DOES UNDOCUMENTED IMMIGRATION INCREASE VIOLENT CRIME?*

• Date: 01 May 2018
• Published date: 01 May 2018
• DOI: http://doi.org/10.1111/1745-9125.12175
• Author: TY MILLER,MICHAEL T. LIGHT

DOES UNDOCUMENTED IMMIGRATION INCREASE

VIOLENT CRIME?∗

MICHAEL T. LIGHT1and TY MILLER2

1Department of Sociology, University of Wisconsin—Madison

2Department of Sociology, Purdue University

KEYWORDS: undocumented immigration, violent crime, immigration enforcement

Despite substantial public, political, and scholarly attention to the issue of immi-

gration and crime, we know little about the criminological consequences of undocu-

mented immigration. As a result, fundamental questions about whether undocumented

immigration increases violent crime remain unanswered. In an attempt to address

this gap, we combine newly developed estimates of the unauthorized population

with multiple data sources to capture the criminal, socioeconomic, and demographic

context of all 50 states and Washington, DC, from 1990 to 2014 to provide the

first longitudinal analysis of the macro-level relationship between undocumented

immigration and violence. The results from fixed-effects regression models reveal that

undocumented immigration does not increase violence. Rather, the relationship be-

tween undocumented immigration and violent crime is generally negative, although

not significant in all specifications. Using supplemental models of victimization data

and instrumental variable methods, we find little evidence that these results are due to

decreased reporting or selective migration to avoid crime. We consider the theoretical

and policy implications of these findings against the backdrop of the dramatic increase

in immigration enforcement in recent decades.

Few topics have more criminological significance and public policy salience than un-

derstanding the impact of undocumented immigration on violent crime. Although the

immigration–crime nexus has been at the fore of criminological inquiry since the Chicago

School of the early 20th century (Shaw and McKay, 1942), this issue has taken on added

importance over the past two decades as the United States has experienced the largest

wave of immigration—in both absolute and relative terms—in its history. Nevertheless,

despite substantial research attention to the association between immigration and crime

(Martinez and Valenzuela, 2006), significant gaps remain in the literature. Most notably,

∗Additional supporting information can be found in the listing for this article in the Wiley Online

Library at http://onlinelibrary.wiley.com/doi/10.1111/crim.2018.56.issue-2/issuetoc.

A previous version of this article was presented at the 2017 American Sociological Association

annual meeting in Montreal, Quebec, Canada, and at the 2017 American Society of Criminology

annual meeting in Philadelphia, Pennsylvania. We are deeply indebted to several of our colleagues

for insightful comments on early drafts, including Ted Gerber, Chad Goldberg, Mike Massoglia,

and especially Shawn Bauldry, Jason Fletcher, and Felix Elwert for invaluable methodological

comments. We also thank the anonymous reviewers for improving our article.

Direct correspondence to Michael Light, Department of Sociology, University of Wisconsin—

Madison, 8128 William H. Sewell Social Sciences Building, 1180 Observatory Drive, Madison, WI

53703 (e-mail: mlight@ssc.wisc.edu).

C2018 American Society of Criminology doi: 10.1111/1745-9125.12175

CRIMINOLOGY Volume 56 Number 2 370–401 2018 370

UNDOCUMENTED IMMIGRATION AND VIOLENT CRIME 371

this body of work has mainly been confined to assessments of the overall or Latino

foreign-born populations (Feldmeyer, 2009; Martinez, Stowell, and Lee, 2010; Ousey and

Kubrin, 2009; Wadsworth, 2010) because of the paucity of data accounting for unautho-

rized immigrants separately.1As Ousey and Kubrin (2017: 1.9) highlighted, “the problem

with these approaches is that they treat immigrants as a homogenous population and fail

to account for significant variation across types of immigrants.” As a result, research on

undocumented immigration remains a substantial lacuna in immigration–crime research.

Indeed, in a recent meta-analysis of the 51 macro-level immigration–crime studies con-

ducted between 1994 and 2014, not one was aimed at explicitly examining unauthorized

immigration flows (Ousey and Kubrin, 2017). Since that time, we are aware of only one

study in which the association between unauthorized immigration and violence was inves-

tigated. In that study, Green (2016) found that undocumented immigration is generally

not associated with violent crime, though unauthorized immigration from Mexico may be

associated with higher rates of violence. Although informative, several limitations of this

study warrant further inquiry. Most notably, the analysis is cross-sectional, thus, limiting

both the substantive questions under consideration and the analytical leverage to answer

them. Substantively, cross-sectional analysis cannot answer the focal question motivating

criminological debates on unauthorized immigration: Has the increase in undocumented

immigration increased violent crime? Because unauthorized immigration is necessarily a

process that unfolds over time, cross-sectional analyses are ill-suited for use in answering

this question. Moreover, the methodological distinction between cross-sectional and lon-

gitudinal analysis in immigration–crime research is a salient one. As Ousey and Kubrin

(2017: 1.13) noted in their meta-analysis, “our findings underscore the fact that the choice

between cross-sectional and longitudinal data and analysis procedures is a critical one

that likely impacts findings and conclusions in this area.” They concluded that because

longitudinal research provides greater analytical rigor, such as superior ability to control

for confounding influences, more weight should be given to the findings from longitudinal

studies. To date, however, the literature currently lacks a longitudinal assessment of the

consequences of undocumented immigration for violent crime (but see Light, Miller, and

Kelly, 2017, for an examination of drug and alcohol crimes).

We seek to fill this gap by providing the first longitudinal empirical analysis of the

macro-level relationship between undocumented immigration and violent crime. In com-

bining newly developed estimates of the unauthorized population with multiple data

sources to capture the criminal, socioeconomic, and demographic context of all 50 states

and Washington, DC, from 1990 to 2014, we use fixed-effects regression models to exam-

ine the effect of increased unauthorized immigration on violent crime rates.

This analysis is timely given the growth of the undocumented population in recent

decades. Between 1990 and 2014, the number of undocumented immigrants more than

tripled, from 3.5 million to 11.3 million (Krogstad, Passel, and Cohn, 2016), accounting

for more than a third of the increase in the total foreign-born population over this period.

This wave of immigration generated substantial public angst regarding the criminality of

unauthorized immigrants, leading to immigration reforms and public policies intended

to reduce the purported crimes associated with undocumented immigration (Bohn,

1. To avoid redundancy, we use the terms “undocumented” and “unauthorized” interchangeably

throughout this article.

372 LIGHT & MILLER

Lofstrom, and Raphael, 2014). Indeed, the presumptive link between unauthorized im-

migration and violent crime has become a core assertion in the anti-immigration narra-

tive in public, political, and media discourse (Chavez, 2008) and has been at the center of

some of the most contentious immigration-reform policies in recent years (e.g., Arizona

SB 1070, 2010).

Moreover, concerns over illegal immigration have arguably been the federal govern-

ment’s highest criminal law enforcement priority in recent decades. Between 1986 and

2008, the number of U.S. Border Patrol officers increased 5-fold while the budget for bor-

der enforcement increased 20-fold (Massey, Pren, and Durand, 2016). As a result, today

the U.S. government spends more on immigration enforcement agencies (U.S. Customs

and Border Protection and U.S. Immigration and Customs Enforcement) than it does on

all other principal criminal law enforcement agencies combined, including the FBI, DEA,

Secret Service, Marshal’s Service, and ATF (Meissner et al., 2013). Reductions in crime

and violence have been primary justifications for this dramatic development (Depart-

ment of Homeland Security Immigration and Customs Enforcement [DHS ICE], 2009).

Yet, this vast apparatus of criminal justice machinery has been built up to increase pub-

lic safety with scant empirical evidence that undocumented immigration and violence are

linked (positively or negatively).

The results of our analysis not only inform this contentious policy debate, but they

also provide us with an opportunity to adjudicate competing theoretical perspectives on

the immigration–crime link. Although the weight of the evidence supports the immigrant

revitalization perspective, whereby immigration is said to reduce crime and violence by

attracting immigrants with low criminal propsensities, strengthening local economies, and

bolstering processes of informal social control2(Lyons, V´

elez, and Santoro, 2013), others

argue that social disorganization better captures the contemporary immigration–crime

relationship (Shihadeh and Barranco, 2013). This perspective may be especially relevant

for the unauthorized population who, unlike their documented counterparts, are hindered

from effectively forming economic and social ties as a result of their lack of legal stand-

ing in the community. It is important to note, however, that almost none (save Green,

2016) of the immigration–violence research to date has systematically examined the un-

documented immigrant population, despite the fact that the patterns of authorized and

unauthorized immigration in recent decades have not been uniform. For example, of the

10 states that experienced the largest percent increase in undocumented immigrants be-

tween 1990 and 2000, only two of them were also in the top 10 for relative increases in

lawful immigrants.3

The remainder of this article is organized as follows. The following sections briefly ex-

plicate the contrasting theoretical perspectives linking immigration and violence. Given

that these perspectives have been given ample treatment elsewhere (Ousey and Kubrin,

2009, 2017), we focus our discussion on the applications of theory to unauthorized

2. In the past, researchers have used terms such as the “immigrant concentration view” (Desmond

and Kubrin, 2009), the “ethnic community model” (Logan, Zhang, and Alba, 2002), the “enclave

hypothesis” (Portes and Jensen, 1992), the “community resource perspective” (Feldmeyer, 2009),

and the “Latino paradox” (Sampson, 2008) to describe these effects. Regardless of the term, for

our purposes, what matters is they make the same directional hypothesis: Immigrants provide pro-

social benefits in ways that reduce the prevalence of violent crime.

3. Authors’ calculations of data from the Center for Migration Studies.