programs were 
implemented (see Table 8) which may imply that many of the effects of programs that were 
measured through individual questionnaires have not yet materialized. It is interesting to see that 
the indicators on which the treatment had a favorable impact, have more to do with effects that 
we anticipated would happen at the early stages of implementation of the programs, such as the 
ones on communication and knowledge of where respondents identify that there are institutions 
working for crime prevention or supporting youth in their community. After-effects, such as 
reduction of fear or crime are captured but not yet at the expected level. 
c) Special Characteristics of the Neighborhoods in Mexico: There are important differences with 
regard to the CARSI evaluations, which could explain differences in results in terms of their 
magnitude; there are two main differences. The first has to do with the fact that the 
neighborhoods in Mexico are much larger in terms of number of households and number of 
square kilometers in cities and of a much greater magnitude and complexity potentially than the 
areas included in Central America. This may explain the fact that the effects of the programs are 
diluted or have minor effects at the level of “the whole community.” The second difference has 
to do with the different situations, in terms of crime and violence, the three cities were 
experiencing during the baseline round and in the second round. For example, Monterrey was 
going through one of its worst moments during the baseline and for the second round, and after 
an intense presence of the Mexican army, experienced, in comparative terms, a time of much 
greater security. It seems that Tijuana experienced apparently minor changes, but in the opposite 
direction. 
d) Reasons of Neighborhood Make-Up: Finally, we think that it is possible, since no indicators 
on crime and violence at the neighborhood level which would have allowed a more 
comprehensive selection of control neighborhoods exist, that elasticity of change in perceptions, 
i.e. the sensitivity of these to similar changes (such as context), had a greater impact on the 
control neighborhoods than in the treatment neighborhoods and therefore the captured effects 
have been lower than expected. This would mean that individuals in the control neighborhoods 
showed more marked changes than the treatment neighborhoods, in situations of adversity or 
improvement. In other words, their responses would be more elastic and the change between the 
baseline and the post-treatment rounds could capture a change that is much more pronounced 
among individuals in the treatment neighborhood that supposedly would be more accustomed to 
extreme situations (especially adverse ones). 
 
Impact Evaluation Report– Mexico 
117
References 
Aos, S., M. Millar and E. Drake, (2006): Evidence-based public policy options to reduce future 
prison construction, criminal justice costs, and crime rates, Washington State Institute for 
Public Policy, Olympia, WA. 
Azpuru, D., S. Finkel, et al. (2008). "Trends in Democracy Assistance: What Has the U.S. Been 
Doing?" Journal of Democracy 19: 150-159. 
Bamberger M., Rao V. & Woolcock M. (2010). Using Mixed Methods in Monitoringand 
Evaluation. Experiences from International Development. Washington, DC: World Bank. 
http://www￾wds.worldbank.org/servlet/WDSContentServer/WDSP/IB/2010/03/23/000158349_ 
20100323100628/Rendered/PDF/WPS5245.pdf 
Brown, B. B., D. D. Perkins, et al. (2004). "Incivilities, Place Attachment and Crime: Block and 
Individual Effects." Journal of Environmental Psychology 24(3): 359-371. 
Casas-Zamora, K. (2013). The Besieged Polis: Citizen Insecurity and Democracy in Latin 
America. Washington, DC.: Latin America Initiative at Brookings and Organization of 
American States. 
Centro de Investigación de la Opinión Pública Salvadoreña (CIOPS). (2013). Opinión de los 
Salvadoreños sobre el desempeño de la Corte Suprema de Justicia, debate entre 
candidatos, campaña electoral y preferencias electorales 2014. Retrieved from 
http://www.utec.edu.sv/utec_files/utec_publicaciones/investigaciones_ciops/agosto2013/
1/files/1.%20boletin%20encuesta%2058%20%28primera%20version%29.pdf. Last 
accesed 28 October, 2013. 
CEPAL (Comisión Económica para América Latina el Caribe, Economic Commission for Latin 
America) (2011). Gasto Público en Seguridad y Justicia en Centroamérica (2006–2011). 
Santiago, Chile: CEPAL 
Córdova Guillén, A. B. (2008). Divided We Failed: Economic Inequality, Social Mistrust, and 
Political Instability in Latin American Democracies. Political Science (Ph.D. 
Dissertation). Nashville, TN, Vanderbilt University. Doctor of Philosophy: 191. 
Finkel, S. E., A. Pérez-Liñán, et al. (2007). "The Effects of U.S. Foreign Assistance on 
Democracy Building, 1990-2003." World Politics 59: 404-439. 
 
Finkel, S. E., A. Pérez-Liñán, et al. (2008). Deepening Our Understanding of the Effects of US 
Foreign Assistance on Democracy Building: Final Report. Washington, D. C., United 
States Agency for International Development: 110. 
 
Impact Evaluation Report - Mexico 
118
Franklin, T. W., C. A. Franklin, et al. (2008). "A Multilevel Analysis of the Vulnerability, 
Disorder, and Social Integration Models of Fear of Crime." Social Justice Research 
21(2): 204-227. 
Freedom House (2013). El Salvador. Retrieved from 
http://www.freedomhouse.org/report/freedom-world/2013/el-salvador. Last accessed 28 
October 2013. 
Hawdon, J. E., J. Ryan, et al. (2003). "Policing Tactics and Perceptions of Police Legitimacy." 
Police Quarterly 6 (4): 469-491. 
 
Hayes, R. J. and L. H. Moulton (2009). Cluster Randomised Trials. New York, Chapman & 
Hall/CRC. 
 
IDB (Inter-American Development Bank). (2012). Citizen Security: Conceptual Framework 
andEmpirical Evidence. Institutions for Development Discussion Paper No. IDB-DP 
232. Washington, DC: Inter-American Development Bank 
Kelling, G. L. and C. M. Coles (1996). Fixing Broken Windows : Restoring Order and Reducing 
Crime in our Communities. New York, Martin Kessler Books. 
 
Kish, L. (1995). Survey Sampling. New York, J. Wiley. 
 
Miller, L. L. and E. Silver (2004). "Sources of informal social control in Chicago 
neighborhoods." Criminology 42(3): 551-583. 
 
Moser, C. and B. Van Bronkhorst (1999). Youth Violence in Latin America and the Caribbean: 
Costs, Causes, and Interventions. Sustainable Development , Working Paper No. 3. 
Urban Peace Program Series. The World Bank. 
 
Moser, C., A. Winton, and A. Moser (2005). Violence, Fear, and Insecurity among the Urban 
Poor in Latin America. The Urban Poor in Latin America. M. Fay. Washington, DC, The 
World Bank. 
 
Moser, C. O. N. and C. McIlwaine (2006). "Latin American Urban Violence as a Development 
Concern: Towards a Framework for Violence Reduction." World Development 34(1): 89-
112. 
 
Murray, D. M. (1998). Design and Analysis of Group-Randomized Trials. New York, Oxford 
University Press. 
 
National Research Council (2008). Improving Democracy Assistance: Building Knowledge 
Through Evaluations and Research. Washington, D. C., The National Academies Press. 
 
Impact Evaluation Report– Mexico 
119
Perkins, D. D. and R. B. Taylor (1996). "Ecological Assessments of Community Disorder: Their 
Relationship to Fear of Crime and Theoretical Implications." American Journal of 
Community Psychology 24(1): 63-107. 
 
Perkins, D. D., A. Wandersman, et al. (1993). "The Physical Environment of Street Crime: 
Defensible Space, Territoriality and Incivilities." Journal of Environmental Psychology 
13(1): 29-49 
Programa de las Naciones Unidas para el Desarrollo (PNUD), (2005). Cuánto le cuesta la 
violencia a El Salvador? San Salvador: PNUD. 
Puddy, R. W. & Wilkins, N. (2011).Understanding Evidence Part 1: Best Available Research 
Evidence. A Guide to the Continuum of Evidence of Effectiveness. Atlanta, GA: Centers 
for Disease Control and Prevention 
http://www.cdc.gov/ViolencePrevention/pdf/Understanding_Evidence-a.pdf 
Renauer, B. C. (2007). "Is Neighborhood Policing Related to Informal Social Control?" 
Policing: An International Journal of Police Strategies & Management 30(1). 
 
Robinson, J. B., B. A. Lawton, et al. (2003). "Multilevel Longitudinal Impacts of Incivilities: 
Fear of crime, Expected Safety, and Block Satisfaction." Journal of Quantitative 
Criminology 19(3): 237-274. 
Sampson, R. J. and S. W. Raudenbush (1999). "Systematic Social Observation of Public Spaces: 
A New Look at Disorder in Urban Neighborhoods." The American Journal of Sociology 
105(3): 603-651. 
 
Sampson, R. J., S. W. Raudenbush, et al. (1997). "Neighborhoods and Violent Crime: A 
Multilevel Study of Collective Efficacy." Science 277: 918-924. 
 
Schafer, J. A., B. M. Huebner, et al. (2003). "Citizen Perceptions of Police Services: Race, 
Neighborhood Context, and Community Policing." Police Quarterly 6(4): 440-468. 
 
Scheider, M. C., T. Rowell, et al. (2003). "The Impact of Citizen Perceptions of Community 
Policing on Fear of Crime: Findings from Twelve Cities." Police Quarterly 6(4): 363-
386. 
 
Seligson, M. A., S. E. Finkel, et al. (2009). Exporting Democracy: Does it Work? Exporting 
Democracy. Z. Barnay and R. E. Moser. Cambridge, Cambridge University Press. 
Sherman, L.; Gottfredson, D.; MacKenzie, D.; Eck, J.; Reuter, P.; Bushway, S. (1998). 
Preventing Crime: What Works, What Doesn't, What's Promising. Report to the U.S. 
Congress. Washington, D.C.: U.S. Dept. of Justice. https://www.ncjrs.gov/works/ 
Impact Evaluation Report - Mexico 
120
Sherman, L. (2012): Desarrollo y evaluación de programas de seguridad ciudadana en América 
Latina. Protocolo para la prevención del delito a partir de la evidencia. Washington, DC: 
Banco Interamericano de Desarrollo. 
http://idbdocs.iadb.org/wsdocs/getdocument.aspx?docnum=37024856 
Skogan, W. G. (1990). Disorder and Decline: Crime and the Spiral Decay in American 
Neighborhoods. New York, Free Press. 
Spybrook, J., S. Raudenbush, X. Liu and R. Congdon (2008). Optimal Design for Longitudinal 
and Multilevel Research: Documentation for the Optimal Design Software 
Villarreal, A. and B. F. A. Silva (2006). "Social Cohesion, Criminal Victimization and Perceived 
Risk of Crime in Brazilian Neighborhoods." Social Forces 84(3): 1725-1753. 
 
Warner, B. D. (2007). "Directly Intervene or Call the Authorities? A Study of Forms of 
Neighborhood Social Control Within a Social Disorganization Framework." Criminology 
45(1): 99-129. 
 
WHO (2002). World Report on Violence and Health, World Health Organization. 
 
WOLA (2008). Daring to Care: Community-Based Responses to Youth Gang in Central 
America and Central American Immigrant Communities in the United States. 
Washington, DC, Washington Office on Latin America. 
Impact Evaluation Report– Mexico 
121
Appendices 
Appendix A: Study Design: A Multi-Site “Cluster Randomized Experiment” (CRE) 
According to a major study undertaken by USAID for the National Academy of 
Sciences, and on which one of the authors of this LAPOP study (Mitchell Seligson) 
participated, the goal of a scientifically rigorous evaluation is to test the counterfactual; 
to be able to determine “not merely what happened after a given assistance program, 
but how much of what happened differs from what would be observed in the absence of 
this program” (National Research Council 2008: 44). As recommended by that study, a 
credible evaluation must contain certain features, all of which are part of the LAPOP 
evaluation design. The main design features of LAPOP’s research design are: 
(1) Collection of data in selected units before and during the implementation of the 
programs; specifically, a baseline study, and two mid-point measurements. Initially, 
LAPOP proposed a post-treatment measurement. However, due to changes in the 
program implementation schedule, in the case of Mexico, only one measurement was 
made during the study and a post-program was not feasible; Thus, the present impact 
evaluation consists of two rounds of surveys (baseline and mid-term). 
(2) The random assignment of units to either a “treatment” or “control” group [i.e. units 
that are to be treated by the program and units that in all other aspects are eligible for 
but do not participate in or benefit from the program]. In the case of Mexico, the 
assignment of treatment units was not random and that is why we refer to a quasi￾experimental design. 
The design used in this study is technically known as a “Cluster Randomized 
Experiment” (CRE), or “Multi-Site Cluster Randomized Experiments” when carried 
out in blocks composed of clusters (in our case, neighborhoods) that are very similar 
(Spybrook, Raudenbush, Liu and Congdon 2008). This type of experimental design is 
widely considered to be among the most reliable and accurate methods for establishing 
cause-and-effect (Puddy and Wilkins 2011; Sherman 2012). The research design used 
by LAPOP for the CARSI evaluation scores a 5 (the highest level) on the Maryland 
Scale of Accuracy in the Design of Evaluation, indicating that the findings yielded by 
the study should be considered strong evidence for a cause-and-effect relationship (or 
non-relationship; Sherman 2012). As described by Murray (1998), a CRE has four 
distinguishing characteristics (note that in this study, the block or site corresponded to a 
country or municipality, and the cluster to a community): 
a) “The unit of assignment is an identifiable group rather than an individual. Such 
groups are not formed at random, but rather through some connection among their 
members.” This design was appropriate for the CARSI and Merida projects, since the 
program was based on the community or neighborhood as the unit of analysis in each 
site or country. 
Impact Evaluation Report - Mexico 
122
b) Different units of assignment are allocated [randomly] to each study condition [either to a 
treatment or control group].” 
c) “Units of observation are members of the groups that serve as the units of assignment.” 
Thus, this type of experiment has a multilevel structure. In this case, the cities or 
countries are the highest aggregate level, units of assignment [polígonos/neighborhoods] 
formed an intermediate level, and individuals were the lowest aggregate level. 
d) CREs “typically involve only a limited number of assignment units in each study 
condition.” 
The CRE design is an especially appropriate research design for community-based programs. 
The group-based unit of assignment corresponds to the neighborhood grouping of individuals. 
Although the interventions may target certain individuals within communities, (for instance, at￾risk youth), random assignment of individuals to treatment and control groups is at times not 
effective since it is difficult to avoid interaction between individuals across study conditions (i.e., 
in any given community, at-risk youth in the treatment group will most likely interact with those 
in the control group). To avoid this problem of contamination, the CRE design uses non￾contiguous clusters (e.g., neighborhoods or communities) that are then randomly assigned to 
each study condition (i.e., treated/non-treated). The literature regards CRE as the “best 
comparative design available” when an intervention “operates at a [cluster] level, [and] 
manipulates the social or physical environment” (Murray 1998). 
The unit of assignment in the study was the “neighborhood or community” [colonia, or 
comunidad]. The units of observation corresponded to “individuals” within neighborhoods, 
including those who were likely to be direct beneficiaries of the programs, those who benefited 
only indirectly, and those who did not participate in the programs’ activities. 
In this type of trial, the multi-level structure of the experiment determines its design and the 
methods used to analyze the data; given the multi-level nature of CRE, the data analysis is more 
complex than in Randomized Experiments (RE). Standard errors must be adjusted to account for 
the study’s “design effect” (Kish 1995; Hayes and Moulton 2009). 
In a study that includes a large number of assignment units (thousands of neighborhoods), 
randomization ensures that units assigned to treatment and control groups will, on average, be 
very similar before the intervention. This is a result of the well-known “law of large numbers,” in 
which a coin-toss will approach 50% heads and 50% tails as the number of tosses approaches 
infinity. For example, in a vaccine trial, there could be thousands of individuals who receive the 
vaccine and a similar group that does not. In large groups like this, the pre-vaccine immune 
levels should be very similar in both groups. However, in experiments involving a relatively 
small number of assignment units (as in the case of CARSI and the CRE methodology more 
generally), the number of treatment areas is relatively limited, if for no other reason than the 
units are not individuals but neighborhoods in which we are sampling individuals. It is not 
feasible to fund an impact evaluation of thousands of communities as the costs would be far too 
high for the limited resources available. 
Impact Evaluation Report– Mexico 
123
Differences in the averages for any given variable in the starting level of the treatment and 
control communities were expected for this study, and in fact that is what was found for many 
variables. The number of communities was just too small to expect the means of treatment and 
control to have converged. Even after 100 coin tosses, there is only about an 8% chance of 
getting a 50/50 split, although getting to a 60/40 split is quite likely. Therefore, for studies such 
as the one LAPOP undertook, in which the number of neighborhoods is relatively small, other 
strategies are implemented in addition to randomization to minimize baseline differences 
between treatment and control groups and increase the internal validity of the study. When the 
number of assignment units is small, a “stratified” design can be used to help minimize 
differences between treatment and control communities. In a stratified design, units from 
relatively homogenous sub-groups (i.e., groups that share similar characteristics) are randomly 
assigned to either the treatment or control. In the specific case of this evaluation where 
assignment of neighborhoods was not random, similarity between groups was sought by pre￾visits to assign control neighborhoods via the matching method comparing aggregate socio￾economic indicators. The matching method is described below and it consists of the comparison 
of economic indicators as well as of a direct observation of the physical conditions of the 
neighborhoods. 
Selection of Study Areas 
The LAPOP’s Merida-CARSI impact evaluation includes four Central American countries 
(Guatemala, El Salvador, Honduras, and Panama) and Mexico. In most Central American 
countries, the selection of treatment and control neighborhoods is done randomly. In the case of 
Mexico, it was not the case because neighborhood treatments were chosen for “convenience” and 
according to the strategy of the federal government’s Live Better program that we describe later. 
Below is a map that locates the three northern cities included in this evaluation and a map for 
each city showing neighborhoods (control, treatment and not at-risk) selected in each one of 
them. 
Regarding the selection of control neighborhoods, we can say that USAID gave us reference 
cards with a detailed description of the sociodemographic conditions of each treatment 
neighborhoods in Tijuana, Ciudad Juárez, and Monterrey. From this information, LAPOP 
implemented a review process in the municipalities and communities of interest ensuring that 
those communities had similar socio-demographic conditions. The information for the selection 
of communities came from the General Population Census of 2010, designed, conducted and 
managed by the National Institute of Statistics, Geography and Informatics (INEGI) in Mexico. 
Then we visited seven control communities across all three cities to select three of them; four 
communities were also considered of no risk (and were shortlisted for socio-demographic 
information available via the 2010 Census). We selected two in Tijuana, two in Ciudad Juárez, 
and one in Monterrey. In the following section we provide further details on the community 
selection process. 
Impact Evaluation Report - Mexico 
124
Map 1: Geographic Location of the Three Cities Included in the Impact Evaluation 
Map 2: Geographic Location of the Treatment, Control, and No-Risk Neighborhoods in Ciudad Juárez 
Impact Evaluation Report– Mexico 
125
Map 3: Geographic Location of the Treatment, Control and No-Risk Neighborhoods in Monterrey 
Map 4: Geographic Location of the Treatment, Control and No-Risk Neighborhoods in Tijuana 
Stages of the Study and Types of Data Collected by LAPOP 
Impact Evaluation Report - Mexico 
126
Once the list of control communities was determined (based on pre-visits and looking to keep it 
as close as possible to the treatment neighborhoods), the process of field work began. This was 
also a multistage effort, as detailed below.
Quantitative Data: 
(1) Baseline Survey 
The baseline data characterized the population of study in both treatment and control groups in 
order to assess the starting condition of each, since the DID method looks at changes from the 
baseline. The baseline survey helped to draw causal inferences about the effect of the 
interventions because it made it possible to compare groups before and after the intervention. 
Phase I: Pre-Visits to the Neighborhoods
In each of the three cities included in the research, it was necessary to select three control 
neighborhoods that did not share a border or that were too close to the treatment group. At the 
sample time, the communities should have similar characteristics to those in the treatment group. 
For selection, we carried out two complementary procedures. The first was clerical work that 
consisted of pre-selection of geographical areas of similar size and characteristics for each 
neighborhood being processed. These areas are the neighborhoods that are candidates to be 
selected as control communities. 
For each of the treatment neighborhoods, we selected based at least two similar geographical 
areas based on 45 economic indicators. Illustrated below is a list of the main indicators: 
 Total Number of People; People Between 0 and 11 Years of Age 
 People Under 15; People Between 5-15 Years of Age; People Between 13-15 Years of 
Age 
 People Between 16-18 Years of Age; People 65 and Older 
 People 5-15 Years of Age Not Attending School 
 People 13-15 Years of Age Not Attending School 
 People 15 and Older Who Cannot Read or Write 
 People 15 and Older Who Have Not Completed Primary School 
 Households Without Access to Social Security 
 Total Number of Households 
 Households Headed By Women; Households With Dirt Floors 
 Proportion of Households Without Electricity 
 Proportion of Households Without Running Water in the House 
 Proportion of Households With No Bathroom or Toilet 
 Proportion of Households Without a Drain or Sewer 
 Proportion of Households Without TV 
 Proportion of Households Without a Refrigerator 
 Proportion of Households Without a Washing Machine 
 Proportion of Households Without a Computers 
Impact Evaluation Report– Mexico 
127
 Proportion of Households Without Internet 
For the selection of candidates in the non-risk neighborhoods, we followed a similar pre￾selection process the only difference being that we aimed to locate geographical areas whose 
socioeconomic indicators were comparatively better than in the treatment/control groups. The 
result of this procedure was the pre-selection of 7 candidates in the control group and 4 non-risk 
candidates in each of the cities for a total of 33 potential neighborhoods (21 control and 12 non￾risk). 
Once we selected these 33 potential neighborhoods, a second method of selection which 
consisted of a personal pre-visit to each of them by means of an observation to compare their 
physical characteristics against treatment neighborhoods was performed. In each city, we first 
visited the treatment neighborhoods and then the control and non-risk neighborhoods, making a 
total of 42 personal pre-visits; nine to the treatment neighborhoods and 33 to the candidates 
described above. All pre-visits were made between July and August 2012 by a team of 
professional interviewers. 
In order to systematize the observations of the pre-visits and standardize the procedures we 
captured data at each site by following a questionnaire designed to capture the general conditions 
of the areas visited as well as the team's observations in the field. Additional data obtained 
during the pre-visit was the assessment of the existence of minimum conditions of access and 
security for the field team so they could perform the subsequent components of the quantitative 
research. 
Once the 42 pre-visits were completed, we proceeded to process and evaluate the information 
collected on them. With the input of information that we have described here so far, we 
proceeded to make recommendations of the control and non-risk neighborhoods that could be 
ideal for the impact evaluation. The proposal, with due justification, was validated by our 
counterparts in USAID. After this important part of the quantitative research, we proceeded to 
prepare the implementation of the three components described below. 
Phase II: Surveys to Households 
Once we had selected the 23 neighborhoods, (nine treatment, nine control, and five non-risk) in 
which the research would be carried out, we proceeded to gather information on three levels 
ranging from the general to the particular, which together make up a comprehensive diagnosis of 
the relevant variables of the investigation. The three levels are: context, household, and 
individual. In this section, we will discuss the methodology used to collect information on 
households and the following sections will cover the other two levels. 
In each of the 23 neighborhoods selected for research, we draw a random sample of households. 
Two criteria for the selection of the sample were defined: (a) having a theoretical margin of error 
in each neighborhood of +/- 5% with a statistical confidence of 95% and (b) a dispersed sample 
in 20 sampling points per neighborhood. The sampling frame used to draw the sample were the 
basic geo-statistic areas from the INEGI also known as the 2010 AGEB Census. 
Impact Evaluation Report - Mexico 
128
Sampling was done in two stages. In the first stage, 20 blocks were randomly selected using a 
systematic random probability method based on the number of households on each block, where 
the blocks with more households are more likely to be selected than blocks with smaller number 
of households, but where all the blocks have a probability different from zero of being selected. 
Such a process is known as probability proportional to size or PPS. 
The second stage consisted of a random selection of households within selected blocks in the 
first stage. At each point, we pre-defined a number of households to be interviewed, which was 
conducted through a pre-established systematic jump starting in a randomly selected household. 
Interviewers walked the block clockwise, selecting homes with this procedure until all interviews 
were completed. All households had a probability different from zero of being selected. 
Within each selected household, we interviewed a member of the household who was 18 years of 
age or older and who could provide information on all household members. Ideally, we were 
looking to speak to the head of the household. Before starting the interview, a letter of informed 
consent where each adult respondent gave us their agreement to participate in the interview was 
read. 
We scheduled a total of 8,560 interviews: 2,760 in Juarez, Tijuana 2,940, and 2,860 in 
Monterrey. It was not possible to complete all the interviews. These reasons range from rejecting 
participation in the study to sending the interviewers to unsafe places. In total, 7,866 valid 
interviews, enough to make estimates and to obtain baseline results, were completed. The 
interviews were conducted between August and September 2012. It was necessary to replace a 
non-risk neighborhood and suspend the individual survey in a treatment neighborhood on 
Monterrey due to insecurity. We pre-selected other neighborhoods in Tijuana and Ciudad Juárez, 
but it was not necessary to replace neighborhoods in these two cities. 
Phase III. Systematic Observations 
Systematic observations constitute a standardized register of physical conditions of the state of 
the streets where the household surveys were applied. As the name implies, they are direct 
observations of a catalog of relevant aspects of the social context and physical characteristics. In 
order to accomplish this, we logged the information for each block for each of the blocks 
included in the sample in an observation questionnaire where the interviewers’ filed team record 
their observations (Appendix B). Among other things, the questionnaire records 20 aspects of the 
physical appearance of the block, the presence of 16 different resources/infrastructure and the 
presence of seven different types of groups of people at the time of observation, as well as 
additional general comments that the interviewer could record. All observations were conducted 
between August and September 2012. 
Figure 63: Example of a Photographic Registry of Systemic Observations 
Impact Evaluation Report– Mexico 
129
The selected neighborhoods represent large-sized populations of more than 1,000 households 
distributed in a considerably large geographic area. Therefore, it was necessary to follow a 
scientific selection process for taking a representative sample based on technical criteria. The 
sampling design was proposed to be representative of all households that made up each 
neighborhood. The description of the sampling procedure is described in Phase II above. For this 
section, is important to note that for household selected in the sample, we conducted a systematic 
observation of the street. 
If two or more households in the sample were on the same street, only one systematic 
observation was conducted. In the data set, each household has associated information about the 
context generated from the registry of systematic observations that we have described here so 
far. 
Additionally, we generated a visual record for most of the streets of at least one photograph per 
street where the systematic observations were carried out. In some neighborhoods, certain risk 
conditions were detected and as such, the team of supervisors decided not to take photographs. 
Further down, we show an example of this photographic archive for one of the sampling point a 
treatment neighborhood in Tijuana where we show the name and photograph of the streets where 
the interviews were conducted. 
Impact Evaluation Report - Mexico 
130
Phase V: Surveys to Individuals 
The survey of individuals is the fourth product of quantitative research conducted during the 
baseline and constitutes the most disaggregate information gathered by capturing the opinions, 
perceptions, preferences, attitudes, and behaviors at the individual level. It is at this level that we 
expected the programs implemented in the treatment neighborhoods to have a greater impact. 
To select the individual who participated in the individual survey, we followed the sample frame 
from the previous research stage. As mentioned above, in the household survey, we recorded 
data of all members of the households. With the information from the adults, we proceeded to 
make a quota sampling within each of the 20 sample points of each neighborhood to interview 
women / men of all ages according to the proportional incidence of them on every block. For 
example, assume that in a neighborhood, we conducted 300 household interviews and in each of 
them, there are four adults. This would yield a total of 1,200 adults in these households. That 
number of adults would be the sampling frame for that particular neighborhood. For each of the 
20 blocks of the neighborhood in this example, we know how many are women and how many 
are men and we know the age of each of them and this information allows us to determine for 
each block of the respondent profile (by gender and age) that we carry out in accordance to the 
proportion / weight of that block out of the total number of blocks. 
For each neighborhood, we scheduled 300 individual interviews for a total of 6,600 interviews 
using the individual questionnaire: 2,400 in Ciudad Juárez, 2,400 in Tijuana, and 1,800 in 
Monterrey. In the latter, we had to suspend the fieldwork at the treatment neighborhood of 
Independencia (Independence) due to the very high crime that the team faced in that area. In 
some cases, it was not possible to complete the prescribed number of interviews for various 
reasons which were documented for each neighborhood and reported to USAID in due course.12
In total, we completed 5,963 valid interviews, enough to make estimates (at the city level and by 
type of neighborhood) and to obtain baseline results. All interviews were conducted between 
September and November 2012. 
In each household, the same questionnaire was applied to one adult member of the household. 
The questionnaire has multiple questions the following main topics: quality of life, social capital, 
community roots, social cohesion, community participation, empowerment, individual safety, 
contextual security, crime prevention, at-risk youth, victimization, corruption, discrimination, 
and political culture. The next section presents the key results of each of these issues. 
 
12 The reports have to do mainly with the risky and unsafe conditions that the teams of interviewers faced while 
carrying out their work as well as the reaction of people in the neighborhoods (sometimes supporting them and 
alerting the team) and other extraordinary events that the teams observed. 
Impact Evaluation Report– Mexico 
131
Second Round Collection 2014 
In the second round of information gathering in 2014 in the communities of Ciudad Juárez, 
Monterrey, and Tijuana, a total of 22 neighborhoods were visited: five non-risk, nine control, and 
eight treatment. The sample was selected in the same blocks and households that were visited in 
the baseline in 2012 (hence the sampling frame and design are the same as the baseline), to 
complete the planned sample, we visited some homes and blocks different from the ones on the 
baseline but always within the limits of each of the originally selected neighborhoods. 
The original design planned the completion of 6,600 interviews; considering the non-response 
rates and times of operation of the project, the final number of interviews was 6,113. As 
described in previous sections, the number of selected communities and the sampling strategy 
was designed by LAPOP. 
The field work conducted from March 25th to May 28th, 2014. Table 6 shows the distribution of 
the 2014 sample. 
Table 6. Sample Distribution in Mexico for the 2014 Collection 
Non-Risk Treatment Control Total 
Number of Communities 5 8 9 22 
Individual Interviews 1,456 2,060 2,597 6,113 
Systematic Observations 229 334 407 970 
Qualitative Data 
Now we describe the scope and general methodology for qualitative phase. 
+ Focus Group Composition 
The focus groups were conducted in the cities of Tijuana, Ciudad Juárez, and Monterrey from 
November 19th – 26th, 2012. The sessions were led by a moderator and recorded only on audio 
requesting the anonymity of the participants. The recruitment was carried out by three teams, 
each of them responsible for a different city. Participants were invited to speak to groups about 
insecurity and how it has impacted their personal, family, community and social environment. 
Four group profiles were carried out in each city: 
• Group 1 Youth at Low-Risk 
• Group 2 Youth at High-Risk 
• Group 3 Community Leaders 
• Group 4 Public Officials 
Impact Evaluation Report - Mexico 
132
The following briefly describes the characteristics of each of the profiles: 
Group 1 was composed of young people who have not had anything to do with the police or have 
ever committed crimes. The age of participants was 18-25. They were made up of students, 
workers, or both. This group was mostly composed of young people of the same socioeconomic 
status as the treatment and control neighborhoods in the quantitative phase. 
Group 2 was composed of young people who had problems with the police, young people who 
had committed crimes; had been in prison, or in trouble with the public prosecutor; or who had a 
violent background. The age of participants was 18-25 and they were students, workers, or both. 
They were mostly composed of young people of similar socioeconomic status to the level of the 
treatment and control neighborhoods in the quantitative phase. 
Group 3 consisted of parents, small and medium-level businessmen, school teachers, and 
religious leaders. The ages of the participants were 26 to 50. They were composed of mostly 
people who lived in similar areas to those in the treatment and control groups in the quantitative 
stage. 
Group 4 consisted of lawyers, school principals, members of social organizations, teachers, small 
and medium-sized entrepreneurs, community leaders, and religious leaders. Composed mostly of 
people from a socio-economic level of D + / C according to the Mexican Association of 
Research Agencies (AMAI). 
Data Collection Sites Mexico 
The information from the communities studied was collected in three cities in northern Mexico; 
these were Ciudad Juárez, Tijuana, and Monterrey. In these three urban areas, we worked in non￾risk, treatment and control communities as shown in Table 7. 
Table 7. Sample Distribution in Mexico by City for the 2014 Data Collection 
City Non-Risk Treatment Control Total 
Ciudad Juárez 581 611 895 2,087 
Monterrey 305 583 883 1,771 
Tijuana 577 865 813 2,255 
Impact Evaluation Report– Mexico 
133
In Table 8, we show general statistics that describe the conditions of households in each of the 
cities included in this impact evaluation. 
Table 8. Statistics on Household Conditions in the Three Cities Where 
Research was Carried Out in Mexico 
Tijuana Juárez Monterrey 
Average of occupants in private dwellings, 2010 3.7 3.7 3.8 
Households without dirt floors, % 94.4 91.2 94.2 
Households that have running water, % 94.9 91.2 95.4 
Households that have sewage/drainage systems % 96.1 92.1 95.8 
Households that have a toilet or a bathroom, % 97.6 92.5 96.0 
Households with electricity, % 98.0 93.0 96.5 
Households that have a refrigerator, % 91.6 89.5 92.8 
Households that have a television % 96.7 91.8 95.0 
Households that have a washing machine % 76.3 77.7 84.0 
Households that have a computer, % 44.2 32.7 44.3 
Considering that the incidence and vulnerability to crime is higher in urban areas (as has been 
shown by the AmericasBarometer surveys), the impact evaluation in Mexico was limited to at￾risk areas (and non-risk) in urban areas. 
Impact Evaluation Report - Mexico 
134
Appendix B: Individual Questionnaire 
LAPOP estudio de comunidades Core Spanish Version # FINAL 
Mexico IRB Approval: 110627 
 
Estudio de comunidades: MEXICO 2014 
© Vanderbilt University 2014. Derechos reservados. All rights reserved. 
PAIS. 
IDNUM. Número de cuestionario [asignado en la oficina]__________________
PROV. Ciudad:
(1) Cd. Juárez (2) Monterrey (3) Tijuana 
MUNICIPIO. Nombre Municipio:
(1) Cd. Juárez 
(2) Monterrey 
 (4) Tijuana 
NOMBAR. Nombre del Barrio/ 
Colonia/Comunidad____________________________________________________
BLOQUE. Número de Polígono 
Manzana. Número de 
manzana_______________________________________________________ 
CUADRA. Número de cuadra dentro de la manzana o 
pasaje__________________________________ 
NUMVIVI.NÚMERO DE LA VIVIENDA DENTRO DE LA MANZANA:______ 
NUMHOGAR.NÚMERO DE ORDEN DEL HOGAR DENTRO DE LA 
VIVIENDA:__________
HORA. Hora de inicio: _____:_____AM PM 
 ------------ 
ATENCIÓN: ES UN REQUISITO LEER SIEMPRE LA HOJA DE CONSENTIMIENTO
 INFORMADO ANTES DE COMENZAR 
Q1. Género [ANOTAR, NO PREGUNTE]: (1) Hombre (2) Mujer
Impact Evaluation Report– Mexico 
135
Ahora para cada una de las siguientes asociaciones que le voy a mencionar, quiero que por favor me diga 
si actualmente es usted miembro y si participa en reuniones de estas asociaciones [Para cada una de las 
asociaciones primero preguntar si es miembro y, después, preguntar si participa en reuniones de 
esa asociación aun cuando NO sea miembro] 
A. ¿Es usted 
miembro? 
B. y diría usted que participa en reuniones ______ 
[LEER ALTERNATIVAS]
 S
í 
N
o 
N
S 
N
R 
Una vez 
 a la 
semana 
Una o 
dos 
veces 
al mes 
Una o 
dos 
veces 
 al año 
Nunc
a 
NS NR 
CP6N. De un grupo religioso 1 0 88 98 1 2 3 4 88 98 
CP7N. De una asociación 
educativa (de la escuela, 
colegio, instituto, o 
universidad) 
1 0 88 98 1 2 3 4 88 98 
CP8N.De un comité o junta de 
mejoras para la comunidad 1 0 88 98 1 2 3 4 88 98 
CP9N. De una asociación de 
profesionales, comerciantes, 
productores, y/u organizaciones 
campesinas 
1 0 88 98 1 2 3 4 88 98 
CP13N.De un partido político 1 0 88 98 1 2 3 4 88 98 
CP14N.De algún otro grupo o 
asociación voluntaria o 
comunitaria?
1 0 88 98 1 2 3 4 88 98 
[RECOGER TARJETA “A”] 
[RECOGER TARJETA “B”] 
[ENTREGAR TARJETA “C”]
LS3. Para comenzar, ¿en general, qué tan satisfecho(a) está con su vida? ¿Usted diría que se encuentra: 
[LEER ALTERNATIVAS] 
(1) Muy satisfecho(a) (2) Algo satisfecho(a) (3) Algo insatisfecho(a) (4) Muy insatisfecho(a) (88) NS (98) 
NR 
[ENTREGAR TARJETA “A”] 
LS6. En esta tarjeta hay una escalera con gradas numeradas del cero al diez. El cero es la 
grada más baja y representa la peor vida posible para usted. El diez es la grada más alta y 
representa la mejor vida posible para usted.
¿En qué grada de la escalera se siente usted en estos momentos? Por favor, escoja la grada 
que mejor represente su opinión. 
[Señale en la tarjeta el número que representa la “peor vida posible” y el que representa 
“la mejor vida posible”. Indíquele a la persona entrevistada que puede seleccionar un 
número intermedio en la escala].
[ENTREGAR TARJETA “B”]
IT1B1. Ahora, usando la misma escalera, suponga que el cero, la grada más baja, significa hay 
que tener mucho cuidado cuando uno trata con otras personas y el diez, la grada más alta, 
significa que se puede confiar en la mayoría de las personas. ¿Qué piensa usted? Por favor 
escoja la grada que mejor represente su opinión. [Señale el número en la tarjeta que 
representa “se puede confiar en la mayoría de las personas” y el que significa “ hay que 
tener mucho cuidado cuando uno trata con otras personas ”.Indíquele a la persona 
entrevistada que puede seleccionar un número intermedio]
Impact Evaluation Report - Mexico 
136
IT1B2. Ahora usando la misma escalera, suponga que el cero, la grada más baja, significa que la 
mayoría de las personas sólo se preocupan de sí mismas y el diez, la grada más alta, significa 
que la mayoría de las personas tratan de ser serviciales. ¿Qué piensa usted? Por favor escoja 
la grada que mejor represente su opinión. 
[RECOGER TARJETA “C”]
[ENTREGAR TARJETA “D”]
IT1B3. Ahora usando la misma escalera, suponga que el cero, la grada más baja, significa que la 
mayoría de las personas tratarían de aprovecharse de usted y el diez, la grada más alta, 
significa que la mayoría de las personas tratarían de ser justas con usted. ¿Qué piensa usted? 
Por favor escoja la grada que mejor represente su opinión. 
[RECOGER TARJETA “D”] 
A continuación me gustaría hacerle unas preguntas sobre este barrio
IT1. Hablando de la gente de aquí, ¿diría que la gente de este barrio es: [LEER ALTERNATIVAS] 
(1) Muy confiable (2) Algo confiable (3) Poco confiable (4) Nada confiable (88) NS (98) NR 
SOCTIE1. ¿Nació usted aquí en este barrio?
(1) Sí[Pase a SOCTIE5] (2)No[Siga] (88) NS [Pase a SOCTIE5] (98) NR[Pase a SOCTIE5]
SOCTIE2. ¿En qué año se vino a vivir aquí? [Si el entrevistado no recuerda, sondee: me podría decir 
más o menos en qué año] 
Año [Cuatro dígitos] |__|__|__|__| (88) NS (98) NR (99) INAP 
SOCTIE5. Sin incluir a sus familiares, ¿cuántos vecinos adultos conoce usted por su nombre en este 
barrio? Diría usted que…[Leer alternativas] 
(1) A ninguno 
(2) A pocos 
(3) A muchos 
(4) A casi todos 
(5) A todos 
(6) (88) NS (98) NR 
SOCTIE8. En general, diría usted que este barrio es [LEER ALTERNATIVAS] : 
(1) Un lugar muy bueno para vivir 
(2) Un lugar bueno para vivir 
(3) [No leer; espontánea: ni bueno, ni malo] 
(4) Un lugar malo para vivir, o 
(5) Un lugar muy malo para vivir 
(88) NS (98) NR 
BAR1. En su opinión, en los últimos 12 meses, este barrio, ha mejorado, ha permanecido igual o 
ha empeorado como lugar para vivir? 
(1) Ha mejorado 
(2) Sigue igual 
(3) Ha empeorado 
 (88) NS (98) NR 
Impact Evaluation Report– Mexico 
137
A4C. En su opinión ¿cuál es el problema más grave que enfrenta este barrio? [NO LEER 
ALTERNATIVAS; SÓLO UNA OPCIÓN] 
Agua, falta de 1
9 
Medio ambiente/Basura en las calles 1
0 
Áreas de recreación, falta de 7
1 
Migración 1
6 
Áreas de recreación, mal estado 7
2 
Narcotráfico 1
2 
Balaceras o tiroteos 7
4 
No hay problemas 7
3 
Caminos/vías en mal estado 1
8 
Pandillas 1
4 
Corrupción 1
3 
Pobreza 0
4 
Delincuencia, crimen 0
5 
Protestas populares (huelgas, cierre 
de carreteras, paros, etc.) 
0
6 
Desempleo/falta de empleo 0
3 
Salud, falta de servicio 2
2 
Desigualdad 5
8 
Secuestro 3
1 
Desnutrición 2
3 
Seguridad (falta de) 2
7 
Desplazamiento forzado 3
2 
Tierra para cultivar, falta de 0
7 
Discriminación 2
5 
Transporte, problemas con el 6
0 
Consumo de drogas 1
1 
Violencia 5
7 
Economía, problemas con, crisis de 0
1 
Vivienda, malas condiciones 5
5 
Educación, falta de, mala calidad 2
1 
Otro 7
0 
Electricidad, falta de 2
4 
NS 8
8 
Explosión demográfica/sobre 
población/hacinamiento 
2
0 
NR 9
8 
Drenajes, falta de (no hay tuberías) 7
5 
 
Ahora, le voy a leer algunas frases acerca de cosas que las personas en este barrio pueden o no hacer. 
Para cada una de estas frases, por favor dígame si usted está muy de acuerdo, de acuerdo, en desacuerdo, 
o muy en desacuerdo. [Repita después de cada pregunta “muy de acuerdo, de acuerdo, en 
desacuerdo, o muy en desacuerdo” para ayudar al entrevistado] 
 Muy de 
acuerdo 
De 
acuerdo 
En 
desacuerdo 
Muy en 
desacuerdo NS NR
SOCO1. La primera frase 
dice…cuando hay un problema en este 
barrio, los vecinos suelen organizarse 
para tratar de resolverlo. ¿Usted diría 
que está muy de acuerdo, de 
acuerdo,en desacuerdo, o muy en 
desacuerdo con esta frase? 
1 2 3 4 88 98 
Impact Evaluation Report - Mexico 
138
SOCO2. Este barrio es un barrio 
unido.¿Usted diría que está muy de 
acuerdo, de acuerdo,en desacuerdo, o 
muy en desacuerdo con esta frase? 
1 2 3 4 88 98 
SOCO3. La gente de este barrio está 
dispuesta a ayudar a sus vecinos. 
¿Usted diría que está muy de acuerdo, 
de acuerdo,en desacuerdo, o muy en 
desacuerdo con esta frase?
1 2 3 4 88 98 
SOCO6. Sus vecinos suelen ayudarle a 
vigilar a su casa cuando no hay 
nadie.¿Usted diría que está muy de 
acuerdo, de acuerdo,en desacuerdo, o 
muy en desacuerdo con esta frase? 
1 2 3 4 88 98 
SOCO8. Usted puede confiar en los 
vecinos de este barrio. ¿Usted diría 
que está muy de acuerdo, de 
acuerdo,en desacuerdo, o muy en 
desacuerdo con esta frase? 
1 2 3 4 88 98 
SOCO9P4.La acción organizada de los 
vecinos de este barrio ayuda a resolver 
los problemas comunitarios.¿Usted 
diría que está muy de acuerdo, de 
acuerdo,en desacuerdo, o muy en 
desacuerdo con esta frase?
1 2 3 4 88 98 
SOCO11. Usted siente temor o 
desconfianza cuando tiene que tratar 
con un policía. ¿Usted diría que está 
muy de acuerdo, de acuerdo,en 
desacuerdo, o muy en desacuerdo con 
esta frase?
1 2 3 4 88 98 
SOCO12. Cuando usted ve a un 
policía, piensa que está allí para 
protegerlo (a). ¿Usted diría que está 
muy de acuerdo, de acuerdo,en 
desacuerdo, o muy en desacuerdo con 
esta frase?
1 2 3 4 88 98 
Para continuar, voy a leerle algunas frases. Para cada una de ellas, por favor digame si está usted muy de 
acuerdo, algo de acuerdo, algo en desacuerdo o muy en desacuerdo. . [Repita después de cada pregunta 
“muy de acuerdo, de acuerdo, en desacuerdo, o muy en desacuerdo” para ayudar al entrevistado] 
 Muy de 
acuerdo 
De 
acuerdo 
En 
desacuerdo 
Muy en 
desacuerdo NS NR
SOC20P4.El gobierno proporciona 
información adecuada sobre cómo 
prevenir los delitos. ¿Usted diría que 
está muy de acuerdo, de acuerdo,en 
desacuerdo, o muy en desacuerdo con 
esta frase? 
1 2 3 4 88 98 
Impact Evaluation Report– Mexico 
139
SOC21P4.El gobierno proporciona 
información adecuada sobre la 
seguridad en esta ciudad. ¿Usted diría 
que está muy de acuerdo, de 
acuerdo,en desacuerdo, o muy en 
desacuerdo con esta frase? 
1 2 3 4 88 98 
SOC22P4.Considera que tiene 
conocimientos básicos de cómo 
prevenir los delitos. ¿Usted diría que 
está muy de acuerdo, de acuerdo, en 
desacuerdo, o muy en desacuerdo con 
esta frase? 
1 2 3 4 88 98 
SOC23P4. En este barrio hay 
suficientes oportunidades para que los 
jóvenes se ganen la vida. ¿Usted diría 
que está muy de acuerdo, de acuerdo, 
en desacuerdo, o muy en desacuerdo 
con esta frase? 
1 2 3 4 88 98 
SOC24P4.Unirse a las organizaciones 
criminales para tener un ingreso mayor 
es una alternativa de vida para los 
jóvenes de este barrio ¿Usted diría que 
está muy de acuerdo, de acuerdo, en 
desacuerdo, o muy en desacuerdo con 
esta frase? 
1 2 3 4 88 98 
Pasando a otro tema, a veces la gente y las comunidades tienen problemas que no pueden resolver por 
sí mismas, y para tratar de resolverlos piden ayuda a algún funcionario u oficina del gobierno.
¿Para tratar de resolver algún problema alguna vez ha pedido usted ayuda o 
cooperación ... [LEA CADA OPCIÓN Y ANOTE LA RESPUESTA]
Sí No NS NR 
CP2MEX. A algún diputado Local o Federal? 1 2 88 98 
CP4A. A alguna autoridad local como el Presidente Municipal o Delegado? 1 2 88 98 
CP4. A algún secretaría de gobierno, institución pública, u oficina del Estado? 1 2 88 98 
PN4. En general, ¿usted diría que está muy satisfecho(a), satisfecho(a), insatisfecho(a) o muy 
insatisfecho(a) con la forma en que la democracia funciona en Mexico?
(1) Muy satisfecho (2) Satisfecho (3) Insatisfecho (4) Muy insatisfecho (88) NS (98) NR
Ahora, cambiando de tema, me podría decir, ¿qué tan seguro o inseguro se sentiría usted en las siguientes 
situaciones: muy seguro, algo seguro, algo inseguro o muy inseguro? 
 Muy 
seguro 
Algo 
seguro 
Algo 
inseguro 
Muy 
inseguro 
NS NR 
FEAR0. Estando en su casa con su 
familia cuando es de día ¿se sintiría… 1 2 3 4 88 98 
FEAR1. Estando solo (a) en su casa 
cuando es de día ¿se sintiría 1 2 3 4 88 98 
FEAR2. Estando solo (a) en su casa 
cuando es de noche ¿se sintiría … 1 2 3 4 88 98 
FEAR3. Caminando solo (a) por este 
barriocuando es de día ¿se sintiría 1 2 3 4 88 98 
FEAR4. Caminando solo (a) por este 
barrio cuando es de noche ¿se sintiría 1 2 3 4 88 98 
Impact Evaluation Report - Mexico 
140
usted… 
FEAR6.En general ¿qué tan preocupado está usted de que alguien entre a su casa a robar cuando no hay 
nadie? Está usted … 
 (1) Muy preocupado (2) Algo preocupado (3) Poco preocupado (4) Nada preocupado (88) NS (98) NR 
FEAR7. En general, ¿qué tan preocupado (a) está usted de que alguien lo (la) detenga en la calle, lo (la) 
amenace, golpee, o le haga daño en este barrio? Está usted… 
 (1) Muy preocupado (2) Algo preocupado (3) Poco preocupado (4) Nada preocupado (88) NS (98) NR 
FEAR9.Ahora hablando de otros lugares de [NOMBRE CIUDAD], ¿qué tan preocupado (a) está usted de 
que alguien lo (a) detenga en la calle, lo (a) amenace, golpee, o le haga daño en este municipio? Está 
usted… 
 (1) Muy preocupado (2) Algo preocupado (3) Poco preocupado (4) Nada preocupado (88) NS (98) NR
FEARP4a. Ahora, cambiando de tema: ¿Considera usted que vivir en el ESTADO de [NOMBRE DEL 
ESTADO]es[LEER ALTERNATIVAS]?
(1) Muy seguro (2) Algo seguro(3) Algo inseguro(4) Muy inseguro (88) NS (98) NR 
FEARP4b. ¿Considera usted que vivir en el municipio de [NOMBRE DEL MUNICIPIO] es:[LEER 
ALTERNATIVAS]? 
(1) Muy seguro (2) Algo seguro(3) Algo inseguro(4) Muy inseguro (88) NS (98) NR 
FEARP4c. ¿Hasta qué grado usted y su familia se sienten amenazados por los problemas de 
criminalidad y violencia que afectan a su barrio? Diria ustes que se sienten [LEER ALTERNATIVAS]? 
(1) Seriamente amenzadaos (2) Algo amenazados 
(3) Poco amenzados(4) Nada amenzados (88) NS (98) NR 
FEARP4d.¿Por temor a ser víctima de algún delito en los últimos 12 meses ¿Ha dejado usted de 
realizar actividades que antes hacía? 
(1) Sí (2) No (88) NS (98) NR 
PESE0. En su opinión, este barrio¿es una comunidad muy segura, algo segura, algo insegura, o muy 
insegura? 
(1) Muy segura (2) Algo segura (3) Algo insegura (4) Muy insegura (88) NS (98) NR 
Mayor Igual Menor NS NR 
PESE1. ¿Considera usted que el nivel de violencia actual 
en este barrio es mayor, igual, o menor que el de otras 1 2 3 88 98 
Para protegerse de la delincuencia, en los últimos 12 meses, 
¿usted realizó algún tipo de medida como... Sí No NS NR 
FEAR10.¿Evitó pasar por algunos lugares en este barrio porque 
pueden ser peligrosos? 1 2 88 98 
FEAR11.¿Tomó medidas conjuntas con sus vecinos para protegerse de 
la delincuencia? 1 2 88 98 
FEAR13. ¿Pidió ayuda a la policía? 1 2 88 98 
FEAR14.¿Pidió ayuda a la municipalidad? 1 2 88 98 
FEAR15.¿Construyó muros, rejas o paredes exteriores adicionales en 
su casa? 1 2 88 98 
FEAR16.¿Puso alambre de púas, “razor”, malla electrificada o vidrio roto 
alrededor de su casa? 1 2 88 98 
FEAR17.¿Puso más candados o chapas a las puertas de su casa? 1 2 88 98 
FEAR21.¿Adquirió o compró un arma de fuego? 1 2 88 98 
FEAR22. ¿Ha sentido usted la necesidad de cambiar de barrio o colonia por temor a la 
delincuencia? 
(1) Sí (2) No (88) NS (98) NR 
Impact Evaluation Report– Mexico 
141
colonias o barrios en este municipio? 
PESE2.¿Considera usted que el nivel de violencia actual 
en este barrio es mayor, igual, o menor que el de hace 
12 meses? 
1 2 3 88 98 
Ahora voy a leerle una lista de situaciones que pueden o no ser problema en algunos barrios. Por favor 
dígame si las siguientes situaciones son un problema muy serio, algo serio, poco serio, nada serio o no son 
un problema en este barrio. [Repita después cada pregunta “es esto un problema muy serio, algo 
serio, poco serio, nada serio o no es un problema” para ayudar al entrevistado] 
 Muy 
serio 
Algo 
serio 
Poco 
serio 
Nada 
Serio 
No es un 
problema 
NS NR 
DISO1. Manchas o graffiti en los muros de este 
barrio 1 2 3 4 5 88 98 
DISO3. Casas abandonadas en este barrio 1 2 3 4 5 88 98 
DISO4. Basura en lascalles de este barrio 1 2 3 4 5 88 98 
DISO5. Predios baldíos con montes altos en 
este barrio 1 2 3 4 5 88 98 
DISO6. Calles o lugares oscuros o sin 
iluminación en este barrio 1 2 3 4 5 88 98 
DISO7. Jóvenes o niños en las calles sin hacer 
nada, que andan vagando en este barrio 1 2 3 4 5 88 98 
DISO8. Jóvenes o niños que viven con este 
barrio las pandillas o maras 
1 2 3 4 5 88 98 
DISO9. Ataques o daños intencionales a la 
propiedad privada en este barrio 1 2 3 4 5 88 98 
DISO10. Venta o tráfico de drogas ilegales en 
este barrio 1 2 3 4 5 88 98 
DISO11. Gente peleando y discutiendo en la 
calle de este barrio 1 2 3 4 5 88 98 
DISO13. Gente borracha o ebria en las calles 
de este barrio 1 2 3 4 5 88 98 
DISO14. Gente drogada en las calles de este 
barrio 1 2 3 4 5 88 98 
DISO15. Robo a viviendas en este barrio. 1 2 3 4 5 88 98 
DISO16. Asaltos a las personas cuando 
caminan por la calle de este barrio. 1 2 3 4 5 88 98 
DISO17. Balaceras en este barrio. 1 2 3 4 5 88 98 
DISO18. Riñas o peleas de pandillas o maras 
en este barrio 1 2 3 4 5 88 98 
Ahora para cada una de las siguientes frases, por favor dígame si es muy probable, algo probable, 
poco probable o nada probable que la gente en su barrio pueda actuar del siguiente modo: 
Muy 
probable 
Algo 
probable 
Poco 
probable 
Nada 
probable NS NR 
ICON1. Si un grupo de niños del 
vecindario estuviera faltando a la 
escuela y vagando en las calles, ¿qué 
tan probable es que sus vecinos hagan 
algo al respecto? 
1 2 3 4 88 98 
ICON2. Si algunos niños estuvieran 
manchando las paredes del vecindario, 
¿qué tan probable es que sus vecinos 
hagan algo al respecto? 1 2 3 4 88 98 
Pasando a otro tema y hablando de cosas que le pueden suceder a uno en el país… 
Impact Evaluation Report - Mexico 
142
VIC1EXT.¿Ha sido usted víctima de algún acto de delincuencia en los últimos 12 meses? Es 
decir, ¿ha sido usted víctima de un robo, hurto, agresión, fraude, chantaje, extorsión, 
amenazas o algún otro tipo de acto delincuencial en los últimos 12 meses? 
(1) Sí [Siga] (2) No [Pasar a VIC1HOGAR] (88) NS [Pasar a VIC1HOGAR] (98) 
NR[Pasar a VIC1HOGAR]
VIC2. Pensando en el último acto delincuencial del cual usted fue víctima, de la lista que le 
voy a leer,¿qué tipo de acto delincuencial sufrió? [Leer alternativas]
(01) Robo sin arma sin agresión o amenaza física 
(02) Robo sin arma con agresión o amenaza física 
(03) Robo con arma 
(04) Agresión física sin robo 
(05) Violación o asalto sexual 
(06) Secuestro 
(07) Daño a la propiedad 
(08) Robo en la casa 
(10) Extorsión [o alguien le sobornó] 
(11) Estafa 
(12) Otro 
(88) NS 
(98) NR 
(99) INAP (no fue víctima)
VIC2B. La última vez que fue víctima de la delincuencia, el o los atacantes eran: 
(1) Un extraño o un grupo de extraños 
(2) Personas de otras comunidades en el municipio 
(3) Vecinos 
(4) Amigos cercanos a usted o su familia 
(5) Miembros de su hogar 
(6) Otros parientes 
(7) No vio 
(8) Policías 
(9) Militares 
(10) Miembros de pandillas o bandas de la zona 
(11) Miembros de un cartel de droga 
(88) No sabe (98) No responde (99) INAP 
AOJ1. ¿La última vez que fue víctima de un acto de delicuencia, denunció el hecho a alguna 
institución? 
(1) Sí (2) No (88) NS (98) NR (99) Inap (no víctima)
VIC1HOGAR. ¿Alguna otra persona que vive en su hogar ha sido víctima de algún acto de 
delincuencia en los últimos 12 meses? Es decir, ¿alguna otra persona que vive en su hogar 
ha sido víctima de un robo, hurto, agresión, fraude, chantaje, extorsión, amenazas o algún 
otro tipo de acto delincuencial en los últimos 12 meses? 
(1) Sí (2) No (88) NS (98)NR
VIC1HOGARb. ¿Conoce usted a algún joven que pertenezca a una pandilla? 
(1) Sí (2) No (88) NS (98)NR
Ahora hablando de su comunidad o barrio… 
Teniendo en cuenta su 
experiencia o lo que ha oído 
mencionar, ¿cuáles de los 
siguientes actos de delincuencia 
han occurrido en los últimos 12 
meses en este barrio? 
Sí 
[Pasar 
a 
VICBA
RXA]
No NS N
R 
VICBARXA¿Cúantas veces occurió 
eso? 
(88) NS 
(98) NR 
(99) INAP 
VICBAR1. Han ocurrido robos en 
los últimos 12 meses en este 1 2 88 98 1 2 3 4 5 6 7 8 9 10 
ma
Impact Evaluation Report– Mexico 
143
barrio? s 
(88) (98) (99) 
VICBAR2. Han ocurrido daños a 
la propiedad privada en los últimos 
12 meses en este barrio? 1 2 88 98 1 2 3 4 5 6 7 8 9
10 
ma
s 
(88) (98) (99) 
VICBAR3. Han ocurrido ventas 
de drogas ilegales en los últimos 
12 meses en este barrio?
1 2 88 98 1 2 3 4 5 6 7 8 9 10 
mas 
(88) (98) (99) 
VICBAR4. Han ocurrido 
extorsiones o cobro de derecho de 
piso en los últimos 12 meses en 
este barrio?
1 2 88 98 
1 2 3 4 5 6 7 8 9 10 
mas 
(88) (98) (99) 
VICBAR5. Han ocurrido 
violaciones sexuales en los 
últimos 12 meses en este barrio?
1 2 88 98 1 2 3 4 5 6 7 8 9 10 
mas 
(88) (98) (99) 
VICBAR8. Han ocurrido balaceras 
en los últimos 12 meses en [este 
barrio? 
1 2 88 98 1 2 3 4 5 6 7 8 9 10 
mas 
(88) (98) (99) 
VICBAR6. Han ocurrido 
secuestros en los últimos 12 
meses en este barrio?
1 2 88 98 1 2 3 4 5 6 7 8 9 10 
mas 
(88) (98) (99) 
VICBAR7. Han ocurrido 
asesinatos en los últimos 12 
meses en este barrio?
1 2 88 98 1 2 3 4 5 6 7 8 9 10 
mas 
(88) (98) (99) 
[Entrevistador: si la respuesta, en el literal A, es NO, NS, o 
NR, marcar y pasar a la siguiente; si la respuesta es Sí 
continuar con los literales B hasta F; las celdas en gris o 
negro indican que las preguntas de ese literal no se hace y 
se pasa al siguiente literal] 
En los últimos 12 meses, le ha occurrido a usted o a alguien que 
vive en su hogar lo siguiente: 
A. 
(1) Sí 
(2) No 
(88) NS 
(98)NR 
(99) 
B. 
¿Cuánta
s veces 
ocurrió 
esto en 
los 
últimos 
12 
meses?
[Leer 
opcione
s] 
(1) Una 
vez 
(2) Dos 
veces 
(3) Tres 
C. La última 
vez que esto 
pasó,¿dónde 
ocurrió el 
hecho? [Leer 
opciones] 
(1) En su 
hogar 
(2) En este 
barrio o 
comunidad 
(3) En este 
municipio 
(4) En otro 
municipio 
(5) Fuera del 
Impact Evaluation Report - Mexico 
144
INAP veces 
(4) Más 
de tres 
veces 
(88) NS 
(98)NR 
(99) 
INAP 
país 
(88) NS 
(98) NR 
(99) INAP 
VIC70. En los últimos 12 meses, ¿alguien entró a robar algo en 
su casa, que no sea un vehículo? 
 
VIC72. En los últimos 12 meses, ¿a usted o a alguien que vive en 
su hogar le robaron algún objeto dejado dentro de su vehículo o 
alguna parte de su vehículo? [(99) INAP No tienen vehículo] 
 
VIC71. En los últimos 12 meses, ¿a usted o alguien que vive en 
su hogar le robaron el vehículo o una moto? [(99) INAP No tienen 
vehículo] 
 
VIC73. En los últimos 12 meses, ¿a usted o alguien que vive en 
su hogar le robaron algo estando fuera de su casa? 
 
VIC74. En los últimos 12 meses, ¿alguien deliberadamente 
destruyó o dañó su casa, negocio o cualquier propiedad suya o 
de alguien que vive en su hogar? 
 
VIC75. En los últimos 12 meses, ¿a usted o alguien que vive en 
su hogar le atacaron a golpes sin que le robaran algo? 
 
VIC78. En los últimos 12 meses, ¿a usted o alguien que vive en 
su hogar lo han tocado o atacado por motivos sexuales de un 
modo ofensivo? 
 
VIC79. En los últimos 12 meses, ¿ha sido usted o alguien que 
vive en su hogar víctima de alguna estafa? 
 
VIC80. En los últimos 12 meses, ¿ha sido usted o alguien que 
vive en su hogar víctima de un chantaje, extorsión o renta? 
 
VIC81. En los últimos 12 meses, ¿ha sido usted o alguien que 
vive en su hogar víctima de un secuestro o un secuestro exprés? 
 
VIC82. En los últimos 12 meses, ¿alguien que vivía en su 
hogar fue asesinado? 
 
AOJ8. Para poder capturar delincuentes, ¿cree usted que las autoridades siempre deben 
respetar las leyes o en ocasiones pueden actuar al margen de la ley? 
(1) Deben respetar las leyes siempre (2) En ocasiones pueden actuar al margen de la ley 
 (88)NS (98) NR 
AOJ12. Si usted fuera víctima de un robo o asalto, ¿cuánto confiaría en que el sistema judicial 
castigaría al culpable? Confiaría…[LEER ALTERNATIVAS]
(1) Mucho (2) Algo (3) Poco (4) Nada (8) NS/NR
Ahora vamos a hablar de su municipio... 
NP1. ¿Ha asistido a una sesión de cabildo abierto o una sesión municipal durante los últimos 12 
meses? 
(1) Sí (2) No (88)NS (98) NR 
Impact Evaluation Report– Mexico 
145
SGL1. ¿Diría usted que los servicios que el municipio está dando a la gente son: [LEER 
ALTERNATIVAS] 
(1) Muy buenos (2) Buenos (3) Ni buenos ni malos (regulares) (4) Malos (5) Muy 
malos (pésimos) (88)NS (98) NR 
NICMUNI6. ¿Qué grado de confianza tiene usted en el buen manejo de los fondos por parte de 
la alcaldía? [Leer alternativas]
(3) Mucha confianza (2) Algo de confianza (1) Poca confianza (0) Nada de confianza 
(88) NS (98) NR
[ENTREGAR TARJETA “E”] 
En esta tarjeta hay una escalera con gradas numeradas del uno al siete, en la cual el 1 es la grada más 
baja y significa NADA y el 7 es la grada más alta y significa MUCHO. Por ejemplo, si yo le preguntara 
hasta qué punto le gusta ver televisión, si a usted no le gusta ver nada, elegiría un puntaje de 1.Si por el 
contrario le gusta ver mucha televisión me diría el número 7. Si su opinión está entre nada y mucho 
elegiría un puntaje intermedio. ¿Entonces, hasta qué punto le gusta a usted ver televisión? Léame el 
número.[ASEGÚRESE QUE EL ENTREVISTADO ENTIENDA CORRECTAMENTE]. 
1 2 3 4 5 6 7 NS=88 
NR=98 
Nada Mucho 
Anotar el número, 1-7, NS=88, NR=98 
Voy a hacerle una serie de preguntas, y le voy a pedir que para darme su respuesta utilice 
los números de esta escalera. Recuerde que puede usar cualquier número. 
B1. ¿Hasta qué punto cree usted que los tribunales de justicia de Honduras 
Mexicogarantizan un juicio justo? (Sondee:Si usted cree que los tribunales no garantizan 
paranada la justicia, escoja el número 1; si cree que los tribunales garantizan mucho la 
justicia escoja el número 7 o escoja un puntaje intermedio ) 
B2. ¿Hasta qué punto tiene usted respeto por las instituciones políticas de Mexico? 
B3. ¿Hasta qué punto cree usted que los derechos básicos del ciudadano están bien 
protegidos por el sistema político mexicano? 
B4. ¿Hasta qué punto se siente usted orgulloso de vivir bajo el sistema político mexicano? 
B6. ¿Hasta qué punto piensa usted que se debe apoyar al sistema político mexicano? 
B10A. ¿Hasta qué punto tiene confianza en el sistema de justicia?
B12. ¿Hasta qué punto tiene confianza usted en las Fuerzas Armadas? 
B13. ¿Hasta qué punto tiene confianza usted en la Congreso Nacional? 
B14. ¿Hasta qué punto tiene confianza usted en el Gobierno Federal? 
B18. ¿Hasta qué punto tiene confianza usted en la Policía?
B20. ¿Hasta qué punto tiene confianza usted en la Iglesia Católica? 
B20A. ¿Hasta qué punto tiene confianza usted en la Iglesia Protestante? 
B21. ¿Hasta qué punto tiene confianza usted en los partidos políticos? 
B21A. ¿Hasta qué punto tiene confianza usted en el presidente?
B31.Hasta qué punto tiene usted confianza en la Suprema Corte de Justicia? 
B32. ¿Hasta qué punto tiene usted confianza en su municipio? 
B43. ¿Hasta qué punto tiene usted orgullo de ser mexicano(a)? 
B60. ¿Hasta qué punto tiene usted confianza en el Comité Municipal de Seguridad?
B62. ¿Hasta qué punto se siente usted orgulloso de vivir en este barrio? 
Usando la misma escalera 
NADA 1 2 3 4 5 6 7 MUCHO
Anotar 1-7, 
NS=88 
NR=98
Impact Evaluation Report - Mexico 
146
Usando la misma escalera 
NADA 1 2 3 4 5 6 7 MUCHO
Anotar 1-7, 
NS=88 
NR=98
N11. ¿Hasta qué punto diría que el gobierno actual mejora la seguridad ciudadana?
N15. ¿Hasta qué punto diría que el gobierno actual está manejando bien la economía?
N16. Y ahora hablando de su municipio,¿hasta qué punto diría que el gobierno municipal 
actual mejora la seguridad ciudadana?
SOCO9.Y ahora hablando de este barrio ¿hasta qué punto diría que los vecinos de este 
barrio están organizados para prevenir la delincuencia y la violencia?
[Recoger tarjeta “E”] 
M1. Y hablando en general del gobierno actual, ¿diría usted que el trabajo que está realizando el 
Enrique Peña Nieto es...?: [LEER ALTERNATIVAS] 
(1) Muy bueno (2) Bueno (3) Ni bueno, ni malo (regular) (4) Malo (5) Muy 
malo (pésimo) (88)NS (98) NR 
POLE2. En general, ¿usted diría que está muy satisfecho(a), satisfecho(a), insatisfecho(a) o muy 
insatisfecho(a) con el trabajo realizado por la Policía en este barrio [Leer alternativas] 
(1) Muy satisfecho (2) Satisfecho (3) Insatisfecho (4) Muy insatisfecho (7) [espontánea no 
leer]La policia no está haciendo nada (88)NS (98) NR 
ICO2.¿Con qué frecuencia la policía patrulla en este barrio? Usted diría: [LEER 
ALTERNATIVAS]
(1) Varias veces al día 
(2) Al menos una vez al día 
(3) Algunas veces por semana 
(4) Algunas veces por mes 
(5) Rara vez 
(6) Nunca 
(88) NS(98) NR 
En los últimos 12 meses, cuáles de las siguientes actividades ha visto a 
la Policía hacer en este barrio 
SÍ No NS NR 
CPOL2.Asistir a reuniones de vecinos de este barrio 1 2 88 98 
CPOL3. Ha visto a la Policía ayudar a realizar actividades de prevención 
de la delincuencia en este barrio 1 2 88 98 
CPOL4. Relacionarse con los niños y jóvenes de este barrio a través de 
actividades recreativas y escolares 1 2 88 98 
Impact Evaluation Report– Mexico 
147
[ENTREGAR TARJETA “G”] 
Ahora, vamos a usar una escalera similar a la anterior, pero el número 1 representa “muy en 
desacuerdo” y el número 7 representa “muy de acuerdo”. Un número entre el 1 y el 7, representa un 
puntaje intermedio.Anotar Número 1-7, 88 para los que NS y 98 para los NR 
1 2 3 4 5 6 7 NS NS
Muy en desacuerdo Muy de acuerdo 88 98 
Anotar Número 1-7, 
NS=88, NR=98
ING4. Puede que la democracia tenga problemas, pero es mejor que cualquier otra forma de 
gobierno. ¿Hasta qué punto está de acuerdo o en desacuerdo con esta frase? 
 
[Recoger tarjeta “G”] 
[ENTRÉGUELE AL ENTREVISTADO LA TARJETA “H”] 
Ahora vamos a cambiar a otra tarjeta. Esta nueva tarjeta tiene una escaleradel 1 a 10, el 1 indica que 
usted desaprueba firmemente y el 10 indica que usted aprueba firmemente. Voy a leerle una lista de 
algunas acciones o cosas que las personas pueden hacer para alcanzar sus metas y objetivos políticos. 
Quisiera que me dijera con qué firmeza usted aprobaría o desaprobaría que las personas hagan las 
siguientes acciones. 
1 2 3 4 5 6 7 8 9 10 
 NS NR
Desaprueba firmemente Aprueba firmemente 88 98 
 1-10, 88, 
98 
E5. Que las personas participen en manifestaciones permitidas por la ley. ¿Hasta qué 
punto aprueba o desaprueba?
E8. Que las personas participen en una organización o grupo para tratar de resolver los 
problemas de las comunidades. ¿Hasta qué punto aprueba o desaprueba?
E11. Que las personas trabajen en campañas electorales para un partido político o 
candidato. ¿Hasta qué punto aprueba o desaprueba?
E15. Que las personas participen en un cierre o bloqueo de calles o carreteras como forma 
de protesta. Usando la misma escala, ¿hasta qué punto aprueba o desaprueba?
E14.Que las personas invadan propiedades o terrenos privados como forma de protesta. 
¿Hasta qué punto aprueba o desaprueba?
E2. Que las personas ocupen (invadan) fábricas, oficinas y otros edificios como forma de 
protesta. ¿Hasta qué punto aprueba o desaprueba?
E3. Que las personas participen en un grupo que quiera derrocar por medios violentos a un 
gobierno electo. ¿Hasta qué punto aprueba o desaprueba?
E16. Que las personas hagan justicia por su propia cuenta cuando el Estado no castiga a 
los criminales. ¿Hasta qué punto aprueba o desaprueba?
Impact Evaluation Report - Mexico 
148
. 
Las preguntas que siguen son para saber su opinión sobre las diferentes ideas que 
tienen las personas que viven en Mexico. Por favor continúe usando la escalera de 10 
puntos
1-10, 
NS=88, 
NR=98 
D1. Hay personas que siempre hablan mal de la forma de gobierno deMexico, no sólo del 
gobierno de turno, sino del sistema de gobierno, ¿con qué firmeza aprueba o desaprueba 
usted el derecho de votar de esas personas? Por favor léame el número de la escala: 
[Sondee: ¿Hasta qué punto?]
D2. ¿Con qué firmeza aprueba o desaprueba usted que éstas personas puedan llevar a 
cabo manifestaciones pacíficas con el propósito de expresar sus puntos de vista? Por 
favor léame el número.
D3. Siempre pensando en los que hablan mal de la forma de gobierno de Mexico. ¿Con 
qué firmeza aprueba o desaprueba usted que estas personas puedan postularse para 
cargos públicos?
D4. ¿Con qué firmeza aprueba o desaprueba usted que estas personas salgan en la 
televisión para dar un discurso?
D5. Y ahora, cambiando el tema, y pensando en los homosexuales. ¿con qué firmeza 
aprueba o desaprueba que estas personas puedan postularse para cargos públicos?
[Recoger tarjeta “H”] 
AUT1. Hay gente que dice que necesitamos un líder fuerte que no tenga que ser electo a través del voto 
popular. Otros dicen que, aunque las cosas no funcionen, la democracia electoral, o sea el voto popular, 
es siempre lo mejor. ¿Qué piensa usted? [LEER ALTERNATIVAS] 
(1) Necesitamos un líder fuerte que no tenga que ser electo, o 
(2) La democracia electoral es lo mejor (88) NS 
(98) NR 
Ahora queremos hablar de su experiencia personal con 
cosas que pasan en la vida...
INAP 
No trató 
o tuvo 
contacto 
No Sí NS NR
EXC2. ¿Algún agente de policía le pidió una mordida 
 enlos últimos 12 meses 0 1 88 98 
EXC6. ¿En los últimos 12 meses, ¿algún empleado público 
le ha solicitado una mordida? 0 1 88 98 
EXC11. ¿Ha tramitado algo enel municipio dónde vive en los 
últimos12 meses? 
No  Marcar 9 
Sí  Preguntar: 
Para tramitar algo en el municipio, como un permiso, por 
ejemplo, durante el último año, ¿ha tenido que pagar alguna 
suma además de lo exigido por la ley? 
9 0 1 88 98 
VB1. ¿Está usted empadronado? [SOLO SI NO COMPRENDE PREGUNTE: ¿Tiene 
credencial para votar?]
(1) Sí (2) No (3) En trámite (88) NS (98) NR 
EXC7. Teniendo en cuenta su experiencia o lo que ha oído mencionar, ¿la corrupción de los 
funcionarios públicos en el país está: [LEER] 
(1) Muy generalizada (2) Algo generalizada (3) Poco generalizada (4) Nada generalizada 
(88) NS (98) NR
Impact Evaluation Report– Mexico 
149
VB2. ¿Votó usted en las últimas elecciones presidenciales del 1º de julio de 2012 
(1) Sí votó [Siga]
(2) No votó [Pasar a VB3B]
(88) NS [Pasar a VB3B] 
(98) NR [Pasar a VB3B]
VB3. ¿Por quién votó para Presidente en las últimas elecciones presidenciales de 2012? 
[NO LEER LISTA] 
(00) Ninguno (fue a votar pero dejó la boleta en blanco o anuló su voto) 
(501) Josefina Vázquez Mota, PAN 
(502) Enrique Peña Nieto, PRI – PVEM 
(503) Andrés Manuel López Obrador, PRD– PT – Movimiento Ciudadano 
(504) Gabriel Ricardo Quadri de la Torre - PANAL 
(77) Otro 
(88) NS 
(98) NR 
(99) Inap (No votó) 
VB3B. ¿Por cual partido político votó en las últimas elecciones para elegir presidente 
municipal? 
[NO LEER LISTA] 
(00) Ninguno (fue a votar pero dejó la boleta en blanco o anuló su voto) 
(1) PAN 
(2) PRI 
(3) PRD 
(4) PVEM 
(5) PANAL 
(6) PT 
(7) MOVIMIENTO CIUDADANO 
(77) Otro 
(88) NS 
(98) NR 
(99) Inap (No votó) 
PROT3. En los últimos 12 meses,¿ha participado en una manifestación o protesta pública? 
(1) Sí ha participado [Siga] (2) No ha participado [Pase a SOCT1] (88) NS [Pase a SOCT1] (98) 
NR [Pase a SOCT1]
PROT4. ¿Cuántas veces ha participado en una manifestación o protesta pública en los últimos 12 
meses? ______________________ (88) NS (98) NR (99) INAP 
SOCT1. Ahora, hablando de la economía… ¿Cómo calificaría la situación económica del país? ¿Diría 
usted que es muy buena, buena, ni buena ni mala, mala o muy mala? 
(1) Muy buena (2) Buena (3) Ni buena, ni mala (regular) (4) Mala (5) Muy mala (pésima) 
(88) NS (98) NR 
SOCT2. ¿Considera usted que la situación económica actual del país es mejor, igual o peor que hace 
doce meses? 
(1) Mejor (2) Igual (3) Peor (88) NS (98) NR 
IDIO1. ¿Cómo calificaría en general su situación económica? ¿Diría usted que es muy buena, buena, ni 
buena ni mala, mala o muy mala?
(1) Muy buena (2) Buena (3) Ni buena, ni mala (regular) (4) Mala 
(5) Muy mala (pésima) (88) NS (98) NR
IDIO2. ¿Considera usted que su situación económica actual es mejor, igual o peor que la de hace doce 
meses?
(1) Mejor (2) Igual (3) Peor (88) NS (98) NR
POL1. ¿Qué tanto interés tiene usted en la política:mucho, algo, poco o nada? 
Impact Evaluation Report - Mexico 
150
(1) Mucho (2) Algo (3) Poco (4) Nada (88) NS (98) NR 
Por propósitos estadísticos, ahora queremos saber cuánta información sobre política y sobre el país 
tiene la gente… 
GI0. ¿Con qué frecuencia sigue las noticias, ya sea en la televisión, la radio, los periódicos, o el internet? 
[Leer opciones]: (1) A diario (2) Algunas veces por semana (3) Algunas veces por mes 
(4) Rara vez (5) Nunca (88) NS (98) NR 
Cambiando de tema... 
¿Alguna vez se ha sentido discriminado o tratado de manera injusta por… [Repetir después de cada 
pregunta: muchas veces, algunas veces, pocas veces, o nunca] 
Mucha
s veces 
Algunas 
veces 
Poca
s 
veces 
Nunca NS NR 
DIS11.Su color de piel, usted diría que 
eso ha sucedido muchas veces, 
algunas veces, pocas veces, o nunca? 
1 2 3 4 88 98 
DIS12. Por su sexo? 1 2 3 4 88 98 
DIS13. Por su condición económica? 1 2 3 4 88 98 
DIS17. Por su manera de hablar o 
acento? 1 2 3 4 88 98 
DIS14. Por su apariencia física? 1 2 3 4 88 98 
DIS15. Por su orientación sexual? 1 2 3 4 88 98 
DIS16. El lugar donde vive? 1 2 3 4 88 98 
DIS17. Por su forma de vestir? 1 2 3 4 88 98 
DIS18. Por alguna otra razón? 1 2 3 4 88 98 
Ahora, cambiando de tema… 
SEG6. ¿Ha visto u oído que alguna institución haya realizado alguna actividad de apoyo dirigidas a los 
jóvenes de este barrio en los últimos 12 meses? 
(1) Sí (2) No [Pase a SEG2](88) NS (98) NR [Pase a SEG2]
SEG6a. Conoce o sabe de alguien de este barrio que haya participado de estas actividades 
en los últimos 12 meses?
(1) Sí (2) No (88) NS (98) NR 
SEG2. En los últimos 12 meses, ¿ha visto u oído que alguna institución haya realizado obras públicas 
en este barrio, tales como mejoramiento del alumbrado público, actividades de limpieza o construcción o 
reparación de calles, canchas, o parques? 
(1) Sí (2) No (88) NS (98) NR 
Sí No NS NR Inap 
CP30. ¿En este barrio hay espacios o lugares donde los vecinos puedan 
reunirse? 1 0 88 98 99 
CP21. ¿Hay una asociaciónde vecinos en este barrio? [Si la respuesta 
es NO, pase a CP26] 1 0 88 98 
CP22. ¿Es usted miembro de esa asociación de vecinos o patronato? 1 0 88 98 99 
CP23. En los últimos tres meses, ¿ha asistido usted a una reunión 
convocada por la asociación de vecinos, o patronato? 1 0 88 98 99 
CP24. En los últimos tres meses, ¿ha realizado usted trabajo voluntario 
para esa asociación o junta directiva? [Si la respuesta es SI Pase a 
CP5]
1 0 88 98 99 
Impact Evaluation Report– Mexico 
151
Sí No NS NR Inap 
CP25. En los últimos tres meses, ¿ha promovido la asociación de 
vecinos de este barrio actividades para prevenir la delincuencia, tales 
como tomar medidas de seguridad para el vecindario u otras actividades? 1 0 88 98 99 
CP26. ¿Hay alguna otra asociación o institución que esté promoviendo 
programas para la prevención de la delicuencia en este barrio? 1 0 88 98 99 
 Una 
vez a 
la 
seman
a 
Una o 
dos 
veces 
al mes 
Una o 
dos 
veces 
al año 
Nunc
a NS NR 
CP5. ¿ Y en los últimos doce meses ha 
contribuido para ayudar a solucionar algún 
problema de su comunidad o de los vecinos de 
su barrio o colonia? Por favor, dígame si lo hizo 
por lo menos una vez a la semana, una o dos 
veces al mes, una o dos veces al año, o nunca 
en los últimos 12 meses. 
(1) (2) (3) (4) (88
) (98) 
 
CP5A. ¿Ha donado Ud. dinero o materiales 
para ayudar a solucionar algún problema de la 
comunidad o de su barrio? 
1 2 3 4 88 98 
CP5B. ¿Ha contribuido Ud. con su propio 
trabajo o mano de obra? 1 2 3 4 88 98 
CP5C. ¿Ha estado asistiendo Ud. a reuniones 
comunitarias sobre algún problema o sobre 
alguna mejora? 
1 2 3 4 88 98 
CP5D. ¿Ha tratado de ayudar Ud. a organizar 
algún grupo nuevo para resolver algún 
problema del barrio, o para buscar alguna 
mejora? 
1 2 3 4 88 98 
SEG3. Comparando la situación actual de esta comunidad con la de hace 12 meses, ¿cree usted que los 
vecinos de este barrio colaboran más, lo mismo, o menos para resolver los problemas de la comunidad? 
(1) Colaboran más 
(2) Colaboran lo mismo 
(3) Colaboran menos 
(88) NS (98) NR 
Ahora, cambiando de tema, para cada una de estas frases, por favor dígame si usted está muy de acuerdo, 
de acuerdo, en desacuerdo, o muy en desacuerdo. [Repita después de cada pregunta “muy de acuerdo, 
de acuerdo, en desacuerdo, o muy en desacuerdo” para ayudar al entrevistado]
 Muy de 
acuerdo 
De 
acuerdo 
En 
desacuerdo 
Muy en 
desacuerdo NS NR 
EMP1.Usted sabe que puede 
hacer una diferencia y contribuir en 
la mejora dela colonia 
1 2 3 4 88 98 
COM1. Se siente responsable de 
participar en actividades de 
mejoramiento de la colonia 
1 2 3 4 88 98 
Impact Evaluation Report - Mexico 
152
COM2. Está involucrado y participa 
de las actividades que se 
organizan en su colonia 
1 2 3 4 88 98 
EMP2. Sabe cómo trabajar con 
otros para resolver problemas de 
su colonia 
1 2 3 4 88 98 
EMP3.Sabe que tiene la capacidad 
de mejorar cosas en esta colonia 1 2 3 4 88 98 
COM3.Tiene el deseo de participar 
más activamente en los proyectos 
de mejoramiento delacolonia 
1 2 3 4 88 98 
EMP5. No tiene la capacidad de 
cambiar las cosas que no le gustan 
de su colonia 
1 2 3 4 88 98 
EMP7.1.Ahora le preguntaré sobre alguna actividad en el barrio o decisión 
que usted y sus vecinos hayan tomado y que haya beneficiado al barrio o 
colonia, como por ejemplo, mejorar un espacio verde o vigilar la seguridad 
del barrio. ¿Podría identificar una actividad en el barrio o decisión importante 
que los vecinos de este barrio hayan realizado en los últimos 6 meses? [A 
anotar SI o NO dependiendo si el entrevistado puede o no describir una 
actividad en el barrio o decisión que el y sus vecinos tomaron]
Sí 
(1) 
No[Pase 
a ED]
(0) NS/NR (99) 
EMP7.2. ¿Fue fácil para usted y sus vecinos realizar esta actividad en el 
barrio o tomar esa decisión? (1) (0) NS/NR (99) 
EMP7.3. En su opinión ¿La actividad en el barrio que realizaron o la decisión 
que se tomó benefició al barrio? (1) (0) NS/NR (99) 
EMP7a. Podría decirme quién o quienes tomaron la iniciativa de realizar esta actividad en el 
barrio o tomaron esta decision? [No leer lista]
(1) Los líderes del patronato 
(2) Los vecinos del barrio 
(3) La municipalidad 
(4) Otra organización (ONG, iglesia, escuelas) 
(88) NS (98) NR 
EMP7b. Ahora, podría decirme por qué motivo realizaron esta actividad en el barrio o se tomó 
la decisión? [No leer lista] 
(1) Para mejorar la seguridad del barrio (prevenir robos, tráfico de drogas) 
(2) Para ayudar a un vecino, una escuela, un club o iglesia del barrio 
(3) Para mejorar la infraestructura del barrio (mejorarmiento de las plaza, calles, alumbrado 
etc.) 
(4) Para mejorar la economía del barrio (generar empleos, oportunidades de negocios) 
(5) Para mejorar la calidad de vida de los niños y jovenes del barrio 
(6) Otros motivos 
 (88) NS (98) NR 
Impact Evaluation Report– Mexico 
153
ED. ¿Cuál fue el último año de educación que usted completó o aprobó?
_____ Año de ___________________ (primaria, secundaria, universitaria, superior no 
universitaria) = ________ años total [Usar tabla a continuación para código] 
10 20 30 40 50 60
Ninguno 0 
Primaria 1 2 3 4 5 6 
Secundaria 7 8 9 
Bachillerato/Profesional Técnico/Media 
Superior 
10 11 12 
Universitaria 13 14 15 16 17 18+ 
Superior no universitaria 13 14 15 16 
NS 88 
NR 98 
EDU1.¿Cuántos niños entre 6 y 17 años viven en esta casa? _______________ 
0. Ninguno [Pase a SEG1] (NS=888, NR=988) [Pase a SEG1] 
EDU2.¿Cuántos de estos [Número de niños en EDU1]niños entre 6 y 17 años estudian o 
van a la escuela actualmente?
_______________ 0. Ninguno (NS=888, NR=988, INAP=999)
EDU3. ¿Cuántos de los [Número de niños en EDU2] niños entre 6 y 17 años que 
estudian participan regularmente en actividades extra escolares, es decir actividades 
después de clases o de la jornada escolar, como por ejemplo actividades culturales o 
deportivas?_______________ 0. Ninguno (NS=888, NR=988) INAP=999 
EDU4.¿Cuántos niños entre 10 y 17 años viven en esta casa? _______________ 0. 
Ninguno [Pase a SEG1] (NS=888, NR=988) [Pase a SEG1] 
EDU2a.¿Cuántos de los niños entre 10 y 17 años que viven en esta casa, no estudian ni 
trabajan? ....................... 0.Ninguno 
(88) NS 
(98) NR
Hablando de los niños y jóvenes de entre 10 y 17 años que viven en su hogar 
TIEM1.¿Podría decirle que hacen principalmente estos jóvenes en su tiempo libre? [No 
leer lista] 
(1) Escuchan música, ven televisión o usan la computadora 
(2) Están en su casa o en la casa de algún amigo 
(3) Están en la calle con sus amigos 
(4) Practican algún deporte o realizan alguna actividad cultural( tocar un instrumento, 
clases de dibujo, música, teatro, manualidades) 
(5) Otras actividades 
(88) NS 
(98) NR 
Impact Evaluation Report - Mexico 
154
SEG1. En los últimos 12 meses, ¿Ha visto u oído que alguna institución haya realizado 
actividades dirigidas a los jóvenes de este barrio para que usen mejor su tiempo libre? 
(1) Si [Si EDU1=0 pase a PROB1] 
(2) No [Pase a PROB1] (88) NS (98) NR 
SEG1a. ¿Cuántos de los niños y jóvenes de entre 10 y 17 años de su hogar han 
participado de estas actividades? 
_______________ 0. Ninguno 
(99) No hay niños entre 10 y 17 años en el hogar 
(88) NS 
(98) NR 
[PARA todas las personas] 
PROB1. Según su opinión, ¿Cuál cree que es problema más importante que afecta 
actualmente a los jóvenes de este barrio? [Leer opciones] 
(1) La falta de oportunidades para acceder a empleos 
(2) La delincuencia/el pandillaje y la violencia 
(3) El consumo excesivo de alcohol y drogas 
(4) La falta de educación o capacitación para trabajar 
(5) La falta de motivación de los jóvenes 
(6) Los embarazos de adolescentes 
 (7) Otros problemas [ 
(88) NS 
(98) NR
 
Ahora cambiando de tema, 
DEL1.Conoce o se enteró de algún jóven de este barrio que haya participado en algún 
delito en los últimos 12 meses? 
(1) Sí (2) No (88) NS (98) NR
DEL2. Conoce o se enteró de algún joven de este barrio que haya consumido drogas en 
los últimos 12 meses? 
(1) Sí (2) No (88) NS (98) NR
DEL3. Conoce o se enteró de algún joven de este barrio que haya consumido alcohol en 
exceso en los últimos 12 meses? 
(1) Sí (2) No (88) NS (98) NR
Ahora, hablando de usted… 
Q2. ¿Cuál es su edad en años cumplidos? __________ años (NS=888, NR=988)
Q3C. Si usted es de alguna religión, ¿podría decirme cuál es su religión? [NO LEER OPCIONES]
(01) Católico 
(02) Protestante, Protestante Tradicional o Protestante no Evangélico (Cristiano, Calvinista; Luterano; 
Metodista; Presbiteriano; Discípulo de Cristo; Anglicano; Episcopaliano; Iglesia Morava). 
(03) Religiones Orientales no Cristianas (Islam; Budista; Hinduista; Taoísta; Confucianismo; Baha’i). 
(04) Ninguna (Cree en un Ser Superior pero no pertenece a ninguna religión) 
(05) Evangélica y Pentecostal (Evangélico, Pentecostal; Iglesia de Dios; Asambleas de Dios; Iglesia 
Universal del Reino de Dios; Iglesia Cuadrangular; Iglesia de Cristo; Congregación Cristiana; Menonita; 
Hermanos de Cristo; Iglesia Cristiana Reformada; Carismático no Católico; Luz del Mundo; Bautista; 
Iglesia del Nazareno; Ejército de Salvación; Adventista; Adventista del Séptimo Día, Sara Nossa Terra). 
(06) Iglesia de los Santos de los Últimos Días (Mormones) y Testigos de Jehová. 
(07) Religiones Tradicionales (Candomblé, Vodou, Rastafari, Religiones Mayas, Umbanda; María Lonza; 
Inti, Kardecista, Santo Daime, Esoterica). 
(10) Judío (Ortodoxo, Conservador o Reformado) 
Impact Evaluation Report– Mexico 
155
(11) Agnóstico o ateo (no cree en Dios) 
(88) NS 
(98) NR 
[ENTREGAR TARJETA “I”] 
Q10NEW. ¿En cuál de los siguientes rangos se encuentran los ingresos familiares mensuales de este 
hogar, incluyendo las remesas del exterior y el ingreso de todos los adultos e hijos que trabajan? 
 [Si no entiende, pregunte: ¿Cuánto dinero entra en total a su casa al mes?] 
(00) Ningún ingreso 
(01) Menos de $480 
(02) Entre $480 - $720 
(03) Entre $721 - $960 
(04) Entre $961 - $1,440 
(05) Entre $1,441 - $2,150 
(06) Entre $2,151 - $2,870 
(07) Entre $2,871 - $3,590 
(08) Entre $3,591 - $4,310 
(09) Entre $4,311 - $6,460 
(10) Entre $6,461 - $7,180 
(11) Entre $7,181 - $8,610 
(12) Entre $8,611 - $10,050 
(13) Entre $10,051 - $11,490 
(14) Entre $11,491 - $12,920 
(15) Entre $12,921 - $14,360 
(16) Más de $14,360 
(88) NS (98) NR[RECOGER TARJETA “I”]
Q10A. ¿Usted o alguien que vive en su casa recibe remesas , es decir, ayuda económica del exterior? 
(1) Sí [Siga] (2) No [Pase a Q14](88) NS[Pase a Q14] (98) NR [Pase a Q14]
Q10B. [Sólo si recibe remesas] ¿Hasta qué punto dependen los ingresos familiares de esta casa de 
las remesas del exterior? [LEER ALTERNATIVAS] 
(1) Mucho (2) Algo (3) Poco (4) Nada (88) NS (98) NR (99) Inap 
Q14. ¿Tiene usted intenciones de irse a vivir o a trabajar a otro país en los próximos tres años? 
(1) Sí (2) No (88) NS (98) NR 
Q10D. El salario o sueldo que usted recibe y el total del ingreso familiar: [LEER ALTERNATIVAS] 
(1) Les alcanza bien y pueden ahorrar 
(2) Les alcanza justo sin grandes dificultades 
(3) No les alcanza y tienen dificultades 
(4) No les alcanza y tienen grandes dificultades 
(88) [No leer] NS 
(98) [No leer] NR 
Q11. ¿Cuál es su estado civil? [NO LEER ALTERNATIVAS] 
(1) Soltero (2) Casado (3) Unión libre (acompañado) (4) Divorciado 
(5) Separado (6) Viudo (88) NS (98) NR 
Q12. ¿Tiene hijos(as)? ¿Cuántos? _________ (00= ninguno) NS=88 NR=98 
PER1. ¿Cuántas personas incluyendo recién nacidos, niños y adultos viven habitualmente 
en este hogar? __________ Anotar Número (NS=888, NR=988)
Para finalizar, podría decirme si en su casa tienen:[LEER ALTERNATIVAS] 
R1. Televisor (0) No (1) Sí
Impact Evaluation Report - Mexico 
156
R3. Refrigeradora (nevera) (0) No (1) Sí
R4. Teléfono convencional /fijo (no celular) (0) No (1) Sí
R4A. Teléfono celular (0) No (1) Sí
R5. Vehículo. ¿Cuántos? (0) No (1) Uno (2) Dos (3) Tres o más
R6. Lavadora de ropa (0) No (1) Sí
R7. Microondas (0) No (1) Sí
R8. Motocicleta (0) No (1) Sí
R12. Agua potable dentro de la casa (0) No (1) Sí
R14. Cuarto de baño dentro de la casa (0) No (1) Sí
R19. Ducha para agua caliente (0) No (1) Sí 
R15. Computadora (0) No (1) Sí
R16. Televisor de pantalla plana (0) No (1) Sí
R18. Servicio de internet (0) No (1) Sí
OCUP4A. ¿A qué se dedica usted principalmente? ¿Está usted actualmente: [LEER ALTERNATIVAS]
(1) ¿Trabajando? [Siga] 
(2) ¿No está trabajando en este momento pero tiene trabajo? [Siga] 
(3) ¿Está buscando trabajo activamente? [TERMINA] 
(4) ¿Es estudiante? [TERMINA]
(5) ¿Se dedica a los quehaceres de su hogar? [TERMINA]
(6) Está jubilado, pensionado o incapacitado permanentemente para trabajar? [TERMINA]
(7) ¿No trabaja y no está buscando trabajo? [TERMINA]
(8) Esta recibiendo entrenamiento a través de algún programa de capacitación? [TERMINA]
 (88) NS [TERMINA] (98) NR [TERMINA] 
OCUP1A. En su ocupación principal usted es: [Leer alternativas]
 (1) ¿Asalariado del gobierno o empresa estatal? 
 (2) ¿Asalariado en el sector privado? 
(3) ¿Patrono o socio de empresa? 
(4) ¿Trabajador por cuenta propia? 
 (5) ¿Trabajador no remunerado o sin pago? 
 (88) NS (98) NR (99) INAP
OCUP1C. ¿Tiene seguro de salud o seguro social a través de su empresa o su empleador? 
(1) Sí (2) No (88) NS (98) NR (99) INAP 
Estas son todas las preguntas que tengo. Muchísimas gracias por su colaboración. 
HORA1. Hora de finalización: _____:_____AM PM
 
Yo juro que esta entrevista fue llevada a cabo con la persona indicada. 
Firma del entrevistador__________________ Fecha ____ /_____ /_____ 
Firma del supervisor de campo _________________ 
Comentarios: 
____________________________________________________________________________________ 
[No usar por PDA] Firma de la persona que digitó los datos __________________________________ 
[No usar por PDA] Firma de la persona que verificó los datos _______________________________ 
Impact Evaluation Report– Mexico 
157
Tarjeta “A” 
¿En qué grada de la escalera se siente usted en estos 
momentos? 
 10 
La mejor 
vida posible 
 9
 8
 7
 6
 5
 4
 3
 2 
1 
La peor vida 
posible 0 
Impact Evaluation Report - Mexico 
158
Tarjeta “B” 
 
 
 10 
Se puede 
confiar en la 
mayoría de 
las personas 
 9
 8
 7
 6
 5
 4
 3
 2 
1 
Hay que tener 
mucho cuidado 
cuando uno trata 
con otras 
personas 0 
 
Impact Evaluation Report– Mexico 
159
Tarjeta “C” 
 
 
 10 
La mayoría de 
las personas 
tratan de ser 
serviciales 
 9
 8
 7
 6
 5
 4
 3
 2 
1 
La mayoría de las 
personas solo se 
preocupan de sí 
mismas 0 
 
Impact Evaluation Report - Mexico 
160
Tarjeta “D” 
 
 
 10 
La mayoría de 
las personas 
tratarían de 
ser justas con 
usted 
 9
 8
 7
 6
 5
 4
 3
 2 
1 
La mayoría de 
las personas 
tratarían de 
aprovecharse 
de usted 0 
 
Impact Evaluation Report– Mexico 
161
Tarjeta “E”
 7 Mucho 
 6 
 5 
 4 
 3 
 2 
Nada 1 
Impact Evaluation Report - Mexico 
162
Tarjeta “G” 
 7 
Muy de 
acuerdo 
 6
 5 
 4
 3
 2
Muy en 
desacuerdo 1 
 
Impact Evaluation Report– Mexico 
163
Tarjeta “H”
 10 
Aprueba 
firmemente 
 9 
 8 
 7 
 6 
 5 
 4 
 3 
 2 
Desaprueba 
firmemente 1 
 
Impact Evaluation Report - Mexico 
164
Tarjeta “I” 
(00) Ningún ingreso 
(01) Menos de $480 
(02) Entre $480 - $720 
(03) Entre $721 - $960 
(04) Entre $961 - $1,440 
(05) Entre $1,441 - $2,150 
(06) Entre $2,151 - $2,870 
(07) Entre $2,871 - $3,590 
(08) Entre $3,591 - $4,310 
(09) Entre $4,311 - $6,460 
(10) Entre $6,461 - $7,180 
(11) Entre $7,181 - $8,610 
(12) Entre $8,611 - $10,050 
(13) Entre $10,051 - $11,490 
(14) Entre $11,491 - $12,920 
(15) Entre $12,921 - $14,360 
(16) Más de $14,360 
Impact Evaluation Report - Mexico 
165
Appendix C: Systematic Observation Questionnaire 
LAPOP estudio de comunidades Observación Spanish Version #F IRB Approval 
Mexico, 2014.- Cuestionario de Observación Sistemática 
© Vanderbilt University 2014. Derechos reservados. All rights reserved. 
PAIS. 
IDNUM. Número de cuestionario [asignado en la 
oficina]__________________ 
PROV. Ciudad 
(1) Cd. Juárez (2) Monterrey (3) Tijuana
MUNICIPIO. Nombre del Municipio: 
(1) Cd. Juarez (2) Monterrey (3) Tijuana 
NOMBAR. Nombre del Barrio/ 
Colonia/Comunidad__________________________ 
BLOQUE. Número polígono 
MANZANA. Número de manzana 
CUADRA. Número de cuadra dentro de la manzana 
FECHA. Fecha 
de la 
entrevista 
día: ____ 
mes:_______ 
año: 2014 
HORA. Hora 
de inicio: 
_____:_____A
M PM 
-----------
- 
Apariencia física de la cuadra – LLENAR UNA HOJA POR CADA MANZANA SELECCIONADA EN 
LA MUESTRA 
Anotar el número de veces que observa lo 
siguiente en esta cuadra del barrio: Por cuadra 
nos referimos al lado de la manzana el cual se 
esta entrevistando. 
0 1 2 3 4 5 6 7 8 9 10 o 
más 
Total 
PHD1. Botaderos de basura al aire libre 
PHD2. Basura o cristales rotos tirados en la calle 
Impact Evaluation Report - Mexico 
166
o acera
PHD6. Paradas de autobuses o microbuses en 
esta cuadra
 
PHD7. Cervecerías o tiendas para la venta de 
alcohol en esta cuadra
PHD9. Nightclubs/antros en esta cuadra 
PHD8. Graffiti o murales alusivos a pandillas
PHD4. Graffiti o murales no alusivos a pandillas
PHD5. Número de viviendas en esta cuadra que 
tienen defensas o barrotes de metal en las 
ventanas
 
FPH1. Número de botellas/latas de licor o cerveza 
vacías tiradas en la calle o acera 
 
FPH2. Número de cajetillas o colillas de cigarro 
tiradas en la calle o acera
 
FPH3. Número de carros u otros medios de 
transporte en mal estado abandonados
 
FPH9. Número de predios baldíos con pasto altos 
o tiraderos
Anotar el número de veces que observa lo 
siguiente en esta cuadra del barrio: 
FPH4. Número de casas abandonadas o 
saqueadas
 
FPH5. Número de casas con cristales o ventanas 
rotas
 
FPH6. Número de casas localizadas en sitios de 
riesgo de desastres naturales (en las cercanías de 
barrancos, ríos, zonas de deslizamientos). 
 
Anotar si observa o no lo siguiente en esta cuadra del barrio: Sí No 
FPH7.Presencia de aguas negras o residuales en las calles. 1 2 
FPH8.Lugares que generen malos olores (mataderos, fábricas, letrinas). 1 2 
FPH9. Alumbrado público o luz en la calle 1 2 
FPH10. Sectores de la cuadra por los cuales los vecinos evitan al pasar 1 2 
Material de la calle de la cuadra 
PHD3. Material de la calle en la cuadra 
1. Pavimento o asfalto 2. Cemento 3. Adoquines 4. Tierra 5. Piedra 6. Grava 7.
Observaciones de personas en la cuadra 
Anotar si observa o no lo 
siguiente en esta cuadra del 
barrio: 
Sí No 
DISO1. Personas discutiendo 
de una forma agresiva o 
violenta (hablando en un tono 
de voz muy alto, con enojo) 
1 2 
DISO2. Niños en la acera o 
calle sin la supervisión de un 
adulto 
1 2 
DISO3. Jóvenes en las calles 
sin hacer nada, vagando 
1 2 
DISO4. Presencia de jóvenes 1 2 
Recursos de la cuadra 
Anotar si esta cuadra cuenta con los 
siguientes recursos: 
Sí No 
RESO1. Alumbrado eléctrico público 1 2 
RESO2. Puesto de policía 1 2 
RESO3. Parques o áreas recreativas 1 2 
RESO4. Iglesia católica 1 2 
RESO5. Iglesia evangélica/protestante 1 2 
RESO6. Centro educativo público 1 2 
RESO7. Centro educativo privado 1 2 
RESO8. Guardería infantil 1 2 
RESO9. Hospital o puesto de salud 
público 
1 2 
Impact Evaluation Report– Mexico 
167
con tatuajes en partes visibles 
del cuerpo 
DISO5. Mendigos o personas 
pidiendo dinero 
1 2 
DISO6. Personas 
prostituyéndose en la calle 
1 2 
DISO7. Personas en estado de 
ebriedad o bajo la influencia de 
drogas en la vía pública 
1 2 
RESO10. Clínica privada 1 2 
RESO11. Casa comunal 1 2 
RESO12. Casa de la cultura 1 2 
RESO13. Otra oficina de una 
institución privada 
1 2 
RESO14. Otra oficina de una 
institución pública 
1 2 
RESO15. Cyber-café, Infocentro u 
otro establecimiento que provea 
servicios de Internet a la comunidad. 
1 2 
RESO16. Biblioteca pública 1 2 
Observación final 
TRAT1. ¿Cómo le trataron las personas residentes en esta cuadra del barrio? 
(1) No había gente alrededor 
(2) No le pusieron mucha atención 
(3) Le trataron bien, lo saludaron 
(4) Le trataron con desconfianza 
(5) Le interrogaron de forma agresiva sobre lo que andaba haciendo 
Comentarios adicionales sobre el aspecto físico/recursos de la cuadra: POR FAVOR ANOTAR 
TODO OBSERVACION ADICIONAL NO CAPTURADA POR LAS PRGUNTAS DE ESTE FORMULARIO 
QUE PUEDAN SERVIR PARA DESCRIBIR EL CONTEXTO DE LA CUADRA 
____________________________________________________________________________ 
____________________________________________________________________________ 
____________________________________________________________________________ 
____________________________________________________________________________ 
____________________________________________________________________________ 
____________________________________________________________________________ 
____________________________________________________________________________ 
____________________________________________________________________________ 
Impact Evaluation Report - Mexico 
168
Appendix D: Household Questionnaire 
LAPOP estudio de comunidades Enumeration Spanish Version # 1 IRB Approval: 110627 
Mexico, 2012 
Cuestionario de enumeración (muestra de hogares) 
© Vanderbilt University 2012. Derechos reservados. All rights reserved. 
SECCION I. LOCALIZACIÓN GEOGRÁFICA 
PAIS. 
IDNUM. Número de cuestionario [asignado en la oficina]__________________
PROV. Ciudad :
(1) Cd. Juárez 
(2) Monterrey 
(3) Tijuana 
MUNICIPIO. Nombre municipio: 
(1) Cd. Juárez 
(2) Guadalupe 
(3) Monterrey 
(4) Tijuana 
NOMBAR. Nombre del Barrio/ 
Colonia/Comunidad____________________________________________________
BLOQUE. Número de Polígono 
MANZANA. Número de manzana 
CUADRA. Número de cuadra dentro de la manzana. 
_________________________________________________ 
NUMVIVI. NÚMERO DE LA VIVIENDA DENTRO DE LA CUADRA:______ 
NUMHOGAR.NÚMERO DEL HOGAR DENTRO DE LA VIVIENDA__________ 
FECHA. Fecha de la entrevista día: ____ mes:_______ año: 2011 
Hora de inicio: _____:_____AM PM ------------ 
ATENCIÓN: ES UN REQUISITO LEER SIEMPRE LA HOJA DE CONSENTIMIENTO INFORMADO 
ANTES DE COMENZAR 
 
Impact Evaluation Report– Mexico 
169 
SECCION II. CARACTERÍSTICAS DE LA VIVIENDA 
VIV1. [No preguntar] La vivienda es:
(1) Casa independiente 
(2) Casa en fila 
(3) En edificio (condominio o apartamento) 
(4) Pieza en casa 
(5) Pieza en mesón 
(6) Rancho o choza 
(7) Otra (vivienda móvil, carpa, etc.) 
(8) Comercial [TERMINA] 
VIV2. [No preguntar] La vivienda está:
(1) Ocupada [Siga] 
(2) Desocupada, de uso ocasional [Termina] 
(3) Desocupada, en alquiler [Termina]
(4) Desocupada, en venta [Termina] 
(5) Desocupada, en reparación o construcción [Termina] 
(6) Desocupada, por otra razón [Termina]
[Para uso del entrevistador] Estado de la 
entrevista 
INT1A. 
Primer 
intento 
INT1B. 
Segundo 
intento 
INT1C. 
Tercer 
intento
INT1D. 
Cuarto 
intento
INT1E. 
Quinto 
intento
Informante apto aceptó ser entrevistado 1 1 1 1 1 
Nadie en casa 2 2 2 2 2 
No hay personas mayores de 18 años 
presentes 
3 3 3 3 3 
Sólo empleada (o) doméstica o visitante 
presente
4 4 4 4 4 
Informante no apto (mudo, enfermos 
mentales, borrachos, personas agresivas, 
etc.) 
5 5 5 5 5 
Informante no habla el idioma de la encuesta 
o es extranjero 
6 6 6 6 6 
Todos los habitantes son extranjeros 
[Terminar cuestionario]
7 7 7 7 7 
Informante presente rehusó pero indicó que 
otro miembro del hogar puede ser 
entrevistado en otro momento 
8 8 8 8 8 
Nadie en el hogar quiso ser entrevistado 9 9 9 9 9 
[Comenzar entrevista] 
VIV4. Para comenzar, me podría decir, ¿De qué material es la mayor parte de las paredes de 
la vivienda? [No leer alternativas] 
(1) Concreto,bloque, o ladrillos pegados con cemento 
(2) Bahareque, vara o caña 
(3) Adobe 
(4) Madera 
(5) Lámina metálica 
(6) Paja, palma u otro vegetal 
(7) Materiales reciclables o desechos (cartón, papel, etc.) 
(8) Otro 
(88) NS 
(98) NR 
VIV5. ¿De qué material es la mayor parte del techo de la vivienda? [No leer alternativas] 
(1) Plataforma de concreto o cemento 
Impact Evaluation Report - Mexico 
170
(2) Lámina de asbesto 
(3) Teja de barro 
(4) Lámina metálica o zinc 
(5) Paja, palma u otro vegetal 
(6) Materiales reciclables 
(7) Madera 
(8) Lamina de azulín 
(9) Otro 
(88) NS 
(98) NR 
VIV6. ¿De qué material es la mayor parte del piso? [Leer alternativas] 
(1) Cerámica 
(2) Ladrillo de cemento 
(3) Ladrillo de barro 
(4) Madera 
(5) Tierra 
(7) Otro 
(88) NS 
(98) NR 
VIV16. Esta vivienda se encuentra ubicada en una zona de riesgo como puede ser….[Leer 
alternativas] 
(1) En zona de hundimientos o deslizamientos de tierra 
(2) En zona inundable 
(3) Al pie o en la ladera de un cerro 
(4) Cerca o en la ribera de un río/lago/laguna/mar 
(5) Cerca de un basurero público 
(6) Otros 
(7) La vivienda no se encuentra en una zona de riesgo 
(88) NS 
(8) (98) NR
Podría decirme si en esta vivienda tienen: [Leer todos]
R1. Televisor (0) No (1) Sí 
R3. Refrigeradora (0) No (1) Sí 
R4. Teléfono convencional/fijo (no celular) (0) No (1) Sí 
R4A. Teléfono celular (0) No (1) Sí 
R5. Carro ¿Cuántos? (0) No (1) 
Uno 
(2) 
Dos (3) Tres o más 
R6. Lavadora de ropa (0) No (1) Sí 
R7. Microondas (0) No (1) Sí 
R8. Motocicleta (0) No (1) Sí 
R14. Cuarto de baño dentro de la casa (0) No (1) Sí 
R19. Ducha para agua caliente (0) No (1) Sí 
R15. Computadora (0) No (1) Sí 
R16. Televisor de pantalla plana (0) No (1) Sí 
R18. Servicio de internet (0) No (1) Sí 
HOG1. ¿Cuántas personas habitan en esta vivienda contando los niños pequeños, los ancianos, y otras 
personas que duermen aquí? [Anotar Número] ___________ 
HOG1a. ¿Cuántas de las personas que habitan esta vivienda tienen entre 6 y 17 años? 
[Anotar Número] ___________ 
Impact Evaluation Report– Mexico 
171 
HOG4. ¿Cuántas de las personas que habitan esta vivienda tienen 18 años o más? 
[Anotar Número] ___________
HOG2. ¿Todas las personas que habitan en esta vivienda comparten un mismo gasto para la comida? 
(1) Sí (2) No 
HOG3. ¿Cuántos hogares o grupos de personas tienen gasto separado para la comida contando el de 
usted? 
[Anotar Número] ___________ 
HOGID. Número de orden del hogar dentro de la vivienda : 
1 2 3 4 5 6 7 8 9 10 
LAS SIGUIENTES PREGUNTAS APLICAN A CADA UNO DE LOS HOGARES DENTRO DE LA 
VIVIENDA. SI EN ESTA VIVIENDA HAY MAS DE UN HOGR DEBERÁ DE COMPPLETAR UN 
FORMALURARIO POR CADA UNO DE ELLOS. EN ESTOS FORMULARIOS LA SECCIÓN I ES LA 
MISMA PERO HABRÁ QUE APLICAR POR SEPARADO LA SECCIÓN II. 
SECCION III. HOGARES 
VIV7. ¿De cuántos cuartos dispone este hogar sin contar baño, corredor, cocina ni garaje? 
_____[Anote uno si no hay divisiones en la vivienda y pase a VIV9]
(88) NS (98) NR
VIV8. Del total de cuartos, ¿Cuántos se utilizan exclusivamente para dormir (como 
dormitorios)? ________ 
(88) NS (98) NR (99) INAP 
VIV9. ¿Cuál es la forma de tenencia de la vivienda/cuarto que ocupa este hogar? Esta 
vivienda/cuarto es [Leer alternativas]: 
(1) Propia 
(2) Propia pagándose a plazos 
(3) Alquilada 
(4) La ocupa gratuitamente 
(7) Otra forma de tenencia 
 (88) NS (98) NR 
VIV10. ¿Cómo se deshacen de las aguas negras en este hogar? [Leer alternativas] 
(1) Por alcantarilla
(2) Por pozo séptico
(3) Por descarga a río, laguna o mar
(4) Por calle o al aire libre
(5) Otra forma
(88) NS (98) NR 
VIV11. ¿Cómo se deshacen de la basura en este hogar? [Leer alternativas]
(1) Hay servicio municipal de recolección de basura 
(2) Hay servicio privado de recolección de basura 
(3) La depositan en contenedores 
(4) La entierran 
(5) La queman 
(6) En botadero al aire libre 
(7) La tiran en cualquier lugar 
(8) Otro 
(88) NS (98) NR 
Impact Evaluation Report - Mexico 
172
VIV12. ¿De dónde proviene el agua que utilizan en este hogar? [Leer alternativas] 
(1) Cañería dentro de la vivienda [Siga] 
(2) Cañería fuera de la vivienda pero dentro de la propiedad [Siga]
(3) Cañería del vecino[Siga] 
(4) Llave pública [Siga] 
(5) Pozo [Pase a VIV14] 
(6) Carro cisterna [Pase a VIV14] 
(7) Pick –up con barriles
(8) Ojo de agua, río, riachuelo o manantial [Pase VIV14] 
(9) Agua lluvia [Pase a VIV14] 
(10) Otro [Pase a VIV14] 
(88) NS [Pase a VIV14] (98) NR [Pase a VIV14] 
VIV13. ¿Con qué frecuencia le llega el agua? [Leer alternativas] 
(1) Todos los días 
(2) Cada dos días 
(3) Cada tres días 
(4) Una vez a la semana 
(5) Una vez cada dos semanas 
(6) Una vez al mes 
(88) NS (98) NR (99) INAP 
VIV14. ¿Qué tipo de alumbrado utilizan principalmente en este hogar? [Leer alternativas] 
(1) Electricidad 
(2) Candil o lámpara de gas o kerosene 
(3) Vela 
(4) Energía solar 
(5) Batería de carro 
(6) Otro 
(88) NS (98) NR
Impact Evaluation Report– Mexico 
173 
SECCION III. MIEMBROS DEL HOGAR Y CARACTERÍSTICAS SOCIODEMOGRÁFICAS [No olvide registrar a los miembros ausentes temporales por un período menor de 4 meses, recién nacidos y 
domésticas que duermen en el hogar, estudiantes en el extranjero y/o enfermos financiados por el hogar] 
ID. 
N 
Ú 
M 
E 
R 
O 
D 
E 
O 
R 
D 
E 
N 
HM01. Por 
favor, dígame 
el primer 
nombre de 
las personas 
que forman 
parte de su 
hogar, 
empezando 
por el jefe o 
jefa del hogar. 
HM02. 
¿Qué relación de 
parentesco tiene 
(NOMBRE)…con 
el jefe(a) del 
hogar? 
(1) Jefe (a) 
(2) Esposo(a), 
compañero(a) 
(3) Hijo(a) 
(4) Padre/madre 
(5) Yerno/nuera 
(6) Nieto(a) 
/bisnieto(a) 
(7) Hermano(a) 
(8) Sobrino(a) 
(9) Padre/madre 
(10) Suegro(a) 
(11) Otros 
familiares (tío, 
etc.) 
(12) Empleada 
doméstica 
(13) Otra 
persona no 
pariente 
(88) NS 
(98) NR
HM03. 
[ANOTAR, 
NO 
PREGUNTE]: 
(1) Hombre 
 (2) Mujer 
HM04. 
¿Qué edad 
tiene 
(NOMBRE) 
en años 
cumplidos? 
Anote 00 
para los 
menores de 
un año. 
(888) NS 
(998) NR
HM08. [Solo 
para 
mayores de 
5 años] 
¿(Nombre) 
asiste o 
asistió a un 
centro de 
educación 
formal? 
(1) Asiste 
actualmente 
o esta de 
vaciones 
peropiensa 
continuar 
[Pasar a 
HM05] 
(2) No asiste 
actualmente, 
pero asistía 
antes [Pasar 
a HM09]
(3) Nunca 
asistió Pasar 
a HM09]
(88) NS 
[Pase a 
HM06] 
(98) NR 
[Pase a 
HM06] 
HM09. ¿Por qué razón 
no asiste, dejó de 
estudiar o no estudio? 
[Leer opciones] 
(1) Porque finalizó sus 
estudios/ el grado 
(2) Porque no quiere o 
quiso estudiar 
(3) Porque realizar o 
ayudar en 
[quehaceres del 
hogar 
(4) Porque el centro 
educativo queda 
lejos 
(5) Por problemas 
familiares o de salud 
(6) Por falta de recursos 
económicos 
(7) Porque está 
trabajando 
(8) Por problemas de 
conducta o 
rendimiento en la 
escuela 
(9) Por embarazo 
(10) Por otro motivo 
HM09B. Aunque no 
estudia actualmente 
piensa retomar sus 
estudios el próximo 
año? 
(1) Si 
(2) No 
(88) NS 
(99) NR 
Si HM08=3 
pasar a HM06
HM05.Solo para los 
que contestaron que 
asisten o asistió 
HM08=1 
HM08=2 
¿Cuál fue el último 
grado o nivel de 
educación que 
(NOMBRE) completó 
o aprobó? 
(0) Ninguno 
(1) Kinder 
(2) Primaria 
inccompleta 
(3) Primaria completa 
(4) Secundaria 
(5) Preparatoria 
(6) Carrea técnica 
(7) Universidad 
(8)Postgrado 
(9) Educación 
especial [para 
personas con 
discapacidades] 
(10) Programa de 
alfabetización 
(11) Otros 
(88) NS 
(98) NR 
(99) INAP 
HM05a. [Solo 
para los que 
contestaron que 
asisten o asistió 
HM08=1 
HM08=2] 
A través de qué 
sistema estudia o 
estudio? 
(1) Escuela pública 
(2) Escuala 
privada 
(3) CONALEP 
(4) Telesecundaria 
(5) Técnica 
superior 
(6) Otro ________ 
(88) NS 
(98) NR 
HM06. [Solo 
para mayores 
de 10 años] 
¿Cuál es el 
estado civil de 
(NOMBRE)? 
(1) Soltero 
(2) Casado 
(3) Unión libre 
(acompañado) 
(4) Divorciado 
(5) Separado 
(6) Viudo 
(88) NS 
(98)NR 
HM07. [Solo para
10 años] 
¿A qué se dedica 
principalmente? [L
alternativas] 
(1) ¿Está Trabajan
(2) ¿No está traba
momento pero tien
(3) ¿Está buscand
activamente? 
(4) ¿Es estudiante
(5) ¿Se dedica a lo
quehaceres de su 
(6) ¿Está jubilado,
o incapacitado 
permanentemente
trabajar? 
(7) ¿No trabaja y n
buscando trabajo?
(88) NS (98)NR 
01 
02 
03 
04 
05 
06 
07 
08 
09 
10 
11 
12 
Impact Evaluation Report - Mexico 
174
SECCION IV. INGRESOS Y MIGRACION 
Q10D. El salario o sueldo que usted recibe y el total del ingreso familiar: [LEER ALTERNATIVAS] 
(1) Les alcanza bien y pueden ahorrar 
(2) Les alcanza justo sin grandes dificultades 
(3) No les alcanza y tienen dificultades 
(4) No les alcanza y tienen grandes dificultades 
(88) [No leer] NS 
(98) [No leer] NR 
Q10A. ¿Usted o alguien que vive en su casa recibe remesas (ayuda económica) del exterior? 
(1) Sí (2) No (88) NS (98) NR 
Q10C. [Preguntar a todos] ¿Tiene usted familiares cercanos que antes vivieron en esta 
casa y que hoy estén residiendo en el exterior? [Si dijo “Sí”, preguntar ¿en dónde?] 
[No leer alternativas] 
(1) Sí, en los Estados Unidos solamente 
(2) Sí, en los estados Unidos y en otros países 
(3) Sí, en otros países (no en estados Unidos) 
(4) No 
(88) NS 
(98) NR 
Q10E. ¿De todos los miembros del hogar, qué persona contribuye en mayor medida al sostenimiento del 
hogar? 
Anotar número de orden: __________________ 
Estas son todas las preguntas que tengo. Muchísimas gracias por su colaboración. Es posible 
que visitemos su hogar nuevamente en los próximos días para saber su opinión sobre lo que 
pasa en el país y en su comunidad. Le agradecemos de antemano su colaboración. 
Hora terminada la entrevista _______ : ______ TI 
TI. Duración de la entrevista [minutos, ver página # 1] _____________ 
Yo juro que esta entrevista fue llevada a cabo con la persona indicada y el lugar indicado. 
Firma del entrevistador__________________ Fecha ____ /_____ /_____ 
 Firma del supervisor de campo _________________ 
Comentarios: 
____________________________________________________________________________________________ 
____________________________________________________________________________________________ 
Firma de la persona que digitó los datos __________________________________ 
Firma de la persona que verificó los datos _______________________________ 
 
Impact Evaluation Report– Mexico 
175 
Appendix E. Pre-Visit Questionnaire 
 
Guía para la realización de visitas previas 
Fecha de la visita: Día: Mes: Año: Hora Inicio: ____ AM ____ PM 
 
 Hora Finalización: ____ AM ____ PM 
I. Ubicación 
MUN.Municipio: 
(1) Ciudad Juarez 
(2) Monterrey (quizas en monterrey se 
incluyan otros municipios 
conurbados) 
(3) Tijuana 
POL.Numero de poligono: _____ 
TIPO. Tipo de polígono
(1) Trtamiento 
(2) Control 
(3) De no riesgo 
ALD. Aldea: _________ CAN. Caserío: ______________ BAR. Barrio: ________________ 
Impact Evaluation Report - Mexico 
176
II. Identificación del polígono 
ID. Número asignado en oficina [Basado en lista de polígonos]: _____________________________ 
NOM.Nombre completo de la comunidad: __________________________________________________ 
NOM1.Nombre como es conocida la comunidad entre los pobladores: ___________________________ 
DIR. Dirección exacta: 
___________________________________________________________________ 
___________________________________________________________________________________
__ 
DIR1.Cómo llegar al polígono: 
DIR2. ¿Tiene este polígono más de dos colonias? 
(1) Sí [Siga]
(2) No [Pasar a CH1]
DIR3. ¿Cuántas colonias tiene este polígono?: [escribir número] 
DIR4. ¿Están las diferentes etapas una a la par de otra, es decir forman un sólo bloque? 
(1) Sí 
(2) No 
Observaciones:____________________________________________________________________
________________________________________________________________________________
________________________________________________________________________________ 
PARA CADA COLONIA DEL POLÍGONO COMPLETAR LOS SIGUEITNES DATOS 
III. Características de CADA colonias 
CH1. Nombre de la colonia _________________________________________ 
Forma de acceso a la comunidad: Sí No 
CH2.Peatonal 1 
CH3.Vehicular 1 
CH4.Ruta de buses, microbuses 1 
Vigilancia 
Hay en la comunidad: 
CH5. Vigilante/Seguridad privada 
2 
CH6. Portón de seguridad privada para entrar a la comunidad 2 
LIM. Que calles delimitan la colonia? 
 Nombre de la calle 
…………………………………….. 
Nombre de la calle 
………………………
…… 
[Nombre de la comunidad] 
Nombre de la calle 
………………………
…… 
Nombre de la calle 
…………………………………………… 
Impact Evaluation Report - Mexico 
177
NVIV. En promedio, ¿cuántas viviendas estima que hay en esta colonio? ____________[Si mayor de 
500 viviendas, contestar siguiente pregunta; si menor de 500 viviendas pasar a CHVIV] 
SECTORES1. Es posible sectorizar la colonia y por lo tanto establecer zonas dentro del barrio? 
(1) SÍ [Siga] (2) No [Pasar a CHVIV], [Sondear con líderes locales] Cuales son esos sectores? 
[Listar y enumerar] 
SECTOR2. ¿Cuáles son esos sectores? SECTOR3. ¿Cuántas viviendas 
aproximadamente comprende cada sector? 
1. 
2. 
3. 
4. 
5. 
CHVIV. Anotar características de la mayoría de las viviendas de esta colonia: 
1) Viviendas en edificios o multifamiliares 2) Casas independientes 3) Algunas en edificios, otras 
independientes 
OTR. ¿Hay otras comunidades a la par o contiguas a esta comunidad? 
(0) No_____ 
(1) Sí_____[Listar nombre (s) abajo] 
Comunidad Ubicación 
 
Por favor indicar si esta colonia cuenta con los siguientes recursos y la cantidad: 
No Sí 
CH5. Canchas (adentro o en las afueras de la comunidad que son 
accesibles a pie) 00 1 2 3 o más 
CH6. Iglesias o templos de cualquier denominación religiosas (adentro o 
en las afueras de la comunidad que son accesibles a pie) 0 1 2 3 o más 
CH7. Escuelas públicas (adentro o en las afueras de la comunidad que 
son accesibles a pie) 0 1 2 3 o más 
CH8. Escuelas/colegios privados adentro o en las afueras de la 
comunidad que son accesibles a pie) 0 1 2 3 o más 
CH9. Centros de salud (adentro o en las afueras de la comunidad que 
son accesibles a pie) 0 1 2 3 o más 
CH10. Casa comunal (adentro o en las afueras de la comunidad que es 
accesible a pie) 00 1 2 3 o más 
CH11. Bibliotecas públicas (adentro o en las afueras de la comunidad 
que es accesible a pie) 00 1 2 
CH12. Plazas o parques públicos (adentro o en las afueras de la 
comunidad que es accesible a pie) 00 1 2 3 o más 
CH13.Otros: 
Impact Evaluation Report - Mexico 
178
JDIR. ¿Hay una organización de vecinos o patronato en esta colonia? 
(1) Sí, una organización de vecinos o patronato……………...Nombre: 
(2) Sí, más de una organización de vecinos o patronato….Nombres: 
(3) No 
ORG. ¿Hay alguna institución del Estado, organización no gubernamental nacional o internacional que 
estén actualmente implementando programas dirigidos a los niños y/o jóvenes de la comunidad? 
(4) Sí [Escribir nombre (s) de las instituciones] 
(5) No 
Nombre(s): 
Experiencia en la comunidad 
EC1. ¿Para realizar la visita tuvo el acompañamiento de alguien que reside en la polígono/colonia (un 
líder comunitario, un miembro de la junta directiva, o vecino residente en el lugar)? 
(1) Sí 
(2) No 
EC2. En general, ¿Cómo le recibió la gente de la colonia? 
EC4. Teniendo en cuenta el estrato socio-económico de otras comunidades/colonias en este 
municipio, según lo que pudo observar, diría usted que en esta colonia/comunidad residen 
principalmente personas de: 
(1) Clase alta 
(2) Clase media alta 
(3) Clase media 
(4) Clase media baja 
(5) Clase baja 
(6) Hay de todos los niveles Socioeconómicos 
EC5. ¿Cuál es la principal lengua que se habla en la colonia? (escriba la que es la más hablada—que la 
mayoría de personas utilizan) 
EC6. ¿Existen otras lenguas que también se hablan? (Escribir todas) 
EC7.Observaciones adicionales: 
___________________________________________________________________________________
_ 
___________________________________________________________________________________
_ 
Si hay Escuela Pública en la Comunidad visítela y anote: 
Tipo de educación 
Cantid
ad 
maestr
os 
Total 
Estudia
ntes 
Turnos: 
Matutino=1 
Vespertino=
2 
Nocturno=3 
Turno 1y 
2=4 
Todos=5 
Anote el 
código 
Total 
estu
diant
es 
Primar
ia 
Secund
aria 
amb
as 
CH7. Si hay Escuelas 1 2 3 
Impact Evaluation Report - Mexico 
179
públicas dentro de la 
comunidad 
Nombre de la Escuela: _________________________________________________ 
Teléfono:_____________________________ 
Dirección:_____________________________________ 
Nombre del director:____________________________ 
Nombre de la Escuela: _________________________________________________ 
Teléfono:_____________________________ 
Dirección:_____________________________________ 
Nombre del director:____________________________ 
Si hay Iglesia en la Comunidad visítela y anote: 
Nombre de la Iglesia Nombre del Pastor o 
Padre 
Dirección Teléfono 
 
Impact Evaluation Report - Mexico 
180
Appendix F: Focus Group Discussion Guide 
A continuación se presentan las guías de discusión que se usaron en las tres ciudades en la 
segunda ronda (2014) para los cuatro diferentes grupos de discusión focalizada. 
IRB# 091334
Estudio de comunidades MEXICO 2014 
© Vanderbilt University 2011. Derechos reservados. All rights reserved. 
Cuestionario/Guía de Discusión para el 
Grupo Focal de Jóvenes que se Consideran en Riesgo 
(Jóvenes Riesgo Menor) 
Introducción General: 
Dar a conocer la mecánica de trabajo a los participantes 
Establecer el rapport con los participantes: compenetración con ellos, crear bases para un 
ambiente de mayor apertura, tranquilidad y confianza de los participantes. 
FG de Jóvenes 
INTRODUCCIÓN (5 min) 
Bienvenida 
Gracias por haber venido a esta reunión para platicar sobre sus opiniones, experiencias de la ciudad y 
temas relacionados. 
Todas las respuestas y comentarios son importantes. No hay respuestas buenas ni malas. 
Participación de todos 
Grabación de audio CONFIDENCIAL –sólo para fines de este trabajo￾PRIMERA INTERVENCIÓN DE CADA PARTICIPANTE (3 min) 
- actividades cotidianas y pasatiempos preferidos 
PROBLEMÁTICA GENERAL (10 mins) 
1. En su opinión, ¿cuáles son los problemas más graves de los jóvenes como usted, aquí en 
mencionar CIUDAD? 
2. ¿Qué se debería de hacer para resolver estos problemas? (o sea, los problemas de X, Y, Z--los que 
han sido nombrados)
3. ¿Hay organismos en mencionar CIUDAD que han hecho esfuerzos para resolver estos problemas? 
¿Qué organizaciones han hecho algo? ¿Qué han hecho, específicamente? ¿Estos esfuerzos, hasta 
qué punto han logrado solucionar los problemas? 
Impact Evaluation Report - Mexico 
181
VIOLENCIA EN ESCUELAS (25 mints) 
SI HAY POCA CONVERSACIÓN SOBRE VIOLENCIA EN ESCUELAS EXPLORAR 
PREGUNTAS SIMILARES SOBRE VIOLENCIA EN SU COLONIA Y LIGAR CON 
SIGUIENTE SECCIÓN 
4. Hablando de la violencia y de los delitos que la gente de mencionar CIUDAD dice que son tan 
comunes hoy en día, en la escuela primaria y secundaria que usted ha asistido, ¿qué tan serio era 
el problema de la violencia?
5. En las escuelas que usted ha asistido, ¿qué tipo de violencia había? Eran los alumnos mismos que 
hicieron actos violentos, o eran jóvenes de afuera, que no eran alumnos de la escuela de usted?
6. Y los delitos que se cometieron en la escuela, ¿qué tipos de delitos eran?
7. Los que cometieron estos actos de delincuencia, eran miembros de alguna pandilla?
8. ¿Cuántas pandillas había en su escuela o colegio? ¿Cómo se llamaban?
9. ¿Cómo se sabía que cierto alumno y cierta alumna eran miembros de una pandilla? ¿Qué señas 
mostraban que pertenecían a alguna pandilla?
10. ¿Qué hacía el director o directora de la escuela o del colegio cuando se enteraba de que se había 
cometido un delito o algún acto de violencia? (¿Me puede dar ejemplos de algunos casos que 
usted recuerda?)
11. ¿Ha sido víctima usted de algún acto de delincuencia en su escuela o colegio? ¿Me puede contar 
el incidente? ¿Cómo se resolvió esto? ¿Qué pasó con el agresor? 
12. En la escuela o colegio que usted había asistido, ¿se expulsaba a los y las estudiantes violentos? 
¿Eran expulsiones permanentes? ¿Qué pasó con estos estudiantes después de su expulsión—se 
trasladaron a otra escuela u otro colegio, o dejaron de estudiar?
13. En su escuela o colegio, ¿se hacía algo para prevenir la delincuencia?
14. En su escuela o colegio, los alumnos y las alumnas tenían algún rol en mediar los conflictos entre 
alumnos? ¿Qué hacían ellos en cuanto a la mediación?
15. ¿Hacía algo la escuela o el colegio para prevenir que los alumnos y las alumnas se metieran en las 
pandillas? ¿Qué hacía?
16. ¿Qué hacen las pandillas para convencerle a un o una joven que se meta en la pandilla?
17. ¿En su escuela o colegio, había charlas sobre el uso de las drogas narcóticas? ¿Quiénes daban 
estas charlas?
CRIMEN ORGANIZADO (25 mints) 
18. ¿Qué tanta actividad del crimen organizado dirían que hay en mencionar CIUDAD?
19. Y en sus colonias, qué tanta actividad? ¿En términos comparativos dirían que viven en colonias 
que son de las que hay mucha o poca actividad del crimen organizado?
20. ¿Qué grupo hay? ¿Qué actividades realizan principalmente?
21. ¿La gente les tiene respeto o miedo o en muchos casos esta involucrados con ellos?
22. ¿Cómo se sabía que cierto joven de su colonia es miembros de un grupo de crimen organizado?
23. ¿Cómo afectan, que impacto tiene el crimen organizado en las colonias donde viven?
24. ¿Resulta atractivo para los jóvenes pertenecer o apoyar las actividades de grupos de crimen 
organizado? ¿Por qué? ¿Qué ventajas o desventajas tiene?
POLICIA Y SISTEMA DE JUSTICIA (20 mints) 
25. ¿La policía estatal hacía algo para proteger a los y las estudiantes de los pandilleros o grupos de 
crimen organizado? ¿Qué hacía la policía?
26. ¿Cómo trata la policía a los jóvenes de mencionar CIUDAD? (pedir ejemplos) desconfía
Impact Evaluation Report - Mexico 
182
27. ¿Tiene usted confianza o desconfianza en la policía? ¿Por qué dice esto?
28. ¿Cómo han sido las experiencias que usted o sus amigos han tenido con la policía?
29. Con respecto a los juzgados y las experiencias de los jóvenes, qué impresión tiene usted de los 
jueces de mencionar CIUDAD y de la justicia que se provee a los jóvenes?
30. Y los fiscales de mencionar CIUDAD, ¿cómo son ellos en su forma de tratar a los jóvenes?
31. ¿Qué sabe usted de las leyes que tienen que ver con la delincuencia juvenil? ¿A qué edad se 
manda a un condenado a una prisión para jóvenes y a qué edad se le manda a una prisión para 
adultos? 
SITUACIÓN FAMILIAR Y TRABAJO (30 mints) 
32. Hablando ahora de su familia, ¿Cómo es la situación de su hogar? ¿Está satisfecho con el nivel 
de apoyo que su mamá y su papá le han dado a usted y a sus hermanos?
33. ¿Qué le hace falta para poder lograr sus metas?
34. ¿Hay alguna institución en mencionar CIUDAD (el municipio, la iglesia, la escuela, una ONG), o 
algún individuo, que le ha ayudado a conseguir lo que necesite? ¿Qué institución o persona será? 
¿En qué forma le han ayudado?
35. ¿Tiene usted un trabajo en este momento? (Si dice que “no” preguntar, “por qué no?”. Si dice 
que “sí”, preguntar: ¿Qué tipo de trabajo es? ¿Está contento con este trabajo, o preferiría tener un 
trabajo mejor? 
( ¿Qué tipo de trabajo le gustaría tener? ¿Hay obstáculos que le previenen conseguir tal trabajo? 
¿Cuáles son estos obstáculos?)
36. ¿Ha completado usted sus estudios, o sea, se ha graduado del colegio? (Si dicen que “no”, 
preguntar, “Por qué dejó sus estudios?”)
37. ¿Hay en mencionar CIUDAD oportunidades para completar los estudios, para los que ya han 
dejado de estudiar en las escuelas o los colegios?
38. ¿Qué tipos de trabajo ha tenido usted desde que dejó de estudiar? ¿Qué tendría que hacer para 
poder conseguir un trabajo mejor?
39. ¿Ha tenido usted problemas en completar sus estudios o en conseguir trabajo por el hecho de que 
usted habla una lengua indígena mejor que el español, o porque tiene problemas en expresarse en 
español? 
CIERRE (15 mints) 
40. ¿Siente algún cambio en la cuestión educativa /escolar en los últimos meses/años?
41. ¿Considera que –si es que hay- algunos cambios en la sociedad, en la escuela, en su comunidad 
que hayan influido en estos cambios? 
42. ¿Cree que el gobierno o alguna otra institución ha implementado programas o planes para mejorar 
la condición social en su comunidad? 
43. ¿Cómo se siente con respecto a su entorno inmediato y la relación que tiene con su vida diaria en 
el aspecto económico, social y familiar? 
44. ¿Conoce usted algún o algunos programas que ha implementado el gobierno o alguna otra 
institución para mejorar la situación en su comunidad? 
45. ¿Ha participado usted en algún o algunos programas para mejorar su vida en general? 
46. ¿Conoce usted a alguien que haya participado en estos programas? 
47. ¿Qué opinión tiene usted de estos programas que se han implementado en su comunidad? 
48. ¿Hay algo que quiere agregar a la conversación, algo que no se ha comentado hasta el momento?
Impact Evaluation Report - Mexico 
183
IRB# 091334
Estudio de comunidades MEXICO 2014 
© Vanderbilt University 2011. Derechos reservados. All rights reserved. 
Cuestionario/Guía de Discusión para el 
Grupo Focal de Jóvenes que se Consideran en Riesgo 
(Jóvenes Riesgo Mayor) 
Introducción General: 
Dar a conocer la mecánica de trabajo a los participantes 
Establecer el rapport con los participantes: compenetración con ellos, crear bases para un 
ambiente de mayor apertura, tranquilidad y confianza de los participantes. 
FG de Jóvenes 
INTRODUCCIÓN (5 mints) 
Bienvenida 
Gracias por haber venido a esta reunión para platicar sobre sus opiniones, experiencias de la ciudad y 
temas relacionados. 
Todas las respuestas y comentarios son importantes. No hay respuestas buenas ni malas. 
Participación de todos 
Grabación de audio CONFIDENCIAL –sólo para fines de este trabajo￾PRIMERA INTERVENCIÓN DE CADA PARTICIPANTE (3 mints) 
- actividades cotidianas y pasatiempos preferidos 
PROBLEMÁTICA GENERAL 10 minutos) 
49. En su opinión, ¿cuáles son los problemas más graves de los jóvenes como usted, aquí en 
mencionar CIUDAD? 
50. ¿Qué se debería de hacer para resolver estos problemas? (o sea, los problemas de X, Y, Z--los 
que han sido nombrados)
51. ¿Hay organismos en mencionar CIUDAD que han hecho esfuerzos para resolver estos 
problemas? ¿Qué organizaciones han hecho algo? ¿Qué han hecho, específicamente? ¿Estos 
esfuerzos, hasta qué punto han logrado solucionar los problemas? 
VIOLENCIA EN ESCUELAS (25 minutos) 
SI HAY POCA CONVERSACIÓN SOBRE VIOLENCIA EN ESCUELAS EXPLORAR 
PREGUNTAS SIMILARES SOBRE VIOLENCIA EN SU COLONIA Y LIGAR CON 
SIGUIENTE SECCIÓN 
Impact Evaluation Report - Mexico 
184
52. Hablando de la violencia y de los delitos que la gente de mencionar CIUDAD dice que son tan 
comunes hoy en día, en la escuela primaria y secundaria que usted ha asistido, ¿qué tan serio era 
el problema de la violencia?
53. En las escuelas que usted ha asistido, ¿qué tipo de violencia había? Eran los alumnos mismos 
que hicieron actos violentos, o eran jóvenes de afuera, que no eran alumnos de la escuela de 
usted?
54. Y los delitos que se cometieron en la escuela, ¿qué tipos de delitos eran?
55. Los que cometieron estos actos de delincuencia, eran miembros de alguna pandilla?
56. ¿Cuántas pandillas había en su escuela o colegio? ¿Cómo se llamaban?
57. ¿Cómo se sabía que cierto alumno y cierta alumna eran miembros de una pandilla? ¿Qué señas 
mostraban que pertenecían a alguna pandilla?
58. ¿Qué hacía el director o directora de la escuela o del colegio cuando se enteraba de que se había 
cometido un delito o algún acto de violencia? (¿Me puede dar ejemplos de algunos casos que 
usted recuerda?)
59. ¿Ha sido víctima usted de algún acto de delincuencia en su escuela o colegio? ¿Me puede contar 
el incidente? ¿Cómo se resolvió esto? ¿Qué pasó con el agresor? 
60. En la escuela o colegio que usted había asistido, ¿se expulsaba a los y las estudiantes violentos? 
¿Eran expulsiones permanentes? ¿Qué pasó con estos estudiantes después de su expulsión—se 
trasladaron a otra escuela u otro colegio, o dejaron de estudiar?
61. En su escuela o colegio, ¿se hacía algo para prevenir la delincuencia?
62. En su escuela o colegio, los alumnos y las alumnas tenían algún rol en mediar los conflictos entre 
alumnos? ¿Qué hacían ellos en cuanto a la mediación?
63. ¿Hacía algo la escuela o el colegio para prevenir que los alumnos y las alumnas se metieran en 
las pandillas? ¿Qué hacía?
64. ¿Qué hacen las pandillas para convencerle a un o una joven que se meta en la pandilla?
65. ¿En su escuela o colegio, había charlas sobre el uso de las drogas narcóticas? ¿Quiénes daban 
estas charlas?
CRIMEN ORGANIZADO (25 minutos) 
66. ¿Qué tanta actividad del crimen organizado dirían que hay en mencionar CIUDAD?
67. Y en sus colonias, qué tanta actividad? ¿En términos comparativos dirían que viven en colonias 
que son de las que hay mucha o poca actividad del crimen organizado?
68. ¿Qué grupo hay? ¿Qué actividades realizan principalmente?
69. ¿La gente les tiene respeto o miedo o en muchos casos esta involucrados con ellos?
70. ¿Cómo se sabía que cierto joven de su colonia es miembros de un grupo de crimen organizado?
71. ¿Cómo afectan, que impacto tiene el crimen organizado en las colonias donde viven?
72. ¿Resulta atractivo para los jóvenes pertenecer o apoyar las actividades de grupos de crimen 
organizado? ¿Por qué? ¿Qué ventajas o desventajas tiene?
SECCION ÚNICA. 
A. ¿Quién nos puede comentar sobre algún evento que haya vivido más cercanamente? 
B. ¿Cómo fue? ¿a qué se debió? 
C. ¿cómo se pudo resolver? ¿cuánto tiempo tardó en resolverlo? 
D. DEJAR QUE LOS PARTICPANTES COMENTEN SOBRE SUS DIFERENTES 
EXPERIENCIAS. 
POLICIA Y SISTEMA DE JUSTICIA (20 mints) 
73. ¿La policía estatal hacía algo para proteger a los y las estudiantes de los pandilleros o grupos de 
crimen organizado? ¿Qué hacía la policía?
Impact Evaluation Report - Mexico 
185
74. ¿Cómo trata la policía a los jóvenes de mencionar CIUDAD? (pedir ejemplos) desconfía
75. ¿Tiene usted confianza o desconfianza en la policía? ¿Por qué dice esto?
76. ¿Cómo han sido las experiencias que usted o sus amigos han tenido con la policía?
77. Con respecto a los juzgados y las experiencias de los jóvenes, qué impresión tiene usted de los 
jueces de mencionar CIUDAD y de la justicia que se provee a los jóvenes?
78. Y los fiscales de mencionar CIUDAD, ¿cómo son ellos en su forma de tratar a los jóvenes?
79. ¿Qué sabe usted de las leyes que tienen que ver con la delincuencia juvenil? ¿A qué edad se 
manda a un condenado a una prisión para jóvenes y a qué edad se le manda a una prisión para 
adultos? 
SITUACIÓN FAMILIAR Y TRABAJO (30 mints) 
80. Hablando ahora de su familia, ¿Cómo es la situación de su hogar? ¿Está satisfecho con el nivel 
de apoyo que su mamá y su papá le han dado a usted y a sus hermanos?
81. ¿Qué le hace falta para poder lograr sus metas?
82. ¿Hay alguna institución en mencionar CIUDAD (el municipio, la iglesia, la escuela, una ONG), o 
algún individuo, que le ha ayudado a conseguir lo que necesite? ¿Qué institución o persona será? 
¿En qué forma le han ayudado?
83. ¿Tiene usted un trabajo en este momento? (Si dice que “no” preguntar, “por qué no?”. Si dice 
que “sí”, preguntar: ¿Qué tipo de trabajo es? ¿Está contento con este trabajo, o preferiría tener un 
trabajo mejor? 
( ¿Qué tipo de trabajo le gustaría tener? ¿Hay obstáculos que le previenen conseguir tal trabajo? 
¿Cuáles son estos obstáculos?)
84. ¿Ha completado usted sus estudios, o sea, se ha graduado del colegio? (Si dicen que “no”, 
preguntar, “Por qué dejó sus estudios?”)
85. ¿Hay en mencionar CIUDAD oportunidades para completar los estudios, para los que ya han 
dejado de estudiar en las escuelas o los colegios?
86. ¿Qué tipos de trabajo ha tenido usted desde que dejó de estudiar? ¿Qué tendría que hacer para 
poder conseguir un trabajo mejor?
87. ¿Ha tenido usted problemas en completar sus estudios o en conseguir trabajo por el hecho de que 
usted habla una lengua indígena mejor que el español, o porque tiene problemas en expresarse en 
español? 
88. ¿Hay algo que quiere agregar a la conversación, algo que no se ha comentado hasta el momento?
CIERRE (15 mints) 
89. ¿Siente algún cambio en la cuestión educativa /escolar en los últimos meses/años?
90. ¿Cree que el crimen organizado ha aumentado o disminuido con los programas de gobierno u 
otras instituciones?
91. ¿Se han incrementado el número de pandillas o grupos de jóvenes con fines de violencia?
92. ¿Considera que –si es que hay- algunos cambios en la sociedad, en la escuela, en su comunidad 
que hayan influido en estos cambios? 
93. ¿Cree que el gobierno o alguna otra institución ha implementado programas o planes para 
mejorar la condición social en su comunidad? 
94. ¿Cómo se siente con respecto a su entorno inmediato y la relación que tiene con su vida diaria en 
el aspecto económico, social y familiar? 
95. ¿Conoce usted algún o algunos programas que ha implementado el gobierno o alguna otra 
institución para mejorar la situación en su comunidad? 
96. ¿Ha participado usted en algún o algunos programas para mejorar su vida en general? 
97. ¿Conoce usted a alguien que haya participado en estos programas? 
98. ¿Qué opinión tiene usted de estos programas que se han implementado en su comunidad? 
99. ¿Hay algo que quiere agregar a la conversación, algo que no se ha comentado hasta el momento?
Impact Evaluation Report - Mexico 
186
IRB# 091334
Estudio de comunidades en MEXICO 2014 
© Vanderbilt University 2011. Derechos reservados. All rights reserved. 
Cuestionario para Líderes 
Introducción: 
Dar a conocer la mecánica de trabajo a los participantes 
Establecer el rapport con los participantes: compenetración con ellos, crear bases para un 
ambiente de mayor apertura, tranquilidad y confianza de los participantes. 
FG de Líderes 
INTRODUCCIÓN (5 mints) 
Bienvenida 
Gracias por haber venido a esta reunión para platicar sobre sus opiniones, experiencias de la ciudad y 
temas relacionados. 
Todas las respuestas y comentarios son importantes. No hay respuestas buenas ni malas. 
Participación de todos pero uno a la vez para escuchar y atender su comentario. 
Grabación de audio CONFIDENCIAL –sólo para fines de este trabajo￾PRIMERA INTERVENCIÓN DE CADA PARTICIPANTE (3 mints) 
Nombre, ocupación y pasatiempos preferidos 
I. Preguntas generales (25 mints) 
1. ¿Cuáles son los problemas más grandes con respecto a los jóvenes de su 
comunidad/mencionar CIUDAD? 
2. ¿Qué tan comunes son los conflictos en su escuela? (entre alumnos; entre alumnos y 
maestros) grupos externos con alumnos? 
3. ¿Cómo se resuelven los conflictos? (en la escuela/en la comunidad/centro de /calle/casa)? 
¿Cuál es el papel de Uds. (padres/maestros/líderes) en la resolución de conflictos? ¿Los 
jóvenes dan el mismo peso a todas las autoridades o creen que alguna tiene mayor 
“influencia” y por qué? 
4. ¿Existen medios de resolver conflictos que sugiere la secretaría de Educación (p.e., 
convivencia escolar)? ¿Cómo funcionan estos medios? 
5. ¿Saben de casos de padres de familia/maestros/empresarios/líderes sociales que aumentaron 
o disminuyeron su participación en a raíz de algún “acontecimiento violento o coyuntural”? 
6. ¿Cómo se comportan los jóvenes en la resolución de conflictos? ¿escuchan, son agresivos, 
receptivos? 
7. En la escuela, por ejemplo: Cuando un alumno sigue comportándose de una manera que no 
se considera aceptable, es posible expulsarlo(la)? Cual es el reglamento de la secretaría de 
Impact Evaluation Report - Mexico 
187
Educación al respecto? De hecho, hay estudiantes que entran en sus escuelas que han sido 
expulsados de otras escuelas? ¿Qué han hecho estos alumnos? ¿Por qué han sido 
expulsados? ¿le dan el mismo trato a ese chico? ¿cómo lo reciben los otros alumnos? De ser 
el caso: ¿cómo perciben los alumnos a los compañeros que han tenido problemas con la ley, 
que han sido detenidos? ¿Cómo reaccionan los padres de familia ante la llegada de un chico 
(a) que ha estado involucrado en delincuencia juvenil? 
8. ¿Su escuela/colonia/iglesia/empresa/organización tiene algún programa de prevención de 
violencia y delincuencia? ¿Cómo es este programa? ¿Hasta qué punto ha tenido éxito este 
programa? ¿Cuáles han sido sus logros y limitantes? 
9. ¿Cuál es la tasa de deserción estudiantil en mencionar CIUDAD? En qué grado tienden a 
abandonar la escuela los alumnos? Cuáles son los motivos más comunes por la deserción? 
¿Qué hacen las ONG’s /empresas/padres para evitar la deserción escolar? 
10. ¿Hay barreras que impidan que un(a) niño(a) se matricule en una escuela? Cuáles son? 
II. Convivencia Escolar (40 mints) 
11. ¿Participa su Organización/Iglesia/comunidad/empresa con escuelas/colegios en algún 
programa de Convivencia Escolar o convivencia comunitaria? 
12. ¿Me puede explicar cómo funciona este programa? 
13. Según este plan, ¿cómo se debe manejar los conflictos entre estudiantes? ¿Cómo se debe 
negociar los conflictos? ¿Quién interviene? 
14. ¿Qué tipos de correcciones y sanciones se usan con los agresores? 
15. ¿Quiénes participan en mantener la seguridad o vigilar por la convivencia escolar y 
comunitaria? 
 a. ¿En qué forma lo hacen? 
 b. ¿Qué tan exitoso han sido sus esfuerzos? (Pídales que den ejemplos.) 
A) Organización.
16. ¿Qué tipo de liderazgo hay en la comunidad?¿Hay comités o clubes de líderes 
estudiantiles/juveniles/laborales/religiosos? 
 17¿Qué hacen estos comités o clubes? 
 18. ¿Cómo trabajan? ¿Quienes los apoyan? 
 19. ¿Qué ha hecho este club para negociar los conflictos entre los jóvenes? (dé ejemplos) 
 20. ¿Han hecho algo estos comités o clubes para prevenir los conflictos de convivencia? ¿Ha 
habido casos que involucraban pandilleros? 
 21. ¿Hay un Comité en su escuela, templo, organización o colonia dedicado a prevenir la 
delincuencia, violencia, drogadicción, etc? 
 22. ¿Qué hace este comité? ¿Cómo funciona? ¿Qué tipo de persona forma parte de él? 
 23.¿Se ha integrado a algún pandillero a esos clubes o comités? De ser el caso ¿Cómo los 
percibe el comité y la comunidad? 
24. ¿Ha oído usted hablar de las Escuelas Seguras? ¿Qué sabe al respecto? ¿Cómo son 
diferentes las Escuelas Seguras de las escuelas ordinarias—cómo funcionan? ¿Hasta qué punto 
han tenido éxito en prevenir la delincuencia? ¿Hay Escuelas Seguras en sus comunidades? 
25.. ¿Conoce usted el programa YYYY? ¿Cómo funciona YYYY? ¿Ha tenido éxito en reducir 
el nivel de delincuencia en las escuelas? ¿Cómo ha logrado esto? (Si no, por qué no?) 
26. ¿Se trabaja con niños y niñas de edad de kinder o primaria, para educarlos en la convivencia 
escolar? 
26a. Pensando específicamente en los últimos dos años, ¿cree usted que algo haya 
cambiado/mejorado en las escuelas o en las calles? 
B). Participación Comunitaria. 
Impact Evaluation Report - Mexico 
188
27. ¿Hay grupos activos en su comunidad/mencionar CIUDAD dedicados a reducir el nivel de 
violencia o delincuencia entre los jóvenes, en las escuelas y colonias? ¿Cuáles son sus nombres? 
 28. Específicamente, qué ha hecho la Escuela, la asociación de Padres y Madres/el templo/su 
empresa para prevenir la violencia entre los jóvenes? ¿Hay relación con otros miembros de la 
comunidad, empresas, ONG’s? ¿Ha habido casos de docentes que se burlen, abusen, violenten a 
los alumnos? ¿Qué hacen los padres? ¿Qué hacen los maestros? ¿y en los casos donde los padres 
violenten a sus hijos? ¿Empresas a los jóvenes empleados en ellas? 
 29. ¿Qué han hecho para prevenir que los jóvenes se metan en pandillas? 
 30. ¿Cuáles son los actos de delincuencia más comunes en su centro de trabajo/colonia? 
 31. ¿Ustedes alguna vez han asistido a la casa de algún(a) estudiante/joven para hablar con su 
padre y madre sobre su conducta (en la escuela, trabajo, colonia)? ¿Por qué fue? ¿Qué resultado 
tuvo esta visita? ¿Cómo se acercan o abordan a la familia de un joven en problemas (cualquier 
tipo) y cómo abordan a la familia de un (a) joven que es maltratado por su 
familia/maestro/empleador? 
32. ¿Qué rol ha desempeñado la policía o autoridades de procuración de justicia en su escuela o 
colonia en torno a la prevención y tratamiento de la delincuencia? ¿Cuáles han sido sus 
actividades? 
 33. ¿Con qué frecuencia viene la policía u otra autoridad a su escuela/colonia? ¿Cómo los 
reciben los chicos? ¿Qué hacen cuando vienen? [Si el/la entrevistado/a menciona que “dan 
charlas”, preguntas sobre qué temas dan charlas.] ¿Acude la misma autoridad o funcionario o los 
rolan? 
 34. ¿Organizan actividades deportivas? ¿Hay más actividades, (como defensa personal, 
medidas en casos de balacera, primeros auxilios, drogas)? 
35. ¿Los sacerdotes católicos o pastores evangélicos vienen con el fin de prevenir la violencia y 
delincuencia? ¿Cuáles son sus actividades? ¿Cómo los reciben los alumnos/jóvenes? ¿se 
muestran interesados? 
36. Hay otras personas o instituciones que acudan a su escuela/comunidad/empresa/templo para 
impartir charlas o actividades los jóvenes? 
III. Los jóvenes y las pandillas (35 mints) 
37. ¿Cómo son los jóvenes de su escuela/comunidad/templo? a) ¿En qué forma se visten y peinan los 
muchachos, las niñas y mujeres? ¿Qué los distingue de los que son pandilleros? 
38. ¿Cómo sabe usted si un joven es miembro de una pandilla? 
39. ¿Qué señas dan los jóvenes que pertenecen a una pandilla? (Sondeo: su apariencia; su 
conducta)? 
40. ¿A qué edad comienzan a afiliarse con las pandillas? 
41. ¿Por qué cree usted que los jóvenes se afilian con las pandillas? ¿Cómo se inician? 
42. ¿Hombres y mujeres se afilian por motivos distintos? ¿Y las niñas y mujeres, por qué se afilian? 
¿Y los hombres? 
a. ¿Qué factores contribuyen a esta afiliación? 
b. ¿Qué estrategias usan los pandilleros para reclutar jóvenes a su pandilla? 
43. ¿Cómo es su forma de hablar? ¿Las niñas y mujeres también hablan así? 
a. ¿Hablan distinto a los otros jóvenes? 
44. ¿Pueden salir de la pandilla? 
45. ¿De qué manera se comportan los miembros de las pandillas? ¿es distinta al comportamiento de 
los otros muchachos? 
46. Hay presencia de pandillas en su colegio/empresa/ ¿Es la escuela/colonia/ un terreno neutral? 
¿Cómo es el control del poder dentro de la escuela/negocio/colonia cuando hay pandillas?¿hay 
disputa geográfica? 
Impact Evaluation Report - Mexico 
189
47. ¿Han tratado las pandillas de interferir en actividades escolares, civiles, laborales, religiosas? 
48. ¿cuál es el trato que dan las pandillas a los docentes/sacerdotes/empleadores/padres? Las 
pandillas los respetan? ¿han sido víctimas de actos delincuenciales por parte de las pandillas? 
(den ejemplos). 
49. ¿Cómo se llaman las pandillas que hay en mencionar CIUDAD? 
50. ¿Cometen delincuencias en su escuela/colegio/barrio? ¿Qué tipo de delincuencias cometen? 
51. ¿Qué hacen cuando se dan cuenta de un acto de delincuencia juvenil? 
52. ¿Qué hacen para prevenir los actos de delincuencia en el Centro Escolar/laboral/templo/barrio? 
53. ¿Hacen los estudiantes/lo jóvenes algo para prevenir o reparar el daño en la propiedad del 
colegio/parroquia/colonia/trabajo? 
54. ¿Escriben graffitis (manchas)? 
a. ¿Entienden los graffitis? 
b. ¿A quién se dirigen los mensajes de graffiti (a los miembros de la pandilla a que 
pertenece el pintor? A los miembros de otra pandilla? A la policía? A otro grupo?) 
IV. Crimen Organizado. (15 mints) 
55. ¿Están las pandillas involucradas con el crimen organizado? 
56. .¿Cómo se da ese acercamiento? 
57. ¿Cómo trabajan? 
58. ¿Hay presencia del crimen organizado en su escuela/negocio/colonia/templo? No ¿Por qué? ¿Qué han 
hecho para prevenirlo? Preguntar si antes lo había. De ser Si ¿cómo es? 
59. Sabe de escuelas/empleos donde se ha infiltrado el crimen organizado ¿de qué modo? 
60. ¿Sabe de alumnos/fieles o colaboradores que sean parte del crimen organizado? 
61. ¿Qué tan grave es el problema del consumo de drogas en la escuela/colonia/empresa/templo? 62¿Y el 
tráfico de estas substancias? 
V. Cierre (5 mints) 
63. ¿Ustedes coordinan sus esfuerzos con otros miembros de la comunidad para prevenir que los y las 
jóvenes se afilien con las pandillas? 
64. ¿Con qué instituciones trabajan los miembros de su comunidad? ¿En qué forma trabajan? ¿Qué 
hacen? ¿Qué planes y proyectos tienen? 
65. ¿Cuál es su impresión sobre el trabajo de las autoridades en torno a prevenir la delincuencia y el 
crimen organizado? ¿Qué se podría hacer para prevenir que los y las jóvenes se metan en pandillas? 
66. ¿Hay algo más que quieran ustedes comentar y que no hayamos mencionado a lo largo de esta 
reunión? ¿Algo que consideren ustedes importante? 
Muchas gracias por su participación. 
Impact Evaluation Report - Mexico 
190
IRB# 091334
Estudio de comunidades en MEXICO 2014 
© Vanderbilt University 2011. Derechos reservados. All rights reserved. 
Cuestionario para Líderes y Funcionarios 
Focus Group, Mexico
Introducción General: 
Dar a conocer la mecánica de trabajo a los participantes 
Establecer el rapport con los participantes: compenetración con ellos, crear bases para un 
ambiente de mayor apertura, tranquilidad y confianza de los participantes. 
INTRODUCCIÓN (5 min) 
Bienvenida 
Gracias por haber venido a esta reunión para platicar sobre sus opiniones, experiencias de la ciudad y 
temas relacionados. 
Todas las respuestas y comentarios son importantes. No hay respuestas buenas ni malas. 
Participación de todos pero uno a la vez para escuchar y atender su comentario. 
Grabación de audio CONFIDENCIAL –sólo para fines de este trabajo￾PRIMERA INTERVENCIÓN DE CADA PARTICIPANTE (3 min) 
Nombre, ocupación y pasatiempos preferidos 
A. Actividades de prevención social de delincuencia (30 minutos) 
1. ¿Qué rol desempeña usted en su institución? 
2. ¿Qué actividades realizan en su institución para prevenir la delincuencia? 
3. ¿Usted o sus colegas hacen algo para facilitar que otros servicios sociales o comunitarios se 
provean a las comunidades? (Dar ejemplos) 
4. ¿Tienen programas contacto o actividades con escuelas, colegios o asociaciones de padres de 
familia? ¿Cuáles son las actividades que realiza en las escuelas y en los colegios? ¿Con qué 
frecuencia asisten? 
5. ¿Tienen reuniones con directores de escuelas y colegios o con sus docentes? ¿De qué platican 
y cuál es su participación? Y ¿con los padres de familia y alumnos? ¿De qué se habla en estas 
reuniones? 
6. ¿Ha oído hablar de algún programa de la secretaría de Educación que fomente la convivencia 
escolar, o sea un programa donde hay mediación de conflictos? ¿Qué sabe de este programa? 
7. ¿Ha oído si ha tenido éxito o no en disminuir o prevenir la violencia entre los alumnos? 
¿Qué sabe usted de la Escuelas Seguras? ¿Cómo funciona este programa? ¿Quién se encarga del 
programa? ¿Cuántas Escuelas Seguras hay, y dónde se están ubicadas? ¿Qué tan exitosas han 
sido hasta el momento, en reducir el nivel de delincuencia y violencia? (¿Cómo sabe esto—a 
base de qué datos/en qué lo nota/quién le dijo?) 
5. Cree que parte de sus tareas es ¿mediar/Conciliar conflictos en su comunidad? 
Impact Evaluation Report - Mexico 
191
¿Qué tipo de asuntos trata usted en su rol de mediador? [Pedir ejemplo ¿qué ha hecho Ud. 
durante un caso de mediación?] 
6. ¿Hay Centros de Mediación en los barrios/zonas donde usted trabaja? ¿Quiénes son las 
personas que llevan a cabo la mediación? ¿Qué tipos de casos se median en esos centros? 
¿Tratan casos de violencia? 
7. Cuando hay casos de violencia doméstica ¿Quién suele solicitar su ayuda? [Sondear: Una 
mujer a quien se está maltratando; un hombre del hogar, un vecino?] 
 ¿Qué hace usted en los casos de violencia doméstica? [¿Me puede dar un ejemplo?] ¿Han 
recibido capacitación al respecto? 
8. Según su opinión, ¿cuáles han sido los logros y los fracasos del municipio hasta el momento 
en prevenir la violencia en mencionar CIUDAD? 
9 ¿Ha tenido usted contacto con algunas organizaciones de la sociedad civil (ONG o asociaciones 
religiosas) que se enfocan en la prevención de la violencia y la delincuencia juvenil? ¿Cuáles 
ONGs? ¿Cuáles iglesias y templos? ¿Tienen programas en común? ¿De qué trata? ¿Cuáles son 
sus resultados? 
10 ¿Qué rol desempeña usted con respecto a estas organizaciones? 
11¿Tiene usted relación con algún programa nacional de Seguridad Ciudadana? 
12 ¿Qué tipo de actividades se llevan a cabo de manera conjunta entre su institución y el 
programa de Seguridad Ciudadana? 
13. ¿Utiliza usted los llamados “mapas de protección” o mapas de recursos socio-comunitarios? 
14. Cuando hay modificaciones legales que repercuten en sus actividades y atribuciones ¿qué tan 
rápido conoce usted estas modificaciones? ¿Hay alguna nueva medida que estén implementando 
para atrapar/consignar/defender (según sea el caso) delincuentes? 
B. Actividades de prevención y combate a la delincuencia juvenil (20 minutos) 
15. ¿Sabe usted en qué formas trabaja su institución junto con las comunidades para prevenir la 
delincuencia juvenil? ¿Cuáles son esas formas? ¿Sabe si en su institución se desarrollan acciones 
para la operación de un Observatorio Ciudadano o Consejo ciudadano? ¿Su institución tiene una 
política dirigida a los jóvenes? 
16. ¿Qué acciones preventivas cree que serían las más eficaces para resolver el problema de la 
delincuencia juvenil? 
17. ¿Qué acciones de control cree que se podrían tomar para tratar el problema de la 
delincuencia juvenil? 
18. ¿Ha elaborado usted (o su institución) instrumentos auxiliares como “mapas de crimen” en 
las comunidades donde le toca trabajar? 
19. ¿Existe en mencionar CIUDAD un programa de Convivencia Pacífica? ¿Qué sabe de este 
programa? 
 ¿Participa usted en este programa? ¿En qué forma? [Dejarle explicar, con ejemplos, detalles] 
 ¿Hasta qué punto cree usted que ha tenido éxito el Programa de Convivencia Pacífica? ¿Por qué 
opina que sí [o no] ha tenido éxito? ¿Qué evidencia hay del éxito [o falta de éxito] de él? 
20. ¿Hay un programa de prevención que involucra la ayuda de los jóvenes en colaborar con la 
policía/Centros de rehabilitación/Tutelares (según sea el caso) para prevenir delitos? En algunos 
países se llama “Jóvenes Contra el Delito”, algo parecido al Vecino Vigilante, pero se incorpora 
la cooperación de los jóvenes. Hay algo de este tipo, parecido, en mencionar CIUDAD? 
 ¿Qué sabe de este program? ¿Cómo funciona? ¿Qué ha logrado hasta el momento, según 
su entender? 
21. ¿Existe en mencionar CIUDAD el programa “YYYY”? ¿Qué sabe al respecto? 
22. ¿Ha oído de un programa enfocado en jóvenes el que se llama ZZZZ? ¿Qué sabe de este 
programa? ¿Hasta qué punto ha tenido éxito? 
22a. ¿Ha notado usted algún cambio (mejoría) a partir de la instalación de alguno de estos 
programas? ¿Cree usted que han funcionado en este último par de años? 
Impact Evaluation Report - Mexico 
192
C. Actividades de toma de denuncia (10 minutos) 
23. Hablando ahora de la toma de denuncia y atención ciudadana, ¿existen en mencionar 
CIUDAD Centros de Denuncia? ¿A dónde acude la comunidad a denunciar? (Centros de 
Denuncia/Ministerio Público) ¿Cómo es el trato que se les da? ¿Qué delitos son los que 
denuncian? 
 *Si las comunidades no tienen acceso a tales oficinas de denuncia, ¿a dónde van las 
personas para hacer las denuncias? 
21. ¿Por qué será que hay personas que no denuncian los delitos? ¿Qué obstáculos enfrentan? 
 ¿Por qué será que hay personas que ni llaman a la policía para informar de un delito? 
22. Cuando alguien de alguna comunidad viene a la unidad policial para hacer una denuncia, 
¿qué tipo de seguimiento y supervisión da su institución a la denuncia? 
22b. Hablando ahora de las víctimas del delito, ¿Qué tipo de atención se da a las víctimas de 
delitos? 
23. ¿Existen otros mecanismos o lugares a dónde puede acudir una persona para hacer una 
denuncia? SI ¿Cómo funciona, en qué consiste? 
D. Características de las comunidades de intervención (10 mints) 
25. ¿Usted se da cuenta de los problemas que las comunidades consideran prioritarios? 
26. En el caso de mencionar CIUDAD, ¿cuáles son los problemas prioritarios que identifican? 
27. ¿Qué tipos de problemas sociales en general enfrentan las colonias de mencionar CIUDAD? 
28. ¿Cuáles son los problemas de violencia y delincuencia que existen en estas comunidades? 
 ¿Cuáles son los delitos más comunes en estas comunidades? 
28. ¿Cuáles son los factores de riesgo social y situacional que amenazan a la comunidad? 
29. ¿Qué tan violentas son las comunidades de mencionar CIUDAD? 
30. ¿Quiénes son los que cometen estos actos de violencia? 
31. ¿Cuáles son las comunidades menos peligrosas en que usted trabaja? ¿Y las más peligrosas? 
32. ¿Qué porcentaje de los delincuentes que viven en las comunidades son personas que fueron 
deportadas/transitan a los Estados Unidos? ¿Cómo impactan estas personas la convivencia en el 
barrio? 
33. ¿Qué porcentaje de ellos han salido directamente de cárceles? (la menor parte, la mayoría, la 
mitad). ¿Cómo impactan estas personas la convivencia en el barrio? 
34.¿Qué y quienes causan mayores problema a los habitantes problemas en mencionar CIUDAD? 
E. Acciones pandillas (30 mints) 
34. ¿Hay pandillas en las comunidades en que usted trabaja? ¿Cuáles son? 
35. ¿Qué tipos de delitos cometen los pandilleros en mencionar CIUDAD? 
 Los pandilleros que viven en las comunidades donde usted trabaja, ¿hay algunos que han sido 
deportados o encarcelados en los Estados Unidos? ¿Qué delitos habían cometido en los Estados 
Unidos? 
 ¿Qué estatus tienen estas personas entre sus compañeros pandilleros? ¿Tienen un rol 
especial en las pandillas? 
36. ¿Cómo sabe usted si un joven es pandillero al verlo en la calle, qué características lo 
identifican como pandillero? [Sondear: Ropa, tatuajes, corte de pelo] 
 ¿Los pandilleros tienen una forma de hablar especial de ellos? ¿Cómo es? 
 ¿Usan una forma de comunicación no verbal -con las manos, gestos- especial? ¿Me las 
puede describir? 
37. Y si es niña o mujer, ¿cuáles son las características que señalan que ella es miembro de 
alguna pandilla? ¿Qué trato le dan las pandillas a las mujeres? ¿la policía trata igual a hombres 
que a mujeres de una pandilla? 
Impact Evaluation Report - Mexico 
193
38. ¿Por qué, cree usted, se meten los jóvenes en las pandillas? ¿A qué edad lo hacen? 
39. Y las niñas y mujeres, ¿por qué se meten en ellas? 
40. ¿Hasta qué edad se quedan en las pandillas? 
41. ¿Hay manera de salir de una pandilla? ¿Cuál es la manera? ¿la policía tiene alguna programa 
para sacar a mujeres u hombres de las pandillas? 
42. Cuando usted ve grafitis (manchas), ¿los entiende? ¿Sabe qué significado tienen y a quién se 
dirigen estos mensajes? [Sondear: A otra pandilla? A los miembros de la misma pandilla? A la 
Policía?] 
43. Hablando ahora de las bandas del crimen organizado, ¿Qué hacen? 
44. ¿En qué forma son distintas a las pandillas? ¿En qué forma son parecidas? O sea, ¿qué tienen 
en común las bandas del crimen organizado con las pandillas? ¿hay otros grupos delincuenciales 
que no se identifican como pandillas ni como grupos de delincuencia organizada ? 
45. Para los pandilleros que salen de las cárceles después de haber cumplido con su sentencia, 
¿hay programas de rehabilitación? ¿Hay suficientes para todos los que salen de los cárceles? 
¿Qué tipos de rehabilitación se ofrecen? 
46. Para los pandilleros menores, ¿hay suficientes Centros de rehabilitación? ¿Qué tipos 
actividades ofrecen estos Centros? 
47. ¿Cómo es la relación entre las pandillas y el crimen organizado? 
48. ¿Ha presenciado altercados o enfrentamientos entre estos? 
49. ¿Cómo se siente ante la presencia del crimen organizado en mencionar CIUDAD? 
49a. ¿Cree usted que en los últimos dos años ha habido algún cambio a partir de los nuevos 
programas puestos a funcionar con respecto a los jóvenes y el crimen organizado? 
F. Actividades de coordinación y participación comunitaria (10 mints) 
50. ¿Realiza usted actividades en conjunto y coordinadamente con las comunidades? [Sondear--
¿Cuáles por ejemplo?] 
51. ¿Planifica usted, juntamente con los líderes comunitarios, servicios sociales o acciones 
preventivas? 
52. ¿Hay algunas responsabilidades que su institución comparte con los líderes comunitarios en 
las comunidades en que usted trabaja? 
53. ¿Qué tipos de líderes son? (líderes de centros comunitarios, sacerdotes/pastores, maestros) 
54. ¿Ha establecido usted contactos con miembros de las comunidades en que usted trabaja, que 
sirvan de informantes sobre hechos de delincuencia y violencia? ¿Con qué frecuencia conversa 
usted con estas personas? 
G. Cierre (10 mints) 
56. ¿Qué opina usted del rol del Ejército en prevenir la delincuencia? ¿De hecho el Ejército 
desempeña el rol de prevenir delincuencia en mencionar CIUDAD? ¿Qué hace exactamente? 
57. ¿Ha habido casos en mencionar CIUDAD de linchamientos (o sea, hacer justicia con la 
propia mano) los últimos 12 meses? ¿Por qué motivo hubo estos linchamientos, cree usted? 
58. ¿Cree que hay medidas legales que entorpecen su labor? 
59. ¿Qué medidas está tomando la ciudad para prevenir el delito. 
60. ¿Hay algo más que quisiera usted agregar a todo esto que hemos platicado durante esta 
reunión? ¿Algo que no hayamos mencionado y que considere usted importante? 
Muchas gracias por su participación.