State of K
Improving Democracy one Question at a Time
NOTE: This is a sample. It is for presentation purposes only.
Some links and buttons are disabled.
Does deporting undocumented immigrants reduce crime?
No. Deporting undocumented immigrants does not reduce crime.
We believe this is the best available short answer based on our reading of the 4 studies we found that examine this question.
However, we believe this question is underresearched. Additional study may change our conclusion.
The 4 studies were published from 2007 to 2014.
All of them focused on the United States of America.
Evidence from three of the studies supports this answer. One other study reached mixed conclusions.
Does immigration enforcement actually reduce crime? Surprisingly,
little evidence exists either way—despite the fact that deporting noncitizens who commit
crimes has been a central feature of American immigration law since the early twentieth
century. We capitalize on a natural policy experiment to address the question and, in the
process, provide the first empirical analysis of the most important deportation initiative to
be rolled out in decades.
The policy initiative we study is “Secure Communities,” a
program designed to enable the federal government to check the immigration status of
every person arrested for a crime by local police. Before this program, the government
checked the immigration status of only a small fraction of arrestees. Since its launch, the
program has led to over a quarter of a million detentions. We exploit the slow rollout of
the program across more than 3,000 U.S. counties to obtain differences-in-differences
estimates of the impact of Secure Communities on local crime rates. We also use rich
data on the number of immigrants detained under the program in each county and
month—data obtained from the federal government through extensive FOIA requests—to
estimate the elasticity of crime with respect to incapacitated immigrants.
Our results show that Secure Communities led to no meaningful reductions in the FBI index crime
rate. Nor has it reduced rates of violent crime—homicide, rape, robbery, or aggravated
assault. This evidence shows that the program has not served its central objective of
making communities safer.
In 2008, the federal government introduced “Secure Communities,” a program that requires local law enforcement agencies to share arrestee information with federal immigration officials. We employed the staggered activation of Secure Communities to examine whether this program has an effect on crime or the behavior of local police. Supporters of the program argue that it enhances public safety by facilitating the removal of criminal aliens. Critics worry that it will encourage discriminatory policing. We found little evidence for the most ambitious promises of the program or for its critics’ greatest fears.
Although a large body of evidence reports that municipal police can have an appreciable effect on crime, involving local police in federal immigration enforcement does not seem to offer measurable public safety benefits. Noncitizens removed through Secure Communities either would have been incapacitated even in the absence of the program or do not pose an identifiable risk to community safety.
Contemporary criminological research on immigration has focused largely on one aspect of the immigration process, namely, the impact of in-migration (i.e., presence or arrival) of foreign-born individuals on crime. A related but understudied aspect of the immigration process is the impact that the removal of certain segments of the foreign-born population, and specifically undocumented or deportable aliens, has on aggregate levels of criminal violence.
In an effort to cast new light on the association between forced out-flows of immigrants and crime, we begin with descriptive analyses of patterns of deportation activity across the continental United States over an eleven-year period (1994–2004). We then examine the relationship between deportation activity and violent crime rates in a multilevel framework wherein Metropolitan Statistical Areas (MSAs) are situated within border patrol sectors.
The results of dynamic regression modeling indicate that changing levels of deportation activity are unrelated to changing levels of criminal violence for the sample of MSAs for the national at large. However, we also detect significant interactions by geographic location for selected violent offenses. For MSAs within sectors along the Mexican border, the deportation measure exhibits a significant negative effect on one indicator of criminal violence—the aggravated assault rate. For MSAs within non-border sectors, the effect of the deportation measures is significantly positive for the violence crime index and the aggravated assault rate.
Overall, our analyses indicate that the relationship between deportation and criminal violence is complex and dependent on local context.
The perception that immigration adversely affects crime rates led to legislation in the 1990s that particularly increased punishment of criminal aliens. In fact, immigrants have much lower institutionalization (incarceration) rates than the native born - on the order of one-fifth the rate of natives. More recently arrived immigrants have the lowest relative incarceration rates, and this difference increased from 1980 to 2000.
We examine whether the improvement in immigrants' relative incarceration rates over the last three decades is linked to increased deportation, immigrant self-selection, or deterrence. Our evidence suggests that deportation does not drive the results. Rather, the process of migration selects individuals who either have lower criminal propensities or are more responsive to deterrent effects than the average native. Immigrants who were already in the country reduced their relative institutionalization probability over the decades; and the newly arrived immigrants in the 1980s and 1990s seem to be particularly unlikely to be involved in criminal activity, consistent with increasingly positive selection along this dimension.