Effects of Internal Forced Displacement on Crime:
Evidence from Colombia
Andrés Sánchez-Saldarriaga, Catalina Gómez-Toro, Hermilson
Velásquez and Juan Felipe Mejía-Mejía
Lecturas de Economía - No. 101. Medellín, enero-junio 2024
Evidence from Colombia
Andrés Sánchez-Saldarriaga, Catalina Gómez-Toro, Hermilson
Velásquez and Juan Felipe Mejía-Mejía
Lecturas de Economía - No. 101. Medellín, enero-junio 2024
Lecturas de Economía, 101 (enero-junio 2024), pp. 135-172Página 1 de 1CROSSMARK_logo_3_Test
25/06/2020file:///D:/Paty/LecturasEconomia/LEvol93_2020/LEVol93Abr26_2020/ART_340990/CR...
Andrés Sánchez-Saldarriaga, Catalina Gómez-Toro, Hermilson Velásquez
and Juan Felipe Mejía-Mejía
Effects of Internal Forced Displacement on Crime: Evidence from Colombia
Abstract: Internal forced displacement, a consequence of territorial control strategies by armed groups,
has resulted in profound socioeconomic challenges for receiving areas. Approximately eight million people
have been affected in Colombia, where internal conflict and criminal networks have prevailed for over
five decades. This paper explores the relationship between internal forced migration and crime rates
in Colombian municipalities from 2003 to 2016, focusing on spatial dynamics. The study utilizes
a spatial panel model and considers five crime categories: homicides, kidnapping, personal injuries,
automobile theft, and residential burglary. The results reveal significant associations between forced
displacement and certain types of crime, highlighting the importance of spatial analysis in understanding
the interaction between migration and crime. The findings offer valuable insights for policymakers
and researchers, facilitating informed policy interventions and community initiatives to address the
repercussions of internal migration on crime dynamics.
Keywords: Internal forced displacement, crime, spatial panel, Colombia, Latin America, internal
migration.
JEL Classification: K42, O15, R23, R59.
Efectos del desplazamiento forzado interno sobre la delincuencia: evidencia de
Colombia
Resumen: El desplazamiento forzado interno, consecuencia de las estrategias de control territorial
de los grupos armados, ha resultado en profundos desafíos socioeconómicos para las zonas receptoras.
Aproximadamente ocho millones de personas se han visto afectadas en Colombia, donde el conflicto
interno y las redes criminales han prevalecido durante más de cinco décadas. Este artículo explora la
relación entre la migración interna forzada y las tasas de criminalidad en los municipios colombianos
de 2003 a 2016, enfocándose en la dinámica espacial. El estudio utiliza un modelo de panel espacial
y considera cinco categorías de delitos: homicidios, secuestros, lesiones personales, robo de automóviles
y robo residencial. Los resultados revelan asociaciones significativas entre el desplazamiento forzado y
ciertos tipos de delincuencia, destacando la importancia del análisis espacial para comprender la interac-
ción entre migración y delincuencia. Los hallazgos ofrecen información valiosa para los formuladores de
políticas y los investigadores, facilitando intervenciones políticas informadas e iniciativas comunitarias
para abordar las repercusiones de la migración interna en la dinámica del crimen.
Palabras clave: desplazamiento interno forzado, criminalidad, panel espacial, Colombia, América
Latina, migración interna.
https://doi.org/10.17533/udea.le.n101a353012
Este artículo y sus anexos se distribuyen por la revista Lecturas de Economía bajo los términos de la Licencia Creative
Commons Atribución-NoComercial-CompartirIgual 4.0. https://creativecommons.org/licenses/by-nc-sa/4.0/
25/06/2020file:///D:/Paty/LecturasEconomia/LEvol93_2020/LEVol93Abr26_2020/ART_340990/CR...
Andrés Sánchez-Saldarriaga, Catalina Gómez-Toro, Hermilson Velásquez
and Juan Felipe Mejía-Mejía
Effects of Internal Forced Displacement on Crime: Evidence from Colombia
Abstract: Internal forced displacement, a consequence of territorial control strategies by armed groups,
has resulted in profound socioeconomic challenges for receiving areas. Approximately eight million people
have been affected in Colombia, where internal conflict and criminal networks have prevailed for over
five decades. This paper explores the relationship between internal forced migration and crime rates
in Colombian municipalities from 2003 to 2016, focusing on spatial dynamics. The study utilizes
a spatial panel model and considers five crime categories: homicides, kidnapping, personal injuries,
automobile theft, and residential burglary. The results reveal significant associations between forced
displacement and certain types of crime, highlighting the importance of spatial analysis in understanding
the interaction between migration and crime. The findings offer valuable insights for policymakers
and researchers, facilitating informed policy interventions and community initiatives to address the
repercussions of internal migration on crime dynamics.
Keywords: Internal forced displacement, crime, spatial panel, Colombia, Latin America, internal
migration.
JEL Classification: K42, O15, R23, R59.
Efectos del desplazamiento forzado interno sobre la delincuencia: evidencia de
Colombia
Resumen: El desplazamiento forzado interno, consecuencia de las estrategias de control territorial
de los grupos armados, ha resultado en profundos desafíos socioeconómicos para las zonas receptoras.
Aproximadamente ocho millones de personas se han visto afectadas en Colombia, donde el conflicto
interno y las redes criminales han prevalecido durante más de cinco décadas. Este artículo explora la
relación entre la migración interna forzada y las tasas de criminalidad en los municipios colombianos
de 2003 a 2016, enfocándose en la dinámica espacial. El estudio utiliza un modelo de panel espacial
y considera cinco categorías de delitos: homicidios, secuestros, lesiones personales, robo de automóviles
y robo residencial. Los resultados revelan asociaciones significativas entre el desplazamiento forzado y
ciertos tipos de delincuencia, destacando la importancia del análisis espacial para comprender la interac-
ción entre migración y delincuencia. Los hallazgos ofrecen información valiosa para los formuladores de
políticas y los investigadores, facilitando intervenciones políticas informadas e iniciativas comunitarias
para abordar las repercusiones de la migración interna en la dinámica del crimen.
Palabras clave: desplazamiento interno forzado, criminalidad, panel espacial, Colombia, América
Latina, migración interna.
https://doi.org/10.17533/udea.le.n101a353012
Este artículo y sus anexos se distribuyen por la revista Lecturas de Economía bajo los términos de la Licencia Creative
Commons Atribución-NoComercial-CompartirIgual 4.0. https://creativecommons.org/licenses/by-nc-sa/4.0/
Effets du déplacement forcé interne sur la criminalité: données probantes
provenant de Colombie
Résumé: Les déplacements forcés internes, conséquence des stratégies de contrôle territorial des groupes
armés, ont entraîné de profonds défis socio-économiques pour les zones d’accueil. Environ huit millions
de personnes ont été touchées en Colombie, où règnent des conflits internes et des réseaux criminels
depuis plus de cinq décennies. Cet article explore la relation entre la migration forcée interne et les
taux de criminalité dans les municipalités colombiennes de 2003 à 2016, en se concentrant sur la
dynamique spatiale. L’étude utilise un modèle de panel spatial et considère cinq catégories de crimes : les
homicides, les enlèvements, les blessures corporelles, le vol d’automobile et le cambriolage résidentiel. Les
résultats révèlent des associations significatives entre les déplacements forcés et certains types de criminalité,
soulignant l’importance de l’analyse spatiale pour comprendre l’interaction entre migration et criminalité.
Les résultats offrent des informations précieuses aux décideurs politiques et aux chercheurs, facilitant
des interventions politiques éclairées et des initiatives communautaires pour faire face aux répercussions
de la migration interne sur la dynamique de la criminalité.
Mots clés: Déplacement forcé interne, criminalité, panel spatial, Colombie, Amérique latine,
migration interne.
Cómo citar / How to cite this item:
Sánchez-Saldarriaga, A., Gómez-Toro, C., Velásquez, H., & Mejía-Mejía, F. (2023).
Effects of Internal Forced Displacement on Crime: Evidence from Colombia. Lecturas
de Economía, 101, 135-172. https://doi.org/10.17533/udea.le.n101a353012
provenant de Colombie
Résumé: Les déplacements forcés internes, conséquence des stratégies de contrôle territorial des groupes
armés, ont entraîné de profonds défis socio-économiques pour les zones d’accueil. Environ huit millions
de personnes ont été touchées en Colombie, où règnent des conflits internes et des réseaux criminels
depuis plus de cinq décennies. Cet article explore la relation entre la migration forcée interne et les
taux de criminalité dans les municipalités colombiennes de 2003 à 2016, en se concentrant sur la
dynamique spatiale. L’étude utilise un modèle de panel spatial et considère cinq catégories de crimes : les
homicides, les enlèvements, les blessures corporelles, le vol d’automobile et le cambriolage résidentiel. Les
résultats révèlent des associations significatives entre les déplacements forcés et certains types de criminalité,
soulignant l’importance de l’analyse spatiale pour comprendre l’interaction entre migration et criminalité.
Les résultats offrent des informations précieuses aux décideurs politiques et aux chercheurs, facilitant
des interventions politiques éclairées et des initiatives communautaires pour faire face aux répercussions
de la migration interne sur la dynamique de la criminalité.
Mots clés: Déplacement forcé interne, criminalité, panel spatial, Colombie, Amérique latine,
migration interne.
Cómo citar / How to cite this item:
Sánchez-Saldarriaga, A., Gómez-Toro, C., Velásquez, H., & Mejía-Mejía, F. (2023).
Effects of Internal Forced Displacement on Crime: Evidence from Colombia. Lecturas
de Economía, 101, 135-172. https://doi.org/10.17533/udea.le.n101a353012
Effects of Internal Forced Displacement on Crime: Evidence
from Colombia
Andrés Sánchez-Saldarriaga a, Catalina Gómez-Toro b, Hermilson
Velásquez c and Juan Felipe Mejía-Mejía d
Introduction. –I. Literature Review. –II. Empirical Strategy. –III. Estimations Results.
–Conclusions. –Appendix. –Ethics Statement. –References.
Original manuscript received on 11 March 2023; final version accepted on 09 January 2024
Introduction
Forced displacement, characterized as a rapid and unexpected form of
migration, has emerged as a profound consequence of territorial control
strategies and attacks on civilian populations by armed illegal and criminal
groups (Roche-Villarreal, 2012). Colombia has been beset by internal
conflict and criminal networks affecting a staggering eight million people
over five decades (InSight Crime, 2021; Unidad para las Víctimas, n.d.). The
appropriation of essential resources by insurgents and criminals in the pursuit
of territorial supremacy has led to the deprivation of welfare and property
rights for the affected civilian population, exposing them to increased risks
of permanent injuries and fatalities (Ibáñez, 2009; Téllez, 2021).
a Andrés Sánchez-Saldarriaga: Research assistant at Universidad EAFIT, Economics and Finance
School, Medellín, Colombia. E-mail: afsanchezs@eafit.edu.co
https://orcid.org/0000-0001-6558-2089
b Catalina Gómez-Toro: Professor and researcher at Universidad EAFIT, Economics Depart-
ment, Medellín, Colombia. E-mail: cgomezt1@eafit.edu.co
https://orcid.org/0000-0003-1851-6753
c Hermilson Velásquez: Professor and researcher at Universidad EAFIT, Economics Depart-
ment, Medellín, Colombia. E-mail: evelas@eafit.edu.co
https://orcid.org/0000-0002-2588-1356
d Juan Felipe Mejía-Mejía: Professor and researcher at Universidad EAFIT, Economics
Department, Medellín, Colombia. E-mail: jfmejia@eafit.edu.co
https://orcid.org/0000-0001-5022-2646
from Colombia
Andrés Sánchez-Saldarriaga a, Catalina Gómez-Toro b, Hermilson
Velásquez c and Juan Felipe Mejía-Mejía d
Introduction. –I. Literature Review. –II. Empirical Strategy. –III. Estimations Results.
–Conclusions. –Appendix. –Ethics Statement. –References.
Original manuscript received on 11 March 2023; final version accepted on 09 January 2024
Introduction
Forced displacement, characterized as a rapid and unexpected form of
migration, has emerged as a profound consequence of territorial control
strategies and attacks on civilian populations by armed illegal and criminal
groups (Roche-Villarreal, 2012). Colombia has been beset by internal
conflict and criminal networks affecting a staggering eight million people
over five decades (InSight Crime, 2021; Unidad para las Víctimas, n.d.). The
appropriation of essential resources by insurgents and criminals in the pursuit
of territorial supremacy has led to the deprivation of welfare and property
rights for the affected civilian population, exposing them to increased risks
of permanent injuries and fatalities (Ibáñez, 2009; Téllez, 2021).
a Andrés Sánchez-Saldarriaga: Research assistant at Universidad EAFIT, Economics and Finance
School, Medellín, Colombia. E-mail: afsanchezs@eafit.edu.co
https://orcid.org/0000-0001-6558-2089
b Catalina Gómez-Toro: Professor and researcher at Universidad EAFIT, Economics Depart-
ment, Medellín, Colombia. E-mail: cgomezt1@eafit.edu.co
https://orcid.org/0000-0003-1851-6753
c Hermilson Velásquez: Professor and researcher at Universidad EAFIT, Economics Depart-
ment, Medellín, Colombia. E-mail: evelas@eafit.edu.co
https://orcid.org/0000-0002-2588-1356
d Juan Felipe Mejía-Mejía: Professor and researcher at Universidad EAFIT, Economics
Department, Medellín, Colombia. E-mail: jfmejia@eafit.edu.co
https://orcid.org/0000-0001-5022-2646
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
Host areas of forced migrants face socioeconomic challenges, including
worsening labor market conditions, increased inequality, and deteriorating
living standards, with limited access to public services and educational systems
(Calderón-Mejía & Ibáñez, 2016; Ibáñez & Moya, 2010). The arrival of
forced migrants also alters the dynamics of the local population and creates
a more heterogeneous society, resembling the consequences observed in
general migration scenarios (Ibáñez & Vélez, 2008). Furthermore, despite
considerable government efforts to combat criminal activities, crime rates
in the receiving areas have persisted and, in some cases, even increased,
negatively affecting the population’s welfare (Roche-Villarreal, 2012).
Mixed findings characterize empirical research on the migration-crime
relationship. Reid et al. (2005) examine the effects of migration on crime rates
in metropolitan areas and conclude that migration does not necessarily lead
to increased crime rates. Similarly, Bianchi et al. (2012) find that migration
primarily affects the incidence of robberies but does not impact other crime
types. Kubrin et al. (2016) identify diverse effects of migration on violent
crime rates across various neighborhoods. International migration studies
exploring spatial heterogeneity effects in specific regions, such as those by
Graif and Sampson (2009) and Arnio and Baumer (2012), have contributed
to the literature by revealing inconsistent relationships between migrants and
crime. These studies underscore the significance of incorporating spatial
dynamics into the analysis of migration and crime interactions.
The main objective of this research paper is to delve into the relationship
between internal forced migration and its potential impact on crime rates. To
achieve this aim, we examine crime rates in various Colombian municipalities
from 2003 to 2016. A crucial spatial component is integrated, recognizing the
growing importance of spatial analysis in elucidating this relationship. Our
work contributes to the existing body of knowledge by providing valuable
insights for policymakers, researchers, and communities into the implications
of internal migration on crime dynamics, thereby facilitating informed policy
interventions and community initiatives.
This study’s novelty lies in its comprehensive investigation of the
relationship between internal forced migration and crime rates in Colombia.
139
Host areas of forced migrants face socioeconomic challenges, including
worsening labor market conditions, increased inequality, and deteriorating
living standards, with limited access to public services and educational systems
(Calderón-Mejía & Ibáñez, 2016; Ibáñez & Moya, 2010). The arrival of
forced migrants also alters the dynamics of the local population and creates
a more heterogeneous society, resembling the consequences observed in
general migration scenarios (Ibáñez & Vélez, 2008). Furthermore, despite
considerable government efforts to combat criminal activities, crime rates
in the receiving areas have persisted and, in some cases, even increased,
negatively affecting the population’s welfare (Roche-Villarreal, 2012).
Mixed findings characterize empirical research on the migration-crime
relationship. Reid et al. (2005) examine the effects of migration on crime rates
in metropolitan areas and conclude that migration does not necessarily lead
to increased crime rates. Similarly, Bianchi et al. (2012) find that migration
primarily affects the incidence of robberies but does not impact other crime
types. Kubrin et al. (2016) identify diverse effects of migration on violent
crime rates across various neighborhoods. International migration studies
exploring spatial heterogeneity effects in specific regions, such as those by
Graif and Sampson (2009) and Arnio and Baumer (2012), have contributed
to the literature by revealing inconsistent relationships between migrants and
crime. These studies underscore the significance of incorporating spatial
dynamics into the analysis of migration and crime interactions.
The main objective of this research paper is to delve into the relationship
between internal forced migration and its potential impact on crime rates. To
achieve this aim, we examine crime rates in various Colombian municipalities
from 2003 to 2016. A crucial spatial component is integrated, recognizing the
growing importance of spatial analysis in elucidating this relationship. Our
work contributes to the existing body of knowledge by providing valuable
insights for policymakers, researchers, and communities into the implications
of internal migration on crime dynamics, thereby facilitating informed policy
interventions and community initiatives.
This study’s novelty lies in its comprehensive investigation of the
relationship between internal forced migration and crime rates in Colombia.
139
140
By integrating spatial analysis, the study offers fresh insights into the
complex interplay between migration patterns and criminal activities across
various geographical locations. This innovative approach adds depth to the
existing literature and contributes to a nuanced understanding of the factors
influencing crime rates in the context of internal forced migration.
This article is organized as follows. In Section I, we conduct a literature
review of the most relevant theories and empirical studies exploring the
relationship between migration and crime. Section II presents the data and
the empirical strategy for spatial panel data, and section III presents the results.
Finally, we present the conclusions.
I. Literature Review
Theoretical approaches to explaining the crime-migration relationship are
founded in sociological and criminological theory, of which the three main
theories are the Social Disorganization Theory, Cultural Theory, and Strain
Theory. The Social Disorganization Theory posits that lack of resources,
inhabitants’ diversity, and residential mobility are determinants of criminal
acts. Due to these factors, communities cannot build strong relations, which
increases social disorganization and ultimately leads to higher crime rates
(Mears, 2001). Cultural Theory proposes that the causes of crime are related
to cultural contradictions since immigrants face different behavior patterns
in host areas; these contradictions exacerbate tensions among the population.
Strain Theory argues that crime is derived from tensions generated by the
pressures to succeed and structural barriers in the receiving areas. Therefore,
tensions arise between immigrants and the local population since each group
strives to improve living conditions (Thomas, 2011). These theories share a
particular characteristic: population heterogeneity, which entails differences
between migrants and local inhabitants and is one of the main determinants
of crime.
Chiswick and Miller (2014) state that if crime can be observed and
analyzed as an economic factor, it is clear how migration comes into play.
One of the first determinants of internal migration studied is related to
the difference in returns between rural and industrial activities. When
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
By integrating spatial analysis, the study offers fresh insights into the
complex interplay between migration patterns and criminal activities across
various geographical locations. This innovative approach adds depth to the
existing literature and contributes to a nuanced understanding of the factors
influencing crime rates in the context of internal forced migration.
This article is organized as follows. In Section I, we conduct a literature
review of the most relevant theories and empirical studies exploring the
relationship between migration and crime. Section II presents the data and
the empirical strategy for spatial panel data, and section III presents the results.
Finally, we present the conclusions.
I. Literature Review
Theoretical approaches to explaining the crime-migration relationship are
founded in sociological and criminological theory, of which the three main
theories are the Social Disorganization Theory, Cultural Theory, and Strain
Theory. The Social Disorganization Theory posits that lack of resources,
inhabitants’ diversity, and residential mobility are determinants of criminal
acts. Due to these factors, communities cannot build strong relations, which
increases social disorganization and ultimately leads to higher crime rates
(Mears, 2001). Cultural Theory proposes that the causes of crime are related
to cultural contradictions since immigrants face different behavior patterns
in host areas; these contradictions exacerbate tensions among the population.
Strain Theory argues that crime is derived from tensions generated by the
pressures to succeed and structural barriers in the receiving areas. Therefore,
tensions arise between immigrants and the local population since each group
strives to improve living conditions (Thomas, 2011). These theories share a
particular characteristic: population heterogeneity, which entails differences
between migrants and local inhabitants and is one of the main determinants
of crime.
Chiswick and Miller (2014) state that if crime can be observed and
analyzed as an economic factor, it is clear how migration comes into play.
One of the first determinants of internal migration studied is related to
the difference in returns between rural and industrial activities. When
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
rural fertility exceeds urban fertility, agricultural labor will grow faster than
industrial employment (Herrick, 1965). In principle, this was perceived as
beneficial due to the scarce labor supply in large industrial centers. However,
migration is increasingly acknowledged as the major contributing factor
to the phenomenon of urban surplus labor and as a force that continues
to exacerbate serious urban unemployment problems caused by growing
economic and structural imbalances between urban and rural areas (Herrick,
1965).1 As for crime and its economic effects, its increase is negatively
related to the welfare of citizens –as evidenced by the reduction in security
perception–, investment decisions, and the reduction of public investment in
sectors that yield higher welfare benefits since resources are diverted to crime
prevention investments.
In some specific cases, migration may be related to crime. When low-
skilled migrants arrive in areas where job opportunities are scarce and their
contact networks are typically reduced, their chances of finding employment
are diminished. At the same time, the possibility of engaging in illegal
activities is heightened. An important factor that increases participation in
illegal activities is that these activities generate high income with little effort.2
The government’s effort to punish crime is a crucial factor in its reduction, as
stated by Ehrlich (1973).
Despite the general belief that there is a link between migration and
crime, empirical research exploring this relationship is sparse (Reid et al.,
2005). Reid et al. (2005) examine the effects of migration on metropolitan
crime rates. After controlling for a host of demographic and economic
characteristics, they find that migration does not increase crime rates in
the receiving areas and that some aspects of migration may even reduce
crime in metropolitan areas. A similar approach was employed by Bianchi
et al. (2012), who found that migration leads to an increase in the incidence
of robberies but does not impact other types of crime. A critical aspect
1 In this perspective, for the Colombian case, Calderon Mejía and Ibáñez (2016), find that
internal migration substantially reduces wages for unskilled urban workers who compete for
jobs with forced migrants.
2 Psychological and social factors are also important determinants of criminal behavior (see
Hirschi and Stark 1969).
141
rural fertility exceeds urban fertility, agricultural labor will grow faster than
industrial employment (Herrick, 1965). In principle, this was perceived as
beneficial due to the scarce labor supply in large industrial centers. However,
migration is increasingly acknowledged as the major contributing factor
to the phenomenon of urban surplus labor and as a force that continues
to exacerbate serious urban unemployment problems caused by growing
economic and structural imbalances between urban and rural areas (Herrick,
1965).1 As for crime and its economic effects, its increase is negatively
related to the welfare of citizens –as evidenced by the reduction in security
perception–, investment decisions, and the reduction of public investment in
sectors that yield higher welfare benefits since resources are diverted to crime
prevention investments.
In some specific cases, migration may be related to crime. When low-
skilled migrants arrive in areas where job opportunities are scarce and their
contact networks are typically reduced, their chances of finding employment
are diminished. At the same time, the possibility of engaging in illegal
activities is heightened. An important factor that increases participation in
illegal activities is that these activities generate high income with little effort.2
The government’s effort to punish crime is a crucial factor in its reduction, as
stated by Ehrlich (1973).
Despite the general belief that there is a link between migration and
crime, empirical research exploring this relationship is sparse (Reid et al.,
2005). Reid et al. (2005) examine the effects of migration on metropolitan
crime rates. After controlling for a host of demographic and economic
characteristics, they find that migration does not increase crime rates in
the receiving areas and that some aspects of migration may even reduce
crime in metropolitan areas. A similar approach was employed by Bianchi
et al. (2012), who found that migration leads to an increase in the incidence
of robberies but does not impact other types of crime. A critical aspect
1 In this perspective, for the Colombian case, Calderon Mejía and Ibáñez (2016), find that
internal migration substantially reduces wages for unskilled urban workers who compete for
jobs with forced migrants.
2 Psychological and social factors are also important determinants of criminal behavior (see
Hirschi and Stark 1969).
141
142
of these studies is that they examine international migration, which means
the opportunity cost of committing a crime is higher for immigrants. In
a similar vein, Kubrin et al. (2016) observed a decrease in violent crime
rates in neighborhoods in southern California that experience a large influx
of immigrants from Northern Africa. However, a higher proportion of
violent crimes are committed in areas with higher rates of Central American
immigrants.
Various studies analyze both the positive and negative effects of the
relationship between migration and crime. Bell and Machin (2013) find
that a higher influx of migrants leads to increased crime rates. They
provide evidence of higher seven categories of crime rates in areas of
Wales where asylum seekers are located. Nunziata (2015) demonstrates a
low positive correlation for three specifications regarding the relationship
between migration and crime in European countries. In the same line of
argumentation, Spenkush (2014) discovers that in the U.S., a 10% increase in
the migrant population leads to a 1.2% rise in property crime rate, with no
effect on violent crime rates. According to Wadsworth (2010), there was a
reduction in homicide and robbery rates in the U.S. between 1990 and 2000,
partially due to increased immigration. These results are similar to those of
Ousey and Kubrin (2009) for 159 large U.S. cities from 1980 to 2000.
Internal migration can be the result of governmental policies. In such
a scenario, a shift in population will occur that will interfere with the
community’s capacity to inhibit crimes. It will increase population hetero-
geneity and promote socioeconomic disadvantages (Treyger, 2013).
An essential attribute of internal migration is that it is predominantly
associated with a movement of individuals from rural to urban areas (Meng
& Zhang, 2013). Consistent with research on the relationship between
international migration and crime, the findings regarding the relationship
between internal migration and crime are also ambiguous. For the Canadian
case, Andresen (2013) affirms that there is a lack of evidence regarding the
effect of local migration on crime rates. Meng and Zhang (2013) concur with
these assertions regarding rural-to-urban domestic migrations in China. On
the contrary, Treyger (2013) demonstrates an increase in crime rates caused
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
of these studies is that they examine international migration, which means
the opportunity cost of committing a crime is higher for immigrants. In
a similar vein, Kubrin et al. (2016) observed a decrease in violent crime
rates in neighborhoods in southern California that experience a large influx
of immigrants from Northern Africa. However, a higher proportion of
violent crimes are committed in areas with higher rates of Central American
immigrants.
Various studies analyze both the positive and negative effects of the
relationship between migration and crime. Bell and Machin (2013) find
that a higher influx of migrants leads to increased crime rates. They
provide evidence of higher seven categories of crime rates in areas of
Wales where asylum seekers are located. Nunziata (2015) demonstrates a
low positive correlation for three specifications regarding the relationship
between migration and crime in European countries. In the same line of
argumentation, Spenkush (2014) discovers that in the U.S., a 10% increase in
the migrant population leads to a 1.2% rise in property crime rate, with no
effect on violent crime rates. According to Wadsworth (2010), there was a
reduction in homicide and robbery rates in the U.S. between 1990 and 2000,
partially due to increased immigration. These results are similar to those of
Ousey and Kubrin (2009) for 159 large U.S. cities from 1980 to 2000.
Internal migration can be the result of governmental policies. In such
a scenario, a shift in population will occur that will interfere with the
community’s capacity to inhibit crimes. It will increase population hetero-
geneity and promote socioeconomic disadvantages (Treyger, 2013).
An essential attribute of internal migration is that it is predominantly
associated with a movement of individuals from rural to urban areas (Meng
& Zhang, 2013). Consistent with research on the relationship between
international migration and crime, the findings regarding the relationship
between internal migration and crime are also ambiguous. For the Canadian
case, Andresen (2013) affirms that there is a lack of evidence regarding the
effect of local migration on crime rates. Meng and Zhang (2013) concur with
these assertions regarding rural-to-urban domestic migrations in China. On
the contrary, Treyger (2013) demonstrates an increase in crime rates caused
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
by the government’s relocation of domestic migrants. According to Schultz
(1971), interregional migration in Colombia occurred prior to the internal
conflict due to market forces that attracted rural labor to the cities from
regions with relatively low returns to labor and an oversupply of labor.
Crime rates fluctuate in response to economic and demographic factors,
as explained by traditional approaches. However, criminal activities exhibit
spatial and temporal concentrations. Hence, an approach that includes
the spatial dynamics of the variables is necessary for analyzing patterns
and causes of crime rates (Almeida et al., 2003). The main theoretical
perspectives are related to the Routine Activities and Crime Hot Spots
approaches. The Routine Activities approach posits that the place determines
favorable or unfavorable conditions for criminal acts through two forms: first,
physical features influence the social control capacities of crime suppressors;
second, criminal actions are not randomly distributed in space since crime is
influenced by routine activities and the characteristics of each place, which
means criminal actions have a spatial concentration in areas conducive to
crime. The Crime Hot Spots approach associates the conditions of the
population and some land uses with the spatial concentration of crime.
Furthermore, physical conditions may serve as indicators of the social control
that the community can exercise in the area (Anselin et al., 2000).
Consequently, some studies include spatial effects to explain the relation-
ship between migration and crime. Arnio and Baumer (2012) provide ev-
idence of spatial heterogeneity effects in Chicago neighborhoods between
2007 and 2009, although the concentration of immigrants does not signifi-
cantly affect crime. Graif and Sampson (2009) suggest the same effect of spa-
tial patterns and migrants on homicide rates across Chicago neighborhoods
between 1990-2000. Using a spatial panel data model for Wales and England’s
migration-crime relationship between 2001 and 2011, Jaitman and Machin
(2013) identify evidence supporting this linkage. Additionally, Cracolici and
Uberti (2009) suggest that including spatial effects for analyzing crime rates
in Italian provinces and the concentration of foreigners in previous periods
are essential determinants of criminal activities in this region.
143
by the government’s relocation of domestic migrants. According to Schultz
(1971), interregional migration in Colombia occurred prior to the internal
conflict due to market forces that attracted rural labor to the cities from
regions with relatively low returns to labor and an oversupply of labor.
Crime rates fluctuate in response to economic and demographic factors,
as explained by traditional approaches. However, criminal activities exhibit
spatial and temporal concentrations. Hence, an approach that includes
the spatial dynamics of the variables is necessary for analyzing patterns
and causes of crime rates (Almeida et al., 2003). The main theoretical
perspectives are related to the Routine Activities and Crime Hot Spots
approaches. The Routine Activities approach posits that the place determines
favorable or unfavorable conditions for criminal acts through two forms: first,
physical features influence the social control capacities of crime suppressors;
second, criminal actions are not randomly distributed in space since crime is
influenced by routine activities and the characteristics of each place, which
means criminal actions have a spatial concentration in areas conducive to
crime. The Crime Hot Spots approach associates the conditions of the
population and some land uses with the spatial concentration of crime.
Furthermore, physical conditions may serve as indicators of the social control
that the community can exercise in the area (Anselin et al., 2000).
Consequently, some studies include spatial effects to explain the relation-
ship between migration and crime. Arnio and Baumer (2012) provide ev-
idence of spatial heterogeneity effects in Chicago neighborhoods between
2007 and 2009, although the concentration of immigrants does not signifi-
cantly affect crime. Graif and Sampson (2009) suggest the same effect of spa-
tial patterns and migrants on homicide rates across Chicago neighborhoods
between 1990-2000. Using a spatial panel data model for Wales and England’s
migration-crime relationship between 2001 and 2011, Jaitman and Machin
(2013) identify evidence supporting this linkage. Additionally, Cracolici and
Uberti (2009) suggest that including spatial effects for analyzing crime rates
in Italian provinces and the concentration of foreigners in previous periods
are essential determinants of criminal activities in this region.
143
144
Due to the decades-long clash between government forces and anti-
government insurgent groups, internal forced displacement has been the
leading cause of internal migration in Colombia. The intensification of the
internal armed conflict and the intention of illegal armed groups to expand
territory control and asset appropriation, in conjunction with the country’s
institutional weakness, have induced the mobility of the population to safer
areas in other municipalities (Ibáñez, 2009). Significant research has been
conducted on the effects of displacement on poverty, inequality, and the
labor market. Ibáñez and Vélez (2008) find that welfare losses caused by
forced displacement in Colombia represent 37% of the net present value of
rural lifetime aggregate consumption. They claim that a violent environment
modifies the net benefits of migration.3 Additionally, displaced persons are
forced to confront a decreased labor income, dire conditions in host areas, and
a severe disruption of risk-sharing mechanisms. The effects of government
programs are transitory and forced migrants are unable to recover the income
and welfare levels they had before displacement (Ibáñez & Moya, 2010).
Using a structural model of crime and inequality, Bourguignon et al.
(2017) show that income distribution in Colombia influences aggregate crime
and find that a significant number of criminals live in households with an
eighty percent below-mean income per capita. Furthermore, inflows of
forced migrants increase the proportion of low-skilled and informal workers,
causing a reduction in informal wages (Calderón-Mejía & Ibáñez, 2016).
Consequently, the welfare loss of forcibly displaced persons can determine
engagement in illegal actions within this population and generate labor market
problems; informal wage reduction and unequal distribution of income
influence the criminal propensity of the entire population.
While internal forced displacement and its associated impacts have been
widely examined, a variety of repercussions of the massive displacement that
occurred in the nation have also been observed. These consequences include
the breakdown of social connections, fragmentation of families, elevated
3 The criminal and reception of forcibly displaced persons rates are obtained by multiplying
each crime variable by a factor of 100,000/total municipality population; the probability of
apprehension for each kind of crime is determined by dividing the number of captures by
the number of crimes committed in the same category.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Due to the decades-long clash between government forces and anti-
government insurgent groups, internal forced displacement has been the
leading cause of internal migration in Colombia. The intensification of the
internal armed conflict and the intention of illegal armed groups to expand
territory control and asset appropriation, in conjunction with the country’s
institutional weakness, have induced the mobility of the population to safer
areas in other municipalities (Ibáñez, 2009). Significant research has been
conducted on the effects of displacement on poverty, inequality, and the
labor market. Ibáñez and Vélez (2008) find that welfare losses caused by
forced displacement in Colombia represent 37% of the net present value of
rural lifetime aggregate consumption. They claim that a violent environment
modifies the net benefits of migration.3 Additionally, displaced persons are
forced to confront a decreased labor income, dire conditions in host areas, and
a severe disruption of risk-sharing mechanisms. The effects of government
programs are transitory and forced migrants are unable to recover the income
and welfare levels they had before displacement (Ibáñez & Moya, 2010).
Using a structural model of crime and inequality, Bourguignon et al.
(2017) show that income distribution in Colombia influences aggregate crime
and find that a significant number of criminals live in households with an
eighty percent below-mean income per capita. Furthermore, inflows of
forced migrants increase the proportion of low-skilled and informal workers,
causing a reduction in informal wages (Calderón-Mejía & Ibáñez, 2016).
Consequently, the welfare loss of forcibly displaced persons can determine
engagement in illegal actions within this population and generate labor market
problems; informal wage reduction and unequal distribution of income
influence the criminal propensity of the entire population.
While internal forced displacement and its associated impacts have been
widely examined, a variety of repercussions of the massive displacement that
occurred in the nation have also been observed. These consequences include
the breakdown of social connections, fragmentation of families, elevated
3 The criminal and reception of forcibly displaced persons rates are obtained by multiplying
each crime variable by a factor of 100,000/total municipality population; the probability of
apprehension for each kind of crime is determined by dividing the number of captures by
the number of crimes committed in the same category.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
unemployment rates, unfavorable socio-economic conditions in numerous
destination cities, and depletion of economic, financial, physical, human, and
social resources throughout the country (Ibáñez & Moya, 2006; Ibáñez &
Velásquez, 2009; Ibáñez & Vélez, 2008). However, due to inconclusive
evidence, comprehensive research on its repercussions on crime rates in
receiving areas still needs to be conducted. Previous studies, such as Roche-
Villarreal’s (2012) investigation into the effect of displacement on property-
related crimes in municipalities, have not established a clear relationship (Reid
et al., 2005). The present study is motivated by the persistent Colombian
conflict, which has spanned over half a century and led to a staggering
7,849,014 documented cases of displacement, as reported by the Unidad para
las Víctimas.
II. Empirical Strategy
Four data sources were used in this research. The annual reports of the
Colombian National Police (Policía Nacional de Colombia, n.d.) provide
the crime categories for the municipalities. Unidad para las Víctimas (2018)
generates statistical data on internal forced displacement. The National
Department of Statistics (DANE, n.d.) is the source of demographic data.
Data from the Municipal Panel of CEDE of the Centro de Estudios para el
Desarrollo Económico (CEDE) at the Universidad de Los Andes (Acevedo
& Bornacelly, 2014) were also used.
Table 1 presents information on the central variables included in the analysis
and the number of occurrences for each crime category. This research examines
five types of crime: homicides, kidnapping, personal injuries, automobile theft,
and residential burglary. These crime categories were selected because they
are the most frequently reported and registered. The primary explanatory
variable is the number of receptions of forced displacement populations in each
municipality. Additionally, this research considers significant crime predictors
and population heterogeneity determinants as control variables.
4 The criminal and reception of forcibly displaced persons rates are obtained by multiplying
each crime variable by a factor of 100,000/total municipality population; the probability of
apprehension for each kind of crime is determined by dividing the number of captures by
145
unemployment rates, unfavorable socio-economic conditions in numerous
destination cities, and depletion of economic, financial, physical, human, and
social resources throughout the country (Ibáñez & Moya, 2006; Ibáñez &
Velásquez, 2009; Ibáñez & Vélez, 2008). However, due to inconclusive
evidence, comprehensive research on its repercussions on crime rates in
receiving areas still needs to be conducted. Previous studies, such as Roche-
Villarreal’s (2012) investigation into the effect of displacement on property-
related crimes in municipalities, have not established a clear relationship (Reid
et al., 2005). The present study is motivated by the persistent Colombian
conflict, which has spanned over half a century and led to a staggering
7,849,014 documented cases of displacement, as reported by the Unidad para
las Víctimas.
II. Empirical Strategy
Four data sources were used in this research. The annual reports of the
Colombian National Police (Policía Nacional de Colombia, n.d.) provide
the crime categories for the municipalities. Unidad para las Víctimas (2018)
generates statistical data on internal forced displacement. The National
Department of Statistics (DANE, n.d.) is the source of demographic data.
Data from the Municipal Panel of CEDE of the Centro de Estudios para el
Desarrollo Económico (CEDE) at the Universidad de Los Andes (Acevedo
& Bornacelly, 2014) were also used.
Table 1 presents information on the central variables included in the analysis
and the number of occurrences for each crime category. This research examines
five types of crime: homicides, kidnapping, personal injuries, automobile theft,
and residential burglary. These crime categories were selected because they
are the most frequently reported and registered. The primary explanatory
variable is the number of receptions of forced displacement populations in each
municipality. Additionally, this research considers significant crime predictors
and population heterogeneity determinants as control variables.
4 The criminal and reception of forcibly displaced persons rates are obtained by multiplying
each crime variable by a factor of 100,000/total municipality population; the probability of
apprehension for each kind of crime is determined by dividing the number of captures by
145
146
Table 1. Description of variables included in the analysis
Variable Description Source
Homicide rates Logarithm of the number of homicides
per 100,000 inhabitants in each
municipality
Policía Nacional Colombia
Kidnapping rate Logarithm of the number of kidnap-
pings per 100,000 inhabitants in each
municipality
Policía Nacional Colombia
Personal injuries rate Logarithm of the number of personal
injuries per 100,000 inhabitants in each
municipality
Policía Nacional Colombia
Residential burglary rate Logarithm of the number of residential
burglaries per 100,000 inhabitants in
each municipality
Policía Nacional Colombia
Automobile theft rate Logarithm of the number of automobile
thefts per 100,000 inhabitants in each
municipality
Policía Nacional Colombia
Reception FDP rate4 Logarithm of the number of recep-
tions of forcibly displaced persons per
100,000 inhabitants in each municipal-
ity
Unidad para las Víctimas
Percentage Male 19-34 Percentage of male population between
15-34 years for each municipality
DANE
Log Population Logarithm of the total population for
each municipality
Panel CEDE
Probability of apprehension Number of arrests divided by the num-
ber of crimes in each crime category and
municipality
Policía Nacional Colombia
Source: Own elaboration.
The control variables were selected based on previous empirical studies
on crime and data availability for Colombian municipalities. Other research
considers population density an essential determinant of criminal acts because
a larger population in an area provides more opportunities for offending and
the number of crimes committed in the same category.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Table 1. Description of variables included in the analysis
Variable Description Source
Homicide rates Logarithm of the number of homicides
per 100,000 inhabitants in each
municipality
Policía Nacional Colombia
Kidnapping rate Logarithm of the number of kidnap-
pings per 100,000 inhabitants in each
municipality
Policía Nacional Colombia
Personal injuries rate Logarithm of the number of personal
injuries per 100,000 inhabitants in each
municipality
Policía Nacional Colombia
Residential burglary rate Logarithm of the number of residential
burglaries per 100,000 inhabitants in
each municipality
Policía Nacional Colombia
Automobile theft rate Logarithm of the number of automobile
thefts per 100,000 inhabitants in each
municipality
Policía Nacional Colombia
Reception FDP rate4 Logarithm of the number of recep-
tions of forcibly displaced persons per
100,000 inhabitants in each municipal-
ity
Unidad para las Víctimas
Percentage Male 19-34 Percentage of male population between
15-34 years for each municipality
DANE
Log Population Logarithm of the total population for
each municipality
Panel CEDE
Probability of apprehension Number of arrests divided by the num-
ber of crimes in each crime category and
municipality
Policía Nacional Colombia
Source: Own elaboration.
The control variables were selected based on previous empirical studies
on crime and data availability for Colombian municipalities. Other research
considers population density an essential determinant of criminal acts because
a larger population in an area provides more opportunities for offending and
the number of crimes committed in the same category.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
less social control over criminals (Reid et al., 2005). Furthermore, males are
more prone to commit criminal acts (Cheng et al., 2017). The empirical
strategy incorporates the proportion of males among the total population
to control for these effects. Additionally, the econometric model includes the
apprehension rate for each crime since a higher probability of arrest reduces
individual incentives to commit a crime (Ehrlich, 1973). Table 2 presents the
descriptive statistics for the variables included in the analysis.
Table 2. Descriptive statistics5
Variable Obs. Mean Std. Dev. Min Max
Homicide rate 15 344 36.087 60.366 0 1 607.467
Kidnapping rate 15 344 20.480 8.874 0 255.427
Personal injuries rate 15 344 98.948 132.339 0 4 723.072
Residential burglary rate 15 344 23.035 44.962 0 1 087.326
Automobile theft 15 344 5.406 17.694 0 886.143
Reception of forcibly displaced
persons rate
15 344 704.94 1 849.586 0 70 958.08
Total population 15 344 41 197.11 251 271.3 837 798 000,1
Percentage of males 15-34 15 344 32.496 2.946 20.743 68.436
Prob. of capture for homicide 15 344 43.016 92.910 0 2900
Prob. of capture for kidnapping 15 344 353.630 2452.473 0 118200
Prob. of capture for personal
injuries
15 344 7.971 51 916 0 2300
Prob. of capture for residential
burglary
15 344 18.898 82.436 0 3 100
Prob. of capture for
Automobile theft
15 344 9.572 50.002 0 1 200
Source: Own elaboration.
Becker (1968) and Ehrlich (1973) provide the foundations for the
base model in this paper. It assumes the number of criminal actions is
monotonically related to the probability of arrest (p), the marginal cost of
punishment (f ), the marginal earning of illegal (wc) and legal (wl) activities,
5 The descriptive statistics in Table 2 correspond to the information of the variables by levels.
147
less social control over criminals (Reid et al., 2005). Furthermore, males are
more prone to commit criminal acts (Cheng et al., 2017). The empirical
strategy incorporates the proportion of males among the total population
to control for these effects. Additionally, the econometric model includes the
apprehension rate for each crime since a higher probability of arrest reduces
individual incentives to commit a crime (Ehrlich, 1973). Table 2 presents the
descriptive statistics for the variables included in the analysis.
Table 2. Descriptive statistics5
Variable Obs. Mean Std. Dev. Min Max
Homicide rate 15 344 36.087 60.366 0 1 607.467
Kidnapping rate 15 344 20.480 8.874 0 255.427
Personal injuries rate 15 344 98.948 132.339 0 4 723.072
Residential burglary rate 15 344 23.035 44.962 0 1 087.326
Automobile theft 15 344 5.406 17.694 0 886.143
Reception of forcibly displaced
persons rate
15 344 704.94 1 849.586 0 70 958.08
Total population 15 344 41 197.11 251 271.3 837 798 000,1
Percentage of males 15-34 15 344 32.496 2.946 20.743 68.436
Prob. of capture for homicide 15 344 43.016 92.910 0 2900
Prob. of capture for kidnapping 15 344 353.630 2452.473 0 118200
Prob. of capture for personal
injuries
15 344 7.971 51 916 0 2300
Prob. of capture for residential
burglary
15 344 18.898 82.436 0 3 100
Prob. of capture for
Automobile theft
15 344 9.572 50.002 0 1 200
Source: Own elaboration.
Becker (1968) and Ehrlich (1973) provide the foundations for the
base model in this paper. It assumes the number of criminal actions is
monotonically related to the probability of arrest (p), the marginal cost of
punishment (f ), the marginal earning of illegal (wc) and legal (wl) activities,
5 The descriptive statistics in Table 2 correspond to the information of the variables by levels.
147
148
the probability of unemployment (u) and a set of variables (π) that may affect
the frequency of crimes.
qi = ψ(p, f, wc, wl, u, π). (1)
This approach assumes that all individuals are identical, a quality that
permits the aggregation of individual functions. Therefore, the number of
crimes in society is defined as an aggregate behavioral function,
Q = Ψ(P, F, Yc, Yl, U, Π), (2)
where the variables denote the mean of the components of the individual
behavioral function. Ehrlich (1973) proposes some behavioral implications
for this model: An increase in the probability of arrest and in the marginal cost
of punishment reduce the incentives to engage in illegal activities. Similarly, an
increase in the probability of being unemployed and in the earnings obtained
through illegal activities generate a rise in incentives for committing a crime.
Consequently, following the approaches of Becker (1968) and Ehrlich
(1973), this paper presents the supply offenses function, where each kind
of crime depends on the reception of forcibly displaced persons and the
determinants of heterogeneity.
( Q
N
)
i
= A F DP iXiW e(μi+τt+εit). (3)
In equation (3), the component
( Q
N
)
i represents the number of specific
crimes in area i, where N is a population scale factor, A is a constant, F DP i
is the number of forcibly displaced persons received in region i, Xi is a
portmanteau of heterogeneity variables of population in i and the probability
of apprehension for the same type of crime in the area, W are the spatial
effects, and ui summarizes the effects of the physic, income and other
nonquantifiable variables in the same area. From the linearization of equation
(3), the econometric specification for the panel model is:
logCit = α + β′ilogF DP it + βiXit + W + μi + τt + εit, (4)
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
the probability of unemployment (u) and a set of variables (π) that may affect
the frequency of crimes.
qi = ψ(p, f, wc, wl, u, π). (1)
This approach assumes that all individuals are identical, a quality that
permits the aggregation of individual functions. Therefore, the number of
crimes in society is defined as an aggregate behavioral function,
Q = Ψ(P, F, Yc, Yl, U, Π), (2)
where the variables denote the mean of the components of the individual
behavioral function. Ehrlich (1973) proposes some behavioral implications
for this model: An increase in the probability of arrest and in the marginal cost
of punishment reduce the incentives to engage in illegal activities. Similarly, an
increase in the probability of being unemployed and in the earnings obtained
through illegal activities generate a rise in incentives for committing a crime.
Consequently, following the approaches of Becker (1968) and Ehrlich
(1973), this paper presents the supply offenses function, where each kind
of crime depends on the reception of forcibly displaced persons and the
determinants of heterogeneity.
( Q
N
)
i
= A F DP iXiW e(μi+τt+εit). (3)
In equation (3), the component
( Q
N
)
i represents the number of specific
crimes in area i, where N is a population scale factor, A is a constant, F DP i
is the number of forcibly displaced persons received in region i, Xi is a
portmanteau of heterogeneity variables of population in i and the probability
of apprehension for the same type of crime in the area, W are the spatial
effects, and ui summarizes the effects of the physic, income and other
nonquantifiable variables in the same area. From the linearization of equation
(3), the econometric specification for the panel model is:
logCit = α + β′ilogF DP it + βiXit + W + μi + τt + εit, (4)
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
where i denotes municipalities and t denotes time. Cit is the crime rate per
100.000 inhabitants in region i and year t, F DP it is the forcibly displaced
persons reception rate per 100.000 inhabitants, Xit is a portmanteau of
heterogeneity variables and the probability of apprehension for the same type
of crime in the area, μi are municipality fixed effects, τt are time fixed effects,
and εit is the error term.
This extension of the Becker-Ehrlich model allows the inclusion of
socioeconomic, geographic, and demographic variables (Cracolici & Uberti,
2009); it is commonly used to analyze the relationship between all types of
migration and crime. This model allows us to consider the reception of
forcibly displaced persons and the spatial allocation of the variables. The
Spatial General Nested (SGN) model is deemed the most appropriate for
our research as it enables the incorporation of spatial determinants of crime
allocation, forcibly displaced persons, and population heterogeneity variables.
Additionally, it includes the spatial error term containing the omitted spatial
variables that are correlated with crime. Equations (5) and (6) present the
SGN model as applied in our work:
logCit = α + δW logCit + β′ilogF Dit + βiXit + W Zitθ + ui,t (5)
uit = λW uit + μi + τt + εit, (6)
where i denotes municipalities, j represents the contiguous municipalities, and
t denotes time. Cit is the crime rate per 100.000 inhabitants in region i and
year t, F DP it is the forcibly displaced persons reception rate per 100.000
inhabitants, Xit is the logarithm of population, the percentage of males
between 19-34 years of age and the probability of apprehension, Zjt denotes
the explanatory variables in the neighbor municipalities, μi are municipality
fixed effects, τt are time fixed effects, εit is the error term, and ujt is the
spatial error term.
Before estimating the dynamics of spatial patterns, the Explanatory
Spatial Data Analysis (ESDA) is applied as a descriptive step for identifying
the presence of spatial effects in the analysis area (Ye & Wu, 2011). In
Colombia, criminal gangs, drug cartels, and illegal armed groups that operate
across geographical areas are essential determinants of crime. The change
in dynamics of these criminal associations in conjunction with governmental
149
where i denotes municipalities and t denotes time. Cit is the crime rate per
100.000 inhabitants in region i and year t, F DP it is the forcibly displaced
persons reception rate per 100.000 inhabitants, Xit is a portmanteau of
heterogeneity variables and the probability of apprehension for the same type
of crime in the area, μi are municipality fixed effects, τt are time fixed effects,
and εit is the error term.
This extension of the Becker-Ehrlich model allows the inclusion of
socioeconomic, geographic, and demographic variables (Cracolici & Uberti,
2009); it is commonly used to analyze the relationship between all types of
migration and crime. This model allows us to consider the reception of
forcibly displaced persons and the spatial allocation of the variables. The
Spatial General Nested (SGN) model is deemed the most appropriate for
our research as it enables the incorporation of spatial determinants of crime
allocation, forcibly displaced persons, and population heterogeneity variables.
Additionally, it includes the spatial error term containing the omitted spatial
variables that are correlated with crime. Equations (5) and (6) present the
SGN model as applied in our work:
logCit = α + δW logCit + β′ilogF Dit + βiXit + W Zitθ + ui,t (5)
uit = λW uit + μi + τt + εit, (6)
where i denotes municipalities, j represents the contiguous municipalities, and
t denotes time. Cit is the crime rate per 100.000 inhabitants in region i and
year t, F DP it is the forcibly displaced persons reception rate per 100.000
inhabitants, Xit is the logarithm of population, the percentage of males
between 19-34 years of age and the probability of apprehension, Zjt denotes
the explanatory variables in the neighbor municipalities, μi are municipality
fixed effects, τt are time fixed effects, εit is the error term, and ujt is the
spatial error term.
Before estimating the dynamics of spatial patterns, the Explanatory
Spatial Data Analysis (ESDA) is applied as a descriptive step for identifying
the presence of spatial effects in the analysis area (Ye & Wu, 2011). In
Colombia, criminal gangs, drug cartels, and illegal armed groups that operate
across geographical areas are essential determinants of crime. The change
in dynamics of these criminal associations in conjunction with governmental
149
150
actions lead to a temporal and spatial fluctuation in crime rates. Therefore,
our study provides an ESDA of the five categories of crime rates; however,
due to space limitations, we only present the analysis for the first, seventh,
and thirteenth years of the period analyzed (2003, 2009, and 2016). This
procedure consists of sampling the selected period to detect any changes in
the spatial dynamics of the variables and to verify the inclusion of these effects
in the empirical model.
The figures depict the spatial dynamics of the five crime categories. The
homicide rate is illustrated in Figure 1, the kidnapping rate in Figure 2,
the personal injuries rate in Figure 3, the automobile theft rate in Figure
4, and the residential burglary rate in Figure 5. Additionally, Figure 6
depicts the spatial dynamics of internal forced displacement rates. Specifically,
municipalities with high homicide rates are concentrated in the following
departments:6 Andean region, Cesar and Guajira (Caribbean region); Valle
del Cauca and Choco (Pacific region); Arauca, Meta and Casanare (Orinoquia
region); Putumayo and Caqueta (Amazonas region). The spatial dynamics
changed for the 6th and 13th years since the number of municipalities with
homicide rates above the mean decreased, particularly in the Pacific region.
Kidnapping rates exhibit a similar spatial distribution but a smaller number
of municipalities above the mean as compared to homicide rates. As in the
case of homicides, kidnapping rates experienced a significant reduction in the
years 2009 and 2016. For this category, the municipalities of Arauca, Casanare,
and Norte de Santander (North-East), and Nari no, Putumayo, and Cauca
(South-West) exhibit high rates.
Personal injury rates showed a similar distribution in 2003, with the
highest rates occurring in municipalities located in the eastern, central, and
western parts of Colombia. In 2009, an upward trend in the number
of municipalities with high rates was observed, especially in the eastern
and northern areas. In 2016, municipalities with rates above the mean
were concentrated in the Andean, northeastern, and southwestern regions.
Residential burglary rates reveal a similar geographic distribution for the three
years of analysis; this rate increases with each passing year and presents
6 Departments are political units equivalent to states or provinces.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
actions lead to a temporal and spatial fluctuation in crime rates. Therefore,
our study provides an ESDA of the five categories of crime rates; however,
due to space limitations, we only present the analysis for the first, seventh,
and thirteenth years of the period analyzed (2003, 2009, and 2016). This
procedure consists of sampling the selected period to detect any changes in
the spatial dynamics of the variables and to verify the inclusion of these effects
in the empirical model.
The figures depict the spatial dynamics of the five crime categories. The
homicide rate is illustrated in Figure 1, the kidnapping rate in Figure 2,
the personal injuries rate in Figure 3, the automobile theft rate in Figure
4, and the residential burglary rate in Figure 5. Additionally, Figure 6
depicts the spatial dynamics of internal forced displacement rates. Specifically,
municipalities with high homicide rates are concentrated in the following
departments:6 Andean region, Cesar and Guajira (Caribbean region); Valle
del Cauca and Choco (Pacific region); Arauca, Meta and Casanare (Orinoquia
region); Putumayo and Caqueta (Amazonas region). The spatial dynamics
changed for the 6th and 13th years since the number of municipalities with
homicide rates above the mean decreased, particularly in the Pacific region.
Kidnapping rates exhibit a similar spatial distribution but a smaller number
of municipalities above the mean as compared to homicide rates. As in the
case of homicides, kidnapping rates experienced a significant reduction in the
years 2009 and 2016. For this category, the municipalities of Arauca, Casanare,
and Norte de Santander (North-East), and Nari no, Putumayo, and Cauca
(South-West) exhibit high rates.
Personal injury rates showed a similar distribution in 2003, with the
highest rates occurring in municipalities located in the eastern, central, and
western parts of Colombia. In 2009, an upward trend in the number
of municipalities with high rates was observed, especially in the eastern
and northern areas. In 2016, municipalities with rates above the mean
were concentrated in the Andean, northeastern, and southwestern regions.
Residential burglary rates reveal a similar geographic distribution for the three
years of analysis; this rate increases with each passing year and presents
6 Departments are political units equivalent to states or provinces.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
Figure 1. Standard deviation map of homicide rate
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
an increase number of municipalities with higher rates in 2016, particularly
in the Orinoquia and Andean regions. The municipalities with the highest
Automobile theft rates exhibit a constant spatial distribution and are located
in central, northeastern, and southwestern areas. Reception rates of forcibly
displaced persons in 2003 were concentrated in bordering areas of the
Colombian territory, especially in Choco, Antioquia, Putumayo, Valle del
Cauca, Guajira, Cesar, and Magdalena. The number of municipalities with
a rate of forcibly displaced persons below the mean increased in 2009, and
this trend was even more pronounced in 2016. Municipalities situated along
the Pacific Coast have exhibited higher rates over the past year.
Spatial panel models offer technical advantages over traditional ap-
proaches, such as less collinearity, increased degrees of freedom, and the abil-
ity to incorporate numerous effects. The procedure for proper model identi-
fication consists of identifying the adequate model without spatial effects and
choosing from the Pooled, Fixed Effects Panel, and Random Effects Panel
models. The second phase involves estimating the general model of the spa-
tial Durbin model (SDM) and testing whether it is more appropriate than the
spatial lag model (SAR), spatial error model (SEM), spatial lag of X (SLX),
spatial Durbin error model (SDEM), general nesting spatial model (GNS), or
spatial autocorrelation model (SAC) (Torres-Preciado et al., 2017).
151
Figure 1. Standard deviation map of homicide rate
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
an increase number of municipalities with higher rates in 2016, particularly
in the Orinoquia and Andean regions. The municipalities with the highest
Automobile theft rates exhibit a constant spatial distribution and are located
in central, northeastern, and southwestern areas. Reception rates of forcibly
displaced persons in 2003 were concentrated in bordering areas of the
Colombian territory, especially in Choco, Antioquia, Putumayo, Valle del
Cauca, Guajira, Cesar, and Magdalena. The number of municipalities with
a rate of forcibly displaced persons below the mean increased in 2009, and
this trend was even more pronounced in 2016. Municipalities situated along
the Pacific Coast have exhibited higher rates over the past year.
Spatial panel models offer technical advantages over traditional ap-
proaches, such as less collinearity, increased degrees of freedom, and the abil-
ity to incorporate numerous effects. The procedure for proper model identi-
fication consists of identifying the adequate model without spatial effects and
choosing from the Pooled, Fixed Effects Panel, and Random Effects Panel
models. The second phase involves estimating the general model of the spa-
tial Durbin model (SDM) and testing whether it is more appropriate than the
spatial lag model (SAR), spatial error model (SEM), spatial lag of X (SLX),
spatial Durbin error model (SDEM), general nesting spatial model (GNS), or
spatial autocorrelation model (SAC) (Torres-Preciado et al., 2017).
151
152
Figure 2. Standard deviation map of kidnapping rate
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
Table 3. Moran test for spatial dependence
variable Test 2003 2009 2016
Homicide Rates χ2 99.44 64.08 30.58
Prob. > χ2 0.000 0.000 0.000
Kidnapping Rate χ2 21.87 11.36 2.4
Prob. > χ2 0.000 0.000 0.121
Personal Injuries Rate χ2 61.95 97.11 80.07
Prob. > χ2 0.000 0.000 0.000
Residential Burglary Rate χ2 221.32 194.79 62.24
Prob. > χ2 0.000 0.000 0.000
Automobile Theft Rate χ2 132.39 139.14 226.14
Prob. > χ2 0.000 0.000 0.000
Source: Own elaboration
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Figure 2. Standard deviation map of kidnapping rate
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
Table 3. Moran test for spatial dependence
variable Test 2003 2009 2016
Homicide Rates χ2 99.44 64.08 30.58
Prob. > χ2 0.000 0.000 0.000
Kidnapping Rate χ2 21.87 11.36 2.4
Prob. > χ2 0.000 0.000 0.121
Personal Injuries Rate χ2 61.95 97.11 80.07
Prob. > χ2 0.000 0.000 0.000
Residential Burglary Rate χ2 221.32 194.79 62.24
Prob. > χ2 0.000 0.000 0.000
Automobile Theft Rate χ2 132.39 139.14 226.14
Prob. > χ2 0.000 0.000 0.000
Source: Own elaboration
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
Figure 3. Standard deviation map of personal injuries rate
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
Figure 4. Standard deviation map of residential burglary rate
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
153
Figure 3. Standard deviation map of personal injuries rate
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
Figure 4. Standard deviation map of residential burglary rate
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
153
154
Figure 5. Standard deviation map of automobile theft rat
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
Figure 6. Standard deviation map of reception of FDP rate
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
III. Estimations Results
In line with the procedure suggested by Elhorst (2014), this paper initially
estimates pooled, fixed effect, and random effect panel models for each
type of crime category. Table 4 provides information for the fixed effects
panel regression. The estimations use a two-period lag of the probability of
apprehension due to the problem of endogeneity and counter-intuitive results
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Figure 5. Standard deviation map of automobile theft rat
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
Figure 6. Standard deviation map of reception of FDP rate
Source: Own elaboration with data from Policia Nacional de Colombia, Panel CEDE and Unidad para las
Víctimas.
III. Estimations Results
In line with the procedure suggested by Elhorst (2014), this paper initially
estimates pooled, fixed effect, and random effect panel models for each
type of crime category. Table 4 provides information for the fixed effects
panel regression. The estimations use a two-period lag of the probability of
apprehension due to the problem of endogeneity and counter-intuitive results
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
derived from using this variable without lag, consistent with the findings of
Roche-Villareal (2012), which indicate a positive relationship between crime
and the probability of apprehension. Tables A1 and A2 of the Appendix
present the estimation results for our panel using this variable without lag
and with one period lag to show the robustness of using the instrument of
the probability of apprehension with a two-period lag in our analysis.
Traditional econometric theory posits that fixed effects is the most
appropriate estimation method because it accounts for the heterogeneity of
each municipality. Furthermore, considering the similarity between the N of
the sample and the total number of municipalities, fixed effects is ostensibly
the correct estimation method. Regardless, in the fixed effect model, as
seen in Table 4, the negative sign for certain crimes in the percentage of
males aged 19-34 years is disputed by the theory that a more significant
proportion of males with a predisposition to commit crimes is found this age
group. Additionally, the lagged probability of apprehension shows a negative
relationship with crime, suggesting a time delay between police efficiency and
its effect on crime levels.
Table 4. Fixed effect estimations
Variable Homicide
rate model
Kidnapping
rate model
Personal
injuries rate
model
Residential
Burglary
model
Automobile
theft rate
Reception FDP
rate
0.020* 0.010 -0.055*** -0.013 -0.008
Population 0.966 -3.437*** -0.328 1.695* 1.875**
Percentage Male
19-34
-0.233*** -0.401*** 0.456*** 0.2088*** 0.030
Probability of ap-
prehension (-2)
-0.001** -0.00003** -0.0009** -0.001 -0.0016**
Constant -1.218 38.309*** -8.684 -24.815*** -24.432***
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
155
derived from using this variable without lag, consistent with the findings of
Roche-Villareal (2012), which indicate a positive relationship between crime
and the probability of apprehension. Tables A1 and A2 of the Appendix
present the estimation results for our panel using this variable without lag
and with one period lag to show the robustness of using the instrument of
the probability of apprehension with a two-period lag in our analysis.
Traditional econometric theory posits that fixed effects is the most
appropriate estimation method because it accounts for the heterogeneity of
each municipality. Furthermore, considering the similarity between the N of
the sample and the total number of municipalities, fixed effects is ostensibly
the correct estimation method. Regardless, in the fixed effect model, as
seen in Table 4, the negative sign for certain crimes in the percentage of
males aged 19-34 years is disputed by the theory that a more significant
proportion of males with a predisposition to commit crimes is found this age
group. Additionally, the lagged probability of apprehension shows a negative
relationship with crime, suggesting a time delay between police efficiency and
its effect on crime levels.
Table 4. Fixed effect estimations
Variable Homicide
rate model
Kidnapping
rate model
Personal
injuries rate
model
Residential
Burglary
model
Automobile
theft rate
Reception FDP
rate
0.020* 0.010 -0.055*** -0.013 -0.008
Population 0.966 -3.437*** -0.328 1.695* 1.875**
Percentage Male
19-34
-0.233*** -0.401*** 0.456*** 0.2088*** 0.030
Probability of ap-
prehension (-2)
-0.001** -0.00003** -0.0009** -0.001 -0.0016**
Constant -1.218 38.309*** -8.684 -24.815*** -24.432***
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
155
156
In view of the above, this paper goes further and corrects estimation
problems by employing spatial models since migration has a significant spatial
component, as suggested by Cracolici and Uberti (2009). Consequently,
under this methodological approach, the spatial model selection criteria
consist of comparing AIC statistics. Our results of the Spatial General
Nested Model (SGN) model yield the lowest values, confirming the empirical
relevance of using this spatial model (see Tables A3-A8 of the Appendix for
SAR, SEM, SLX, SAC, SDM, and SDEM estimations). The estimation result
is displayed in Table 5. Additionally, Table 6 presents the decomposition of
the direct, indirect, and total effects of the SGN for the models.
The coefficients in section 1 of Table 5 correspond to the effect of
the variables on the same type of crime within the same municipality;
consequently, section 2 of Table 5 presents the spatial effects of the variables
in the nearest municipalities. The significance of most of the coefficients
associated with the spatial effects of the variables confirms the relevance
of spatial effects in the analysis of crime and displacement. Furthermore,
economic theory posits that this spatial coefficient presents the correct sign:
a percentage of males between 19 and 34 causes an increase in crime, and a
probability of apprehension is a deterrent to crime (Ehrlich, 1973). According
to Anselin et al. (2000), criminals may commit illegal actions in transit areas
near their residences; this supports the significance in most spatial coefficients
but less so in local coefficients.
Additionally, the spatial effects of forcibly displaced persons yield a
positive sign for homicide rates and a negative sign for personal injuries and
residential burglary. These findings explain that forcibly displaced persons, in
some cases, increase social control for certain crimes while simultaneously
increasing the occurrence of others (Bucheli et al.,2019). Similar effects
of immigrants’ concentration on crime rates were observed in Los Angeles
neighborhoods (Macdonald et al., 2013) and neighborhoods in England (Bell
& Machin, 2013).
Consequently, the results of spatial effects on the reception of forcibly
displaced persons suggest that only personal injury cases have direct and
significant effects; that is, a 1% increase in the number of forcibly displaced
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
In view of the above, this paper goes further and corrects estimation
problems by employing spatial models since migration has a significant spatial
component, as suggested by Cracolici and Uberti (2009). Consequently,
under this methodological approach, the spatial model selection criteria
consist of comparing AIC statistics. Our results of the Spatial General
Nested Model (SGN) model yield the lowest values, confirming the empirical
relevance of using this spatial model (see Tables A3-A8 of the Appendix for
SAR, SEM, SLX, SAC, SDM, and SDEM estimations). The estimation result
is displayed in Table 5. Additionally, Table 6 presents the decomposition of
the direct, indirect, and total effects of the SGN for the models.
The coefficients in section 1 of Table 5 correspond to the effect of
the variables on the same type of crime within the same municipality;
consequently, section 2 of Table 5 presents the spatial effects of the variables
in the nearest municipalities. The significance of most of the coefficients
associated with the spatial effects of the variables confirms the relevance
of spatial effects in the analysis of crime and displacement. Furthermore,
economic theory posits that this spatial coefficient presents the correct sign:
a percentage of males between 19 and 34 causes an increase in crime, and a
probability of apprehension is a deterrent to crime (Ehrlich, 1973). According
to Anselin et al. (2000), criminals may commit illegal actions in transit areas
near their residences; this supports the significance in most spatial coefficients
but less so in local coefficients.
Additionally, the spatial effects of forcibly displaced persons yield a
positive sign for homicide rates and a negative sign for personal injuries and
residential burglary. These findings explain that forcibly displaced persons, in
some cases, increase social control for certain crimes while simultaneously
increasing the occurrence of others (Bucheli et al.,2019). Similar effects
of immigrants’ concentration on crime rates were observed in Los Angeles
neighborhoods (Macdonald et al., 2013) and neighborhoods in England (Bell
& Machin, 2013).
Consequently, the results of spatial effects on the reception of forcibly
displaced persons suggest that only personal injury cases have direct and
significant effects; that is, a 1% increase in the number of forcibly displaced
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
Table 5. SGN model estimations
Variable Homicide
rate model
Kidnapping
rate model
Personal
injuries rate
model
Residential
burglary rate
model
Automobile theft
rate model
Reception FDP
rate
0.009 -0.009 -0.013 .02403* -0.005
Population -1.501 0.379 -3.497*** -2.620** -1.507
Percentage
Male 19-34
-0.213*** 0.076 -0.065 -0.159** -0.030
Probability of
apprehension
(-2)
-0.001*** -0.00003** 0.000 0.000 -0.0014*
Reception FDP
rate
0.059*** 0.020 -0.040** -0.0601*** -0.003
Dependent
variable
4.724*** -3.097*** 3.531*** 3.487** 4.600***
Spatial error
term
0.038 -.411*** 0.266*** 0.236*** 0.007
Population 0.002*** 0.000 -0.001 -0.001 0.000
Percentage
Male 19-34
.319*** .475*** .688*** 0.718*** 0.494***
Probability of
apprehension
(-2)
-0.253*** -0.425*** -0.583*** -0.648*** -0.392***
Constant 3.928*** 3.168*** 3.228*** 4.287*** 3.755***
Statistics
AIC 67576.001 62919.690 64402.695 71542.683 67013.236
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
persons causes a nearly 2% decline in personal injury cases in the same
municipality of reception. Furthermore, the indirect effect of the reception of
forcibly displaced persons implies the existence of significant spatial spillovers.
It suggests that a 1% increase in neighboring municipalities has a positive
157
Table 5. SGN model estimations
Variable Homicide
rate model
Kidnapping
rate model
Personal
injuries rate
model
Residential
burglary rate
model
Automobile theft
rate model
Reception FDP
rate
0.009 -0.009 -0.013 .02403* -0.005
Population -1.501 0.379 -3.497*** -2.620** -1.507
Percentage
Male 19-34
-0.213*** 0.076 -0.065 -0.159** -0.030
Probability of
apprehension
(-2)
-0.001*** -0.00003** 0.000 0.000 -0.0014*
Reception FDP
rate
0.059*** 0.020 -0.040** -0.0601*** -0.003
Dependent
variable
4.724*** -3.097*** 3.531*** 3.487** 4.600***
Spatial error
term
0.038 -.411*** 0.266*** 0.236*** 0.007
Population 0.002*** 0.000 -0.001 -0.001 0.000
Percentage
Male 19-34
.319*** .475*** .688*** 0.718*** 0.494***
Probability of
apprehension
(-2)
-0.253*** -0.425*** -0.583*** -0.648*** -0.392***
Constant 3.928*** 3.168*** 3.228*** 4.287*** 3.755***
Statistics
AIC 67576.001 62919.690 64402.695 71542.683 67013.236
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
persons causes a nearly 2% decline in personal injury cases in the same
municipality of reception. Furthermore, the indirect effect of the reception of
forcibly displaced persons implies the existence of significant spatial spillovers.
It suggests that a 1% increase in neighboring municipalities has a positive
157
158
impact of approximately 8.7% on homicide rates and a negative impact of
nearly 14.8% on residential burglary and 14.2% on personal injuries.
Conclusions
This paper presents arguments for the relationship between forced
displacement and crime rates in the receiving area. Economic theory
suggests that individuals with adverse socioeconomic conditions are prone to
committing crimes. Indeed, all types of forced migration alter the conditions
of municipalities and affect crime rates due to the lack of opportunities
and resources in every location. As individuals accustomed to a particular
lifestyle arrive in the reception areas, they start to understand that they can
no longer support themselves. The prospect of a new location may scare
forced migrants, who have also lost their possessions and need fast incomes.
Sadly, the easiest way to attain these is through illegal methods; therefore,
displacement increases crime.
The findings suggest the presence of a correlation among three of the
crime types examined. The estimation results for homicide, personal injuries,
and residential burglary rates indicate a significant relationship with the
allocation of forcibly displaced persons in densely populated areas. In the case
of homicide rates, a positive relationship between the reception of forcibly
displaced persons in nearby areas and an increase in local homicides can be
observed. These patterns suggest the mobility of criminals to nearby places
to commit these types of crimes. The effect of the reception of forcibly
displaced persons on the rate of personal injuries presents a contrary sign in
the results; this indicates an increase in social control in adjacent areas due to
the reception of displaced persons, thereby reducing the incidence of personal
injuries among the population in local areas. As for residential burglary,
an increase in forced migrants in nearby places contributes to reducing
this type of crime. Conversely, automobile theft and kidnapping rates are
unrelated to the reception of forcibly displaced persons in main and adjacent
areas. Furthermore, the results exhibit problems of endogeneity when the
contemporary probability of apprehension is used. Using this variable with a
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
impact of approximately 8.7% on homicide rates and a negative impact of
nearly 14.8% on residential burglary and 14.2% on personal injuries.
Conclusions
This paper presents arguments for the relationship between forced
displacement and crime rates in the receiving area. Economic theory
suggests that individuals with adverse socioeconomic conditions are prone to
committing crimes. Indeed, all types of forced migration alter the conditions
of municipalities and affect crime rates due to the lack of opportunities
and resources in every location. As individuals accustomed to a particular
lifestyle arrive in the reception areas, they start to understand that they can
no longer support themselves. The prospect of a new location may scare
forced migrants, who have also lost their possessions and need fast incomes.
Sadly, the easiest way to attain these is through illegal methods; therefore,
displacement increases crime.
The findings suggest the presence of a correlation among three of the
crime types examined. The estimation results for homicide, personal injuries,
and residential burglary rates indicate a significant relationship with the
allocation of forcibly displaced persons in densely populated areas. In the case
of homicide rates, a positive relationship between the reception of forcibly
displaced persons in nearby areas and an increase in local homicides can be
observed. These patterns suggest the mobility of criminals to nearby places
to commit these types of crimes. The effect of the reception of forcibly
displaced persons on the rate of personal injuries presents a contrary sign in
the results; this indicates an increase in social control in adjacent areas due to
the reception of displaced persons, thereby reducing the incidence of personal
injuries among the population in local areas. As for residential burglary,
an increase in forced migrants in nearby places contributes to reducing
this type of crime. Conversely, automobile theft and kidnapping rates are
unrelated to the reception of forcibly displaced persons in main and adjacent
areas. Furthermore, the results exhibit problems of endogeneity when the
contemporary probability of apprehension is used. Using this variable with a
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
lag of two periods is the optimal solution since the results are consistent with
the economic theory of crime.
Moreover, the evidence found in this study is consistent with previous
studies conducted for developing countries such as South Africa and
Brazil. In the case of South Africa, evidence supports a positive and
significant effect of internal migration on serious crimes, including homicides
(Kollamparambil, 2019). These effects are similar to those observed in Brazil,
where a 10% increase in internal migrants generated a 9.4% increase in the
homicide rate in a Brazilian microregions during the period from 2005 to 2010
(Egger, 2022).
Our estimations indicate that each crime category has distinct determi-
nants and dynamics; this confirms the differences in the spatial distribution
of crime and the reception of forcibly displaced persons across the Colombian
territory. Due to these differences, the consolidation of a composite crime
index is unsuitable. Consequently, the spatial allocation of the population het-
erogeneity variables is relevant for explaining crime levels; this demonstrates
that previous estimates present bias error since they do not consider the spa-
tial patterns of the variables.
In light of these expected results, the government’s effort should be
directed to generate opportunities for this vulnerable population who are
victims of both internal conflict in their regions of origin and the limited
job opportunities in the regions where they settle. It is then necessary to
encourage these families to train their members according to the demands
of the productive sector and to promote labor-intensive and ideally qualified
activities, which will generate higher economic returns for these households.
Therefore, social and economic programs to alleviate the adverse conditions
of the vulnerable population must be implemented through plans that impact
the local population of the municipalities and neighboring areas.
159
lag of two periods is the optimal solution since the results are consistent with
the economic theory of crime.
Moreover, the evidence found in this study is consistent with previous
studies conducted for developing countries such as South Africa and
Brazil. In the case of South Africa, evidence supports a positive and
significant effect of internal migration on serious crimes, including homicides
(Kollamparambil, 2019). These effects are similar to those observed in Brazil,
where a 10% increase in internal migrants generated a 9.4% increase in the
homicide rate in a Brazilian microregions during the period from 2005 to 2010
(Egger, 2022).
Our estimations indicate that each crime category has distinct determi-
nants and dynamics; this confirms the differences in the spatial distribution
of crime and the reception of forcibly displaced persons across the Colombian
territory. Due to these differences, the consolidation of a composite crime
index is unsuitable. Consequently, the spatial allocation of the population het-
erogeneity variables is relevant for explaining crime levels; this demonstrates
that previous estimates present bias error since they do not consider the spa-
tial patterns of the variables.
In light of these expected results, the government’s effort should be
directed to generate opportunities for this vulnerable population who are
victims of both internal conflict in their regions of origin and the limited
job opportunities in the regions where they settle. It is then necessary to
encourage these families to train their members according to the demands
of the productive sector and to promote labor-intensive and ideally qualified
activities, which will generate higher economic returns for these households.
Therefore, social and economic programs to alleviate the adverse conditions
of the vulnerable population must be implemented through plans that impact
the local population of the municipalities and neighboring areas.
159
160
Appendix
Table A1. Fixed effect without lag of probability of apprehension
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP rate 0.046*** 0.027*** -0.054*** -0.013 0.00004
Population 2.335*** -3.979*** -0.051 2.645*** 0.415
Percentage Male
19-34
-0.300*** -0.610*** 0.328*** 0.1434*** -0.081**
Probability of ap-
prehension
0.009*** 0.0003*** 0.001*** 0.006*** 0.0129***
Constant -12.520** 50.437*** -7.260 -31.934*** -6.831
Observations 15344 15344 15344 15344 15344
Significance: * p < .1; **p < .05; *** p < .01
Source: Own elaboration.
Table A2. Fixed effect with lag of probability of apprehension
Variable Homicide
rate
Kidnapping rate Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP
rate
0.029*** 0.014 -.054*** -0.014 -0.003
Population 1.721*** -3.713*** -0.167 1.967** 1.134*
Percentage
Male 19-34
-0.269*** -0.520*** 0.387*** 0.156*** -0.019
Probability of
apprehension
(-1)
-0.001*** 0.000 -0.001*** -0.002*** -0.0012*
Constant -7.238 44.947864*** -8.021 -25.746*** -15.727***
Observations 15344 15344 15344 15344 15344
Significance: * p < .1; **p < .05; *** p < .01
Source: Own elaboration.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Appendix
Table A1. Fixed effect without lag of probability of apprehension
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP rate 0.046*** 0.027*** -0.054*** -0.013 0.00004
Population 2.335*** -3.979*** -0.051 2.645*** 0.415
Percentage Male
19-34
-0.300*** -0.610*** 0.328*** 0.1434*** -0.081**
Probability of ap-
prehension
0.009*** 0.0003*** 0.001*** 0.006*** 0.0129***
Constant -12.520** 50.437*** -7.260 -31.934*** -6.831
Observations 15344 15344 15344 15344 15344
Significance: * p < .1; **p < .05; *** p < .01
Source: Own elaboration.
Table A2. Fixed effect with lag of probability of apprehension
Variable Homicide
rate
Kidnapping rate Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP
rate
0.029*** 0.014 -.054*** -0.014 -0.003
Population 1.721*** -3.713*** -0.167 1.967** 1.134*
Percentage
Male 19-34
-0.269*** -0.520*** 0.387*** 0.156*** -0.019
Probability of
apprehension
(-1)
-0.001*** 0.000 -0.001*** -0.002*** -0.0012*
Constant -7.238 44.947864*** -8.021 -25.746*** -15.727***
Observations 15344 15344 15344 15344 15344
Significance: * p < .1; **p < .05; *** p < .01
Source: Own elaboration.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
Table A3. SAR Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP
rate
0.018* 0.006 -.041*** -0.002 -0.007
Population 0.763 -2.862*** -0.807 0.874 1.381*
Percentage Male
19-34
-0.215*** -0.330*** 0.303*** 0.128*** 0.025
Probability of ap-
prehension (-2)
-0.0011** -0.00003** -0.0006* -0.0008 -0.001**
Spatial
Dependent vari-
able
0.102*** 0.158*** .337*** .332*** .191***
Constant 3.987*** 3.298*** 3.495*** 4.705*** 3.889***
Statistics
AIC 67600.092 63051.669 64655.179 71813.779 67051.579
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
161
Table A3. SAR Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP
rate
0.018* 0.006 -.041*** -0.002 -0.007
Population 0.763 -2.862*** -0.807 0.874 1.381*
Percentage Male
19-34
-0.215*** -0.330*** 0.303*** 0.128*** 0.025
Probability of ap-
prehension (-2)
-0.0011** -0.00003** -0.0006* -0.0008 -0.001**
Spatial
Dependent vari-
able
0.102*** 0.158*** .337*** .332*** .191***
Constant 3.987*** 3.298*** 3.495*** 4.705*** 3.889***
Statistics
AIC 67600.092 63051.669 64655.179 71813.779 67051.579
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
161
162
Table A4. SLX Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP rate 0.014 -0.003 -0.033*** 0.004 -0.004
Population -1.061 -0.168 -3.244*** -1.573 -0.601
Percentage Male 19-
34
-0.223*** -0.002 0.019 -0.115 -0.021
Probability of ap-
prehension (-2)
-0.001*** -0.00003** -0.00006 -0.0007 -0.001**
Spatial
Reception FDP rate 0.074*** 0.027 -0.114*** -0.109*** -0.008
Population 5.363*** -4.969*** 3.611*** 5.398*** 5.871***
Percentage Male 19-
34
-0.030 -0.613*** 0.595*** 0.385*** -0.007
Probability of ap-
prehension (-2)
0.002** -0.00001 -0.002*** -0.002 -0.001
Constant 3.993*** 3.304*** 3.605*** 4.873*** 3.927***
Statistics
AIC 67620.626 63051.607 65152.215 72421.595 67219.647
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; *** p < .01
Source: Own elaboration.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Table A4. SLX Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP rate 0.014 -0.003 -0.033*** 0.004 -0.004
Population -1.061 -0.168 -3.244*** -1.573 -0.601
Percentage Male 19-
34
-0.223*** -0.002 0.019 -0.115 -0.021
Probability of ap-
prehension (-2)
-0.001*** -0.00003** -0.00006 -0.0007 -0.001**
Spatial
Reception FDP rate 0.074*** 0.027 -0.114*** -0.109*** -0.008
Population 5.363*** -4.969*** 3.611*** 5.398*** 5.871***
Percentage Male 19-
34
-0.030 -0.613*** 0.595*** 0.385*** -0.007
Probability of ap-
prehension (-2)
0.002** -0.00001 -0.002*** -0.002 -0.001
Constant 3.993*** 3.304*** 3.605*** 4.873*** 3.927***
Statistics
AIC 67620.626 63051.607 65152.215 72421.595 67219.647
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; *** p < .01
Source: Own elaboration.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
Table A5. SEM Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP
rate
0.017 0.006 -0.036*** 0.003 -0.007
Population 0.582 -2.983*** -1.105 0.628 1.182
Percentage Male
19-34
-0.227*** -0.341*** 0.304*** 0.108* 0.0323
Probability of ap-
prehension (-2)
-0.001*** -.00003** -0.0005 -0.0007 -0.001**
Spatial
Spatial error term 0.102*** 0.148*** 0.336*** 0.332*** 0.190***
Constant 3.987*** 3.301*** 3.501*** 4.706*** 3.889***
Statistics
AIC 67600.47 63072.694 64691.959 71819.989 67053.566
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
163
Table A5. SEM Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP
rate
0.017 0.006 -0.036*** 0.003 -0.007
Population 0.582 -2.983*** -1.105 0.628 1.182
Percentage Male
19-34
-0.227*** -0.341*** 0.304*** 0.108* 0.0323
Probability of ap-
prehension (-2)
-0.001*** -.00003** -0.0005 -0.0007 -0.001**
Spatial
Spatial error term 0.102*** 0.148*** 0.336*** 0.332*** 0.190***
Constant 3.987*** 3.301*** 3.501*** 4.706*** 3.889***
Statistics
AIC 67600.47 63072.694 64691.959 71819.989 67053.566
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
163
164
Table A6. SAC Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP
rate
0.019* 0.003 -0.031*** -0.002 -0.006
Population 1.004 -2.017*** -0.728 0.140 1.238**
Percentage
Male 19-34
-0.179*** -0.234*** 0.177*** 0.074** 0.009
Probability of
apprehension
(-2)
-0.0009** -0.00003** -0.0005* -0.0007 -0.001**
Spatial
Dependent
variable
0.325*** 0.518*** 0.703*** 0.722*** 0.497***
Spatial error
term
-0.257*** -0.476*** -0.602*** -0.650*** -0.395***
Constant 3.930*** 3.150*** 3.219*** 4.288*** 3.755***
Statistics
AIC 67596.491 62971.356 64440.093 71562.201 67019.308
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Table A6. SAC Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP
rate
0.019* 0.003 -0.031*** -0.002 -0.006
Population 1.004 -2.017*** -0.728 0.140 1.238**
Percentage
Male 19-34
-0.179*** -0.234*** 0.177*** 0.074** 0.009
Probability of
apprehension
(-2)
-0.0009** -0.00003** -0.0005* -0.0007 -0.001**
Spatial
Dependent
variable
0.325*** 0.518*** 0.703*** 0.722*** 0.497***
Spatial error
term
-0.257*** -0.476*** -0.602*** -0.650*** -0.395***
Constant 3.930*** 3.150*** 3.219*** 4.288*** 3.755***
Statistics
AIC 67596.491 62971.356 64440.093 71562.201 67019.308
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
Table A7. SDM Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP rate 0.013 -0.004 -0.027*** 0.010 -0.005
Population -1.191 -0.137 -3.278*** -1.332 -0.716
Percentage Male 19-
34
-0.222*** 0.006 0.00004 -0.119* -0.018
Probability of ap-
prehension (-2)
-0.001*** -.00003** -0.00001 -0.0006 -0.001**
Spatial
Reception FDP rate 0.066*** 0.022 -0.071*** -0.077*** -0.006
Dependent variable 5.060*** -4.225*** 3.638*** 3.446** 5.014***
Population -0.005 -0.535*** 0.415*** 0.316*** -0.004
Percentage Male 19-
34
0.002** -0.000009 -0.001** -0.001 -0.0005
Probability of ap-
prehension (-2)
0.097*** 0.136*** 0.321*** 0.326*** 0.188***
Constant 3.983*** 3.286*** 3.486*** 4.700*** 3.887***
Statistics
AIC 67580.286 62961.433 64577.127 71787.311 67019.308
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
165
Table A7. SDM Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal
injuries rate
Residential
burglary
Automobile
theft rate
Reception FDP rate 0.013 -0.004 -0.027*** 0.010 -0.005
Population -1.191 -0.137 -3.278*** -1.332 -0.716
Percentage Male 19-
34
-0.222*** 0.006 0.00004 -0.119* -0.018
Probability of ap-
prehension (-2)
-0.001*** -.00003** -0.00001 -0.0006 -0.001**
Spatial
Reception FDP rate 0.066*** 0.022 -0.071*** -0.077*** -0.006
Dependent variable 5.060*** -4.225*** 3.638*** 3.446** 5.014***
Population -0.005 -0.535*** 0.415*** 0.316*** -0.004
Percentage Male 19-
34
0.002** -0.000009 -0.001** -0.001 -0.0005
Probability of ap-
prehension (-2)
0.097*** 0.136*** 0.321*** 0.326*** 0.188***
Constant 3.983*** 3.286*** 3.486*** 4.700*** 3.887***
Statistics
AIC 67580.286 62961.433 64577.127 71787.311 67019.308
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
165
166
Table A8. SDEM Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal injuries rate Residential
burglary
Automobile
theft rate
Reception FDP rate 0.015 -0.003 -0.033*** 0.004 -0.005
Population -1.065 -0.329 -2.945*** -0.799 -0.362
Percentage Male 19-
34
-0.225*** -0.020 0.057 -0.075 -0.014
Probability of ap-
prehension (-2)
-.001*** -.00003** -0.0001 -0.0007 -0.001**
Spatial
Reception FDP rate 0.068*** 0.022 -0.081*** -0.076** -0.009
Spatial error term 0.096*** 0.132*** 0.319*** 0.326*** 0.187***
Population 5.168*** -4.683*** 3.570*** 3.002 5.091***
Percentage Male 19-
34
-0.025 -0.577*** 0.528*** 0.380*** -0.003
Probability of ap-
prehension (-2)
0.002** -0.00001 -0.002** -0.001 -0.0007
Constant 3.983*** 3.287*** 3.489*** 4.702*** 3.887***
Statistics
AIC 67581.701 62968.34 64594.445 71795.462 67047.534
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
Ethics Statement
This research article did not work with a person or groups of persons for
the generation of data used in the methodology, therefore it did not require
the endorsement of an Ethics Committee for its realization.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Table A8. SDEM Models estimation
Variable Homicide
rate
Kidnapping
rate
Personal injuries rate Residential
burglary
Automobile
theft rate
Reception FDP rate 0.015 -0.003 -0.033*** 0.004 -0.005
Population -1.065 -0.329 -2.945*** -0.799 -0.362
Percentage Male 19-
34
-0.225*** -0.020 0.057 -0.075 -0.014
Probability of ap-
prehension (-2)
-.001*** -.00003** -0.0001 -0.0007 -0.001**
Spatial
Reception FDP rate 0.068*** 0.022 -0.081*** -0.076** -0.009
Spatial error term 0.096*** 0.132*** 0.319*** 0.326*** 0.187***
Population 5.168*** -4.683*** 3.570*** 3.002 5.091***
Percentage Male 19-
34
-0.025 -0.577*** 0.528*** 0.380*** -0.003
Probability of ap-
prehension (-2)
0.002** -0.00001 -0.002** -0.001 -0.0007
Constant 3.983*** 3.287*** 3.489*** 4.702*** 3.887***
Statistics
AIC 67581.701 62968.34 64594.445 71795.462 67047.534
Observations 13152 13152 13152 13152 13152
Significance: *p < .1; **p < .05; ***p < .01
Source: Own elaboration.
Ethics Statement
This research article did not work with a person or groups of persons for
the generation of data used in the methodology, therefore it did not require
the endorsement of an Ethics Committee for its realization.
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
References
Acevedo, K., & Bornacelly Olivella, I. D. (2014). Panel municipal del CEDE.
Almeida, E., Haddad, E., & Hewings, G. (2003). The Spatial Pattern of
Crime in Minas Gerais: An Exploratory Analysis. Economia Aplicada,
9(1), 39-55. https://www.revistas.usp.br/ecoa/article/view/221387
Andresen, M. A. (2013). International immigration, internal migration,
and homicide in Canadian provinces. International Journal of Offender
Therapy and Comparative Criminology, 57 (5), 632-657. https://doi.or
g/10.1177/0306624X12436798
Anselin, L., Cohen, J., Cook, D., Gorr, W., & Tita, G. (2000). Spatial
Analyses of Crime. In D. Duffe (ed.), Measurement and Analysis of Crime
and Justice, Volume 4 (pp. 213-262). US Department of Justice, Office
of Justice Programs and National Institute of Justice.
Arnio, A. N., & Baumer, E. P. (2012). Demography, Foreclosure, and
Crime: Assessing Spatial Heterogeneity in Contemporary Models of
Neighborhood Crime Rates. Demographic Research, 26, 449-488. https:
//doi.org/10.4054/DemRes.2012.26.18
Becker, G. S. (1968). Crime and Punishment: An Economic Approach.
Journal of Political Economy, 76 (2), 169-217. https://www.jstor.org/
stable/1830482
Bell, B., Fasani, F., & Machin, S. (2013). Crime and Immigration: Evidence
from Large Immigrant Waves. Review of Economics and Statistics,
95(4), 1278-1290. https://doi.og/10.1162/REST_a_00337
Bell, B., & Machin, S. (2013). Immigrant Enclaves and Crime. Journal of
Regional Science, 53(1), 118-141. https://onlinelibrary.wiley.com/doi/
abs/10.1111/jors.12003
Bianchi, M., Buonanno, P., & Pinotti, P. (2012). Do Immigrants Cause
Crime? Journal of the European Economic Association, 10(6), 1318-1347.
https://doi.org/10.1111/j.1542-4774.2012.01085.x
167
References
Acevedo, K., & Bornacelly Olivella, I. D. (2014). Panel municipal del CEDE.
Almeida, E., Haddad, E., & Hewings, G. (2003). The Spatial Pattern of
Crime in Minas Gerais: An Exploratory Analysis. Economia Aplicada,
9(1), 39-55. https://www.revistas.usp.br/ecoa/article/view/221387
Andresen, M. A. (2013). International immigration, internal migration,
and homicide in Canadian provinces. International Journal of Offender
Therapy and Comparative Criminology, 57 (5), 632-657. https://doi.or
g/10.1177/0306624X12436798
Anselin, L., Cohen, J., Cook, D., Gorr, W., & Tita, G. (2000). Spatial
Analyses of Crime. In D. Duffe (ed.), Measurement and Analysis of Crime
and Justice, Volume 4 (pp. 213-262). US Department of Justice, Office
of Justice Programs and National Institute of Justice.
Arnio, A. N., & Baumer, E. P. (2012). Demography, Foreclosure, and
Crime: Assessing Spatial Heterogeneity in Contemporary Models of
Neighborhood Crime Rates. Demographic Research, 26, 449-488. https:
//doi.org/10.4054/DemRes.2012.26.18
Becker, G. S. (1968). Crime and Punishment: An Economic Approach.
Journal of Political Economy, 76 (2), 169-217. https://www.jstor.org/
stable/1830482
Bell, B., Fasani, F., & Machin, S. (2013). Crime and Immigration: Evidence
from Large Immigrant Waves. Review of Economics and Statistics,
95(4), 1278-1290. https://doi.og/10.1162/REST_a_00337
Bell, B., & Machin, S. (2013). Immigrant Enclaves and Crime. Journal of
Regional Science, 53(1), 118-141. https://onlinelibrary.wiley.com/doi/
abs/10.1111/jors.12003
Bianchi, M., Buonanno, P., & Pinotti, P. (2012). Do Immigrants Cause
Crime? Journal of the European Economic Association, 10(6), 1318-1347.
https://doi.org/10.1111/j.1542-4774.2012.01085.x
167
168
Bourguignon, F., Nu nez, J., Sanchez, F., May, A., Bourguignon, F., &
Sanchez, F. (2017). A Structural Model of Crime and Inequality in
Colombia. Journal of the European Economic Association, 1(2), 440-449.
https://www.jstor.org/stable/40005193
Bucheli, J., Fontenla, M., & Wadedll, B. (2019). Return Migration and
Violence. World Development, 116, 113-124. https://doi.org/10.101
6/j.worlddev.2018.12.010
Calderón-Mejía, V., & Ibáñez, A. M. (2016). Labour Market Effects of
Migration-Related Supply Shocks: Evidence from Internal Refugees
in Colombia. Journal of Economic Geography, 16 (3), 695-713. https:
//doi.org/10.1093/jeg/lbv030
Cheng, J., Liu, J., & Wang, J. (2017). Domestic Migration, Home Rentals,
and Crime Rates in China. The Journal of Chinese Sociology, 4(1), 8.
https://doi.org/10.1186/s40711-017-0056-3
Chiswick, B., & Miller, P. (2014). Handbook of the Economics of International
Migration. North Holland.
Cracolici, M. F., & Uberti, T. E. (2009). Geographical Distribution of Crime
in Italian Provinces: A Spatial Econometric Analysis. Jahrbuch Fur
Regionalwissenschaft, 29(1), 1-28. https://doi.org/10.1007/s10037
-008-0031-1
Departamento Administrativo Nacional de Estadística (DANE). (n.d.). Es-
tadísticas por tema: demografía y población. Accessed on 15 November
2022. https://www.dane.gov.co/index.php/estadisticas-por-tema/de
mografia-y-poblacion
Egger, E. M. (2022). Internal Migration and Crime in Brazil. Economic
Development and Cultural Change, 71(1), 223-259. https://doi.org/
10.1086/714740
Ehrlich, I. (1973). Participation in Illegitimate Activities: A Theoretical and
Empirical Investigation. Journal of Political Economy, 81(3). https:
//doi.org/10.1086/260058
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Bourguignon, F., Nu nez, J., Sanchez, F., May, A., Bourguignon, F., &
Sanchez, F. (2017). A Structural Model of Crime and Inequality in
Colombia. Journal of the European Economic Association, 1(2), 440-449.
https://www.jstor.org/stable/40005193
Bucheli, J., Fontenla, M., & Wadedll, B. (2019). Return Migration and
Violence. World Development, 116, 113-124. https://doi.org/10.101
6/j.worlddev.2018.12.010
Calderón-Mejía, V., & Ibáñez, A. M. (2016). Labour Market Effects of
Migration-Related Supply Shocks: Evidence from Internal Refugees
in Colombia. Journal of Economic Geography, 16 (3), 695-713. https:
//doi.org/10.1093/jeg/lbv030
Cheng, J., Liu, J., & Wang, J. (2017). Domestic Migration, Home Rentals,
and Crime Rates in China. The Journal of Chinese Sociology, 4(1), 8.
https://doi.org/10.1186/s40711-017-0056-3
Chiswick, B., & Miller, P. (2014). Handbook of the Economics of International
Migration. North Holland.
Cracolici, M. F., & Uberti, T. E. (2009). Geographical Distribution of Crime
in Italian Provinces: A Spatial Econometric Analysis. Jahrbuch Fur
Regionalwissenschaft, 29(1), 1-28. https://doi.org/10.1007/s10037
-008-0031-1
Departamento Administrativo Nacional de Estadística (DANE). (n.d.). Es-
tadísticas por tema: demografía y población. Accessed on 15 November
2022. https://www.dane.gov.co/index.php/estadisticas-por-tema/de
mografia-y-poblacion
Egger, E. M. (2022). Internal Migration and Crime in Brazil. Economic
Development and Cultural Change, 71(1), 223-259. https://doi.org/
10.1086/714740
Ehrlich, I. (1973). Participation in Illegitimate Activities: A Theoretical and
Empirical Investigation. Journal of Political Economy, 81(3). https:
//doi.org/10.1086/260058
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
Elhorst, J. P. (2014). Spatial Econometrics from Cross-Sectional Data to Spatial
Panels. Springer. https://doi.org/10.1007/978-3-642-40340-8
Engel, S., & Ibáñez, A. M. (2007). Displacement Due to Violence in
Colombia: A Household-Level Analysis. Economic Development and
Cultural Change, 55(2), 335-365. https://econpapers.repec.org/art
icle/ucpecdecc/y_3a2007_3av_3a55_3ai_3a2_3ap_3a335-65.htm
Fernandez, M., Ibáñez, A. M., & Pe na, X. (2014). Adjusting the Labour
Supply to Mitigate Violent Shocks: Evidence from Rural Colombia.
The Journal of Development Studies, 50(8), 1135-1155. https://doi.org/
10.1080/00220388.2014.919384
Graif, C., & Sampson, R. J. (2009). Spatial Heterogeneity in the Effects
of Immigration and Diversity on Neighborhood Homicide Rates.
Homicide Studies, 13(3), 2-20. https://doi.org/10.1177/10887679
09336728
Herrick, B. H. (1966). Urban Migration and Economic Development in Chile.
MIT Press Books.
Hirschi, T., & Stark, R. (1969). Hellfire and Delinquency. Social Problems,
17 (2), 202-213. https://doi.org/10.2307/799866
Ibáñez, A. M. (2009). Forced Displacement in Colombia: Magnitude and
Causes. The Economics of Peace and Security Journal, 4(1), 48-54. https:
//ideas.repec.org/a/epc/journl/v4y2009i1p48-54.html
Ibáñez, A. M., & Moya, A. (2010). Vulnerability of Victims of Civil Conflicts:
Empirical Evidence for the Displaced Population in Colombia. World
Development, 38(4), 647-663. https://doi.org/10.1016/j.worlddev.200
9.11.015
Ibáñez, A. M., Rodríguez, C., & Zarruk, D. (2013). Crime, Punishment,
and Schooling Decisions: Evidence from Colombian Adolescents [IDB
working paper No. IDB-WP-413]. Inter-American Development Bank.
https://publications.iadb.org/en/publication/11290/crime-punishmen
t-and-schooling-decisions-evidence-colombian-adolescents
169
Elhorst, J. P. (2014). Spatial Econometrics from Cross-Sectional Data to Spatial
Panels. Springer. https://doi.org/10.1007/978-3-642-40340-8
Engel, S., & Ibáñez, A. M. (2007). Displacement Due to Violence in
Colombia: A Household-Level Analysis. Economic Development and
Cultural Change, 55(2), 335-365. https://econpapers.repec.org/art
icle/ucpecdecc/y_3a2007_3av_3a55_3ai_3a2_3ap_3a335-65.htm
Fernandez, M., Ibáñez, A. M., & Pe na, X. (2014). Adjusting the Labour
Supply to Mitigate Violent Shocks: Evidence from Rural Colombia.
The Journal of Development Studies, 50(8), 1135-1155. https://doi.org/
10.1080/00220388.2014.919384
Graif, C., & Sampson, R. J. (2009). Spatial Heterogeneity in the Effects
of Immigration and Diversity on Neighborhood Homicide Rates.
Homicide Studies, 13(3), 2-20. https://doi.org/10.1177/10887679
09336728
Herrick, B. H. (1966). Urban Migration and Economic Development in Chile.
MIT Press Books.
Hirschi, T., & Stark, R. (1969). Hellfire and Delinquency. Social Problems,
17 (2), 202-213. https://doi.org/10.2307/799866
Ibáñez, A. M. (2009). Forced Displacement in Colombia: Magnitude and
Causes. The Economics of Peace and Security Journal, 4(1), 48-54. https:
//ideas.repec.org/a/epc/journl/v4y2009i1p48-54.html
Ibáñez, A. M., & Moya, A. (2010). Vulnerability of Victims of Civil Conflicts:
Empirical Evidence for the Displaced Population in Colombia. World
Development, 38(4), 647-663. https://doi.org/10.1016/j.worlddev.200
9.11.015
Ibáñez, A. M., Rodríguez, C., & Zarruk, D. (2013). Crime, Punishment,
and Schooling Decisions: Evidence from Colombian Adolescents [IDB
working paper No. IDB-WP-413]. Inter-American Development Bank.
https://publications.iadb.org/en/publication/11290/crime-punishmen
t-and-schooling-decisions-evidence-colombian-adolescents
169
170
Ibáñez, A. M., & Vélez, C. E. (2008). Civil Conflict and Forced Migration:
The Micro Determinants and Welfare Losses of Displacement in
Colombia. World Development, 36 (4), 659-676. https://doi.org/10
.1016/j.worlddev.2007.04.013
Ibáñez, A., & Moya, A. (2006). The Impact of Intra-State Conflict on
Economic Welfare and Consumption Smoothing: Empirical Evidence for
The Displaced Population in Colombia [HiCN working papers 23]. https:
//ideas.repec.org/p/hic/wpaper/23.html
Ibáñez, A., & Velásquez, A. (2009). Identifying Victims of Civil Conflicts:
An Evaluation of Forced Displaced Households in Colombia. Journal
of Peace Research, 46 (3), 431-451. https://www.jstor.org/stable/25654
414
Ibáñez, A., & Vélez, C. (2008). Civil Conflict and Forced Migration: The
Micro Determinants and Welfare Losses of Displacement in Colombia.
World Development, 36 (4), 659-676. https://doi.org/10.1016/j.worldd
ev.2007.04.013
InSight Crime (2021, January) Colombia Profile. https://insightcrime.org/c
olombia-organized-crime-news/colombia/
Jaitman, L., & Machin, S. (2013). Crime and immigration: new evidence
from England and Wales. IZA Journal of Migration, 2(1). https://doi.or
g/10.1186/2193-9039-2-19
Kollamparambil, U. (2019). Immigration, Internal Migration and Crime in
South Africa: A Multi-Level Model Analysis. Development Policy, 37
(5), 672- 691. https://doi.org/10.1111/dpr.12382
Kubrin, C. E., Hipp, J. R., & Kim, Y. A. (2016). Different than the Sum
of Its Parts: Examining the Unique Impacts of Immigrant Groups on
Neighborhood Crime Rates. Journal of Quantitative Criminology, 34,
1-36. https://doi.org/10.1007/s10940-016-9320-y
Macdonald, J. M., Hipp, J. R., & Gill, C. (2013). The Effects of Immigrant
Concentration on Changes in Neighborhood Crime Rates. Journal of
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Ibáñez, A. M., & Vélez, C. E. (2008). Civil Conflict and Forced Migration:
The Micro Determinants and Welfare Losses of Displacement in
Colombia. World Development, 36 (4), 659-676. https://doi.org/10
.1016/j.worlddev.2007.04.013
Ibáñez, A., & Moya, A. (2006). The Impact of Intra-State Conflict on
Economic Welfare and Consumption Smoothing: Empirical Evidence for
The Displaced Population in Colombia [HiCN working papers 23]. https:
//ideas.repec.org/p/hic/wpaper/23.html
Ibáñez, A., & Velásquez, A. (2009). Identifying Victims of Civil Conflicts:
An Evaluation of Forced Displaced Households in Colombia. Journal
of Peace Research, 46 (3), 431-451. https://www.jstor.org/stable/25654
414
Ibáñez, A., & Vélez, C. (2008). Civil Conflict and Forced Migration: The
Micro Determinants and Welfare Losses of Displacement in Colombia.
World Development, 36 (4), 659-676. https://doi.org/10.1016/j.worldd
ev.2007.04.013
InSight Crime (2021, January) Colombia Profile. https://insightcrime.org/c
olombia-organized-crime-news/colombia/
Jaitman, L., & Machin, S. (2013). Crime and immigration: new evidence
from England and Wales. IZA Journal of Migration, 2(1). https://doi.or
g/10.1186/2193-9039-2-19
Kollamparambil, U. (2019). Immigration, Internal Migration and Crime in
South Africa: A Multi-Level Model Analysis. Development Policy, 37
(5), 672- 691. https://doi.org/10.1111/dpr.12382
Kubrin, C. E., Hipp, J. R., & Kim, Y. A. (2016). Different than the Sum
of Its Parts: Examining the Unique Impacts of Immigrant Groups on
Neighborhood Crime Rates. Journal of Quantitative Criminology, 34,
1-36. https://doi.org/10.1007/s10940-016-9320-y
Macdonald, J. M., Hipp, J. R., & Gill, C. (2013). The Effects of Immigrant
Concentration on Changes in Neighborhood Crime Rates. Journal of
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Sánchez Saldarriaga, A., Gómez Toro, C., Velásquez, H. and Mejía Mejía J. F.: Effects of...
Quantitative Criminology, 29(2), 191-215. https://doi.org/10.1007/S1
0940-012-9176-8
Mears, D. P. (2001). The Immigration-Crime Nexus: Toward an Analytic
Framework for Assessing and Guiding Theory, Research, and Policy.
Sociological Perspectives, 44(1), 1-19. https://doi.org/10.1525/sop.20
01.44.1.1
Meng, X., & Zhang, D. (2013). The Social Impact of Rural-Urban Migration
on Urban ‘Natives’ [working paper]. https://citeseerx.ist.psu.edu/docu
ment?repid=rep1&type=pdf&doi=d70c5fb879d8268c2dda80a9999
7b028b96e628d
Nunziata, L. (2015). Immigration and Crime: Evidence from Victimization
Data. Journal of Population Economics, 28(3), 697-736. https://doi.org/
10.1007/s00148-015-0543-2
Ousey, G. C., & Kubrin, C. E. (2009). Exploring the Connection between
Immigration and Violent Crime Rates in U.S. Cities, 1980-2000. Social
Problems, 56 (3), 447-473. https://doi.org/10.1525/sp.2009.56.3.447
Policía Nacional de Colombia (n.d.). Estadística delictiva. Accessed on March
1 of 2018. https://www.policia.gov.co/grupo-informaci%C3%B3n-c
riminalidad/estadistica-delictiva
Reid, L. W., Weiss, H. E., Adelman, R. M., & Jaret, C. (2005). The
Immigration-Crime Relationship: Evidence across US Metropolitan
Areas. Social Science Research, 34(4), 757-780. https://doi.org/10.1
016/j.ssresearch.2005.01.001
Roche-Villarreal, L. (2012). Forced Displacement and Crime in Colombia
[Memories]. https://citeseerx.ist.psu.edu/document?repid=rep1&
type=pdf&doi=6d3e0399ec3f1ed2ea0e23fcc2b977059e1a7347
Schultz, T. P. (1971). Rural-Urban Migration in Colombia. The Review of
Economics and Statistics, 53(2), 157-163. https://www.jstor.org/stable
/1925711
171
Quantitative Criminology, 29(2), 191-215. https://doi.org/10.1007/S1
0940-012-9176-8
Mears, D. P. (2001). The Immigration-Crime Nexus: Toward an Analytic
Framework for Assessing and Guiding Theory, Research, and Policy.
Sociological Perspectives, 44(1), 1-19. https://doi.org/10.1525/sop.20
01.44.1.1
Meng, X., & Zhang, D. (2013). The Social Impact of Rural-Urban Migration
on Urban ‘Natives’ [working paper]. https://citeseerx.ist.psu.edu/docu
ment?repid=rep1&type=pdf&doi=d70c5fb879d8268c2dda80a9999
7b028b96e628d
Nunziata, L. (2015). Immigration and Crime: Evidence from Victimization
Data. Journal of Population Economics, 28(3), 697-736. https://doi.org/
10.1007/s00148-015-0543-2
Ousey, G. C., & Kubrin, C. E. (2009). Exploring the Connection between
Immigration and Violent Crime Rates in U.S. Cities, 1980-2000. Social
Problems, 56 (3), 447-473. https://doi.org/10.1525/sp.2009.56.3.447
Policía Nacional de Colombia (n.d.). Estadística delictiva. Accessed on March
1 of 2018. https://www.policia.gov.co/grupo-informaci%C3%B3n-c
riminalidad/estadistica-delictiva
Reid, L. W., Weiss, H. E., Adelman, R. M., & Jaret, C. (2005). The
Immigration-Crime Relationship: Evidence across US Metropolitan
Areas. Social Science Research, 34(4), 757-780. https://doi.org/10.1
016/j.ssresearch.2005.01.001
Roche-Villarreal, L. (2012). Forced Displacement and Crime in Colombia
[Memories]. https://citeseerx.ist.psu.edu/document?repid=rep1&
type=pdf&doi=6d3e0399ec3f1ed2ea0e23fcc2b977059e1a7347
Schultz, T. P. (1971). Rural-Urban Migration in Colombia. The Review of
Economics and Statistics, 53(2), 157-163. https://www.jstor.org/stable
/1925711
171
172
Spenkuch, J. L. (2014). Understanding the Impact of Immigration on Crime.
American Law and Economics Review, 16 (1), 177-219. https://www.js
tor.org/stable/24735701
Thomas, P. M. (2011). Theoretical Articulation on Immigration and Crime.
Homicide Studies, 15(4), 382-403. https://doi.org/10.1177/10887679
11424539
Tita, G. E., & Radil, S. M. (2010). Spatial Regression Models in Criminology:
Modeling Social Processes in the SpatialWeights Matrix. Handbook of
Quantitative Criminology. https://doi.org/10.1007/978-0-387-776
50-7
Téllez, J. (2022). Land, Opportunism, and Displacement in Civil Wars:
Evidence from Colombia.American Political Science Review, 116 (2),
403-418. https://doi.org/10.1017/S0003055421001003
Torres-Preciado, V. H., Polanco-Gaytán, M., & Tinoco-Zerme no, M. A.
(2017). Crime and Regional Economic Growth in Mexico: A Spatial
Perspective. Papers in Regional Science, 96 (3), 477-494. https://doi.or
g/10.1111/pirs.12205
Treyger, E. (2013). Migration and Violent Crime: Lessons from the Russian
Experience. Georgetown Immigration Law Journal, 27 (2), 257-310. ht
tps://doi.org/10.1146/annurev.soc.31.041304.122308
Unidad de Víctimas (n.d.). Víctimas de conflicto armado. UpToDate. Ac-
cessed on March 1, 2018. https://www.unidadvictimas.gov.co/es/p
ublicacion-de-datos-abiertos/
Wadsworth, T. (2010). Is Immigration Responsible for The Crime Drop? An
Assessment of the Influence of Immigration on Changes in Violent
Crime Between 1990 and 2000. Social Science Quarterly, 91(2), 531-
553. https://doi.org/10.1111/j.1540-6237.2010.00706.x
Ye, X., & Wu, L. (2011). Analyzing the Dynamics of Homicide Patterns
in Chicago: ESDA and Spatial Panel Approaches. Applied Geography,
31(2), 800-807. https://doi.org/10.1016/j.apgeog.2010.08.00
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024
Spenkuch, J. L. (2014). Understanding the Impact of Immigration on Crime.
American Law and Economics Review, 16 (1), 177-219. https://www.js
tor.org/stable/24735701
Thomas, P. M. (2011). Theoretical Articulation on Immigration and Crime.
Homicide Studies, 15(4), 382-403. https://doi.org/10.1177/10887679
11424539
Tita, G. E., & Radil, S. M. (2010). Spatial Regression Models in Criminology:
Modeling Social Processes in the SpatialWeights Matrix. Handbook of
Quantitative Criminology. https://doi.org/10.1007/978-0-387-776
50-7
Téllez, J. (2022). Land, Opportunism, and Displacement in Civil Wars:
Evidence from Colombia.American Political Science Review, 116 (2),
403-418. https://doi.org/10.1017/S0003055421001003
Torres-Preciado, V. H., Polanco-Gaytán, M., & Tinoco-Zerme no, M. A.
(2017). Crime and Regional Economic Growth in Mexico: A Spatial
Perspective. Papers in Regional Science, 96 (3), 477-494. https://doi.or
g/10.1111/pirs.12205
Treyger, E. (2013). Migration and Violent Crime: Lessons from the Russian
Experience. Georgetown Immigration Law Journal, 27 (2), 257-310. ht
tps://doi.org/10.1146/annurev.soc.31.041304.122308
Unidad de Víctimas (n.d.). Víctimas de conflicto armado. UpToDate. Ac-
cessed on March 1, 2018. https://www.unidadvictimas.gov.co/es/p
ublicacion-de-datos-abiertos/
Wadsworth, T. (2010). Is Immigration Responsible for The Crime Drop? An
Assessment of the Influence of Immigration on Changes in Violent
Crime Between 1990 and 2000. Social Science Quarterly, 91(2), 531-
553. https://doi.org/10.1111/j.1540-6237.2010.00706.x
Ye, X., & Wu, L. (2011). Analyzing the Dynamics of Homicide Patterns
in Chicago: ESDA and Spatial Panel Approaches. Applied Geography,
31(2), 800-807. https://doi.org/10.1016/j.apgeog.2010.08.00
Lecturas de Economía -Lect. Econ. - No. 101. Medellín, enero-junio 2024