Issue Brief: Volume 3, Number 6

Trumped Up Charges: Empirical Effects of U.S. Immigration Reform on Crime and Jobs

Authors: Emily Owens, PhD, Matthew Freedman, PhD, and Sarah Bohn, PhD

Public opinion and political debate on immigration reform have focused largely on two things: the impact of immigration on the economic well-being of U.S. citizens, and the perception that immigrants are prone to crime.


  • Polls show a clear majority of Americans, concerned over issues such as crime and economic competition, wants to see Congress pass a new, comprehensive immigration reform bill.
  • This brief examines the effects, both intended and unintended, of two previous efforts to deter and regulate unauthorized immigration.
  • Recent research shows that the Immigration and Reform Act of 1986, by cutting off access to legal employment for undocumented immigrants, actually increases the probability that some people who are not in the country legally will engage in crime.
  • And despite heightened efforts to police and deport undocumented immigrants in accordance with Section 287(g) of the Immigration and Nationality Act, studies indicate that any resultant decreases in immigrant populations did nothing to improve employment opportunities for low-skilled native residents.
  • While past research reveals some of the limitations and unintended negative impacts of laws aimed at making the U.S. less attractive to unauthorized immigrants, less is known about the potential positive impacts of laws aimed at inclusion. Such knowledge could help to enlighten future debates over comprehensive reform.

According to a recent poll by the Pew Research Center, only about half (51%) of Americans believe that immigrants strengthen the U.S. economically through their hard work and talents—down 6 percentage points from last year—while a sizable minority (41%), including a majority of Republicans (63%), say that immigrants impose an economic burden by taking jobs, housing, and health care resources.[1] Views of the relationship between immigration and crime are even more negative. A Gallup poll in 2007 asking whether immigration makes the crime situation in the U.S. better, or worse, or carries little effect, found that 58% of respondents believed it makes crime worse.[2] When business mogul and Republican presidential contender Donald Trump, in articulating his views on immigration, alleged that Mexico is “sending” people to the U.S. “that have lots of problems”—“they’re bringing drugs, they’re bringing crime”—he was reflecting fears and mistrusts held my many Americans.[3] Not surprisingly, a poll taken earlier this year showed that 73% of Americans want to see Congress deal forcefully with the immigration issue by passing a comprehensive immigration reform bill.[4]

It has been almost 30 years since the U.S. instituted comprehensive immigration legislation. In 1986, Congress passed the Immigration Reform and Control Act (IRCA), creating a pathway to legal status for previously undocumented immigrants. Any non-citizen that could document having entered the U.S. prior to January 1, 1982, and resided continuously in the country since then, could apply for amnesty and permanent legal residency until May 4, 1988. Non-citizens working in agriculture were given a slightly longer amnesty period, through November 30, 1988. In all, IRCA enabled around 2.7 million people to gain legal resident status. At the same time, however, to help stem illegal immigration, IRCA made new demands on employers by instituting the I-9 verification process for new hires, making it illegal for firms to knowingly hire or recruit unauthorized immigrants. By cutting off access to legal employment, IRCA sought to discourage immigrants that could not meet IRCA’s amnesty requirements from staying in the U.S., and deter new immigrants from entering the county.

Ten years later, in 1996, Congress acted again in passing the Illegal Immigration Reform and Immigrant Responsibility Act (IIRIRA). A sprawling piece of legislation, IIRIRA sought to strengthen border controls, tighten enforcement against immigrant smuggling and visa document fraud, establish new rules for apprehending, detaining, and deporting unauthorized immigrants, and buttress the employment restrictions enacted under IRCA (e.g., it introduced the E-Verify program). One important innovation of IIRIRA was the addition of Section 287(g) to the Immigration and Nationality Act, which authorized the federal government to enter into agreements with state and local law enforcement agencies, allowing designated officers at those agencies to enforce immigration law, under the supervision of U.S. Immigration and Customs Enforcement (ICE). Although no actions initially resulted from 287(g), interest in that provision escalated after the 9/11 terrorist attacks, giving rise to enforcement agreements with Florida, Alabama, and several counties in Southern California, among other localities.[5]

Despite popular demands for further action on immigration, recent efforts to build on or modify current legislation either have stalled (Senate bill S.744) or come under legal dispute (President Obama’s executive order from November 2014). As the nation tries to move forward on this issue, it is important to consider what lessons have been learned from previous reform efforts. In particular, have the enforcement policies enacted under IRCA and 287(g) succeeded in addressing the issues that are most on the minds of Americans? Have the employment restrictions and broadened policing responsibilities contributed to preventing crime and improving economic opportunities for native citizens?

Current research suggests not—that U.S. immigration policy, in its punitive approach to discouraging immigration, has had little impact on the economic lot of native workers, and has created an environment that actually invites certain types of criminal activity. The remainder of this brief will summarize what the research indicates, recommend areas for further study, and offer guidelines for shaping the ongoing discussion of immigration reform.

A Pivotal Moment— Criminal Activity and Immigrant Employment Before and After IRCA

The implementation of IRCA offers a unique window for studying U.S. immigration policy. It provided new national standards for immigrant employment that differed significantly from what existed prior to IRCA’s implementation. It therefore gives a means for assessing whether such employment standards constituted an effective policy tool for controlling immigration flows and enhancing the well-being of American communities.

IRCA, in practice, consisted of two primary legalization programs: a general program for Legally Authorized Workers (LAW), which granted legalization to 1.1 million immigrants nationwide, and a program specific to seasonal agricultural workers (SAW), which legalized an additional 1.6 million people. Most applications for long-term residency were approved, including 95.5% of all requests for amnesty filed in Bexar County, Texas. For University of Pennsylvania professor Emily Owens and her coauthors, Matthew Freedman and Sarah Bohn, the heavy flow of immigrants into that county made it an ideal choice for studying the behavior of immigrants before and after IRCA.[6]

Immigration and Naturalization Service records reveal that nearly 29,000 people in Bexar County, which includes the city of San Antonio and is about a two hour drive from the Mexican border, applied for and received permanent legal status under IRCA—about 2.2 times the number of undocumented immigrants estimated by the Census Bureau as living in the county in 1986. To better understand the relationship between immigration and criminal activity, Owens and her coauthors used administrative records from Bexar County that detailed every felony charge between 1980 and 1994. The individual-level data identified both the type of crime and the residence of the alleged offender. They then relied on comprehensive neighborhood characteristics to identify Hispanics who were most likely recent immigrants, in order to distinguish them from the largely U.S.-born or legalized Hispanic population of the county.[7]

IRCA unquestionably improved the labor market prospects of immigrants who applied for and were granted amnesty between 1986 and 1988. Immigrants who chose not to apply for amnesty, were rejected, or came to the U.S. post-IRCA, found fewer economic opportunities available to them, most likely due to the threat of employment sanctions. Without amnesty, immigrant wages dropped as much as 24%, working conditions deteriorated, and job search durations rose.[8]

Such changes to the economic prospects of unauthorized immigrants were predictable results of the employment sanctions in IRCA, but they also had clear, unintended consequences on the subsequent criminal activity of residents in Bexar County. Census data on inmates in prison shows that foreign-born individuals are less likely than native-born citizens to commit crimes overall. And before the end of IRCA’s amnesty period in 1988, Hispanics and non-Hispanics committed roughly an equal number of crimes annually in the county. But the research by Owens, Freedman, and Bohn finds evidence that U.S. immigration policy established by IRCA actually increases the probability that some people who are not in the country legally will engage in crime. When the IRCA amnesty expired in 1988, and unauthorized immigrants were cut off from legal employment opportunities, alleged felonies committed by Hispanics rose 59% in the next few years [Figure 1]. Limiting job opportunities through IRCA increased incidences of crime, but in a very specific way. Empirically, this rise was heavily concentrated in non-violent, felony drug offenses and other income generating crimes that effectively were substitutes for formal employment—crimes like prostitution, gambling, fraud, forgery, car theft, burglary, robbery, and larceny. Income generating crime charges were actually three times as likely to be filed as non-income generating crime charges across all Bexar County neighborhoods.[9]

Figure 1

One complicating factor in calculating the relationship between IRCA and immigrant-specific crime is the introduction at this time of new, stricter anti-drug policies, as well as the possibility that the police may have started focusing more attention on Hispanics after the expiration of IRCA’s amnesty. But the analysis of criminal conviction rates by Owens, Freedman, and Bohn determined that the uptick in felony charges was not driven by changes in drug laws or policing practices, but rather by the new U.S. policy that reduced employment opportunities for immigrants without legal status. Not only was the rise in income generating criminal activity non-trivial, but tellingly, the change in the crime rate among immigrants was in line with other economically high-risk groups, like native residents living below the poverty line.

287(g)’s Inability to Curb Illegal Immigration and Affect the Supply of Native Jobs

Just as IRCA provided a turning point for evaluating the state of immigrant life and the broad effects of reform, so too do two key occurrences from the end of the last decade allow for a better understanding of immigration’s effects on native residents: the economic crisis and the adoption of 287(g) agreements across the country.

A common motivation for stricter immigration enforcement policies, whether law enforcement partnership agreements like 287(g) or enhanced border security, stems from a prevalent, public belief that restricting the inflow of immigrants, especially those unauthorized to live and work in the U.S., will increase employment opportunities for low-skilled, native residents who are looking for work. But there are two issues with this belief. First, it is not clear that policy actions, even if desirable, actually have the ability to affect population movements, as the decision to emigrate is based largely on economic considerations. A second and related point is this: in the face of an economic downturn, immigration becomes a less attractive proposition and the supply of immigrants decreases naturally.[10]

The fact remains, however, that 287(g) did have some effect on local foreign-born populations. By 2011, the program had identified about 186,000 immigrants for deportation and around 126,000 left the country voluntarily.[11] Therefore, understanding the extent to which 287(g) impacted the size of these populations and increased native employment and wages, if at all, may be instrumental in informing and shaping immigration enforcement policies in the future.

New research conducted by Sarah Bohn and Rob Santillano looks at the local economic impact of immigrant removal through 287(g) by comparing counties that participated in a 287(g) agreement versus neighboring counties that did not.[12] Using data from Quarterly Census of Employment and Wages reports, the Department of Homeland Security, and the National Conference of State Legislatures, Bohn and Santillano utilized what is known as a difference-in-difference approach—similar to what the researchers studying IRCA used—in order to control for time-varying local economic shocks. This approach isolates the economic consequences of implementing 287(g) and provides a clearer assessment of 287(g)’s effects on employment and wages than, for instance, comparing 287(g) localities with non-participating localities in different regions of the country based on population alone. These researchers found that compared to neighboring counties, wages in counties with 287(g) agreements did not uniformly rise. In fact, wages in these localities were lower in administrative services (e.g., janitorial work) and slightly lower in construction, although not in other immigrant-heavy industries like manufacturing or accommodation and food services.

Additional findings, looking specifically at immigration from Mexico, confirm that 287(g) did not have a substantial impact on aggregate economic activity, and that there is no evidence that the program improved employment opportunities for native, low-skilled residents or helped to protect them from the negative effects of the recession.[13] Tellingly, there is no indication of correlation between changes in local foreign-born Mexican populations and jobs in an industry that was hit especially hard during the recession and that employed many workers, undocumented and native alike: construction [see Figure 2]. The implication here is that native and foreign-born workers are likely complements, not competitors, because each group fills specialized needs in respective industries.

Figure 2

Ultimately, 287(g) was effective at removing unauthorized immigrants only when applied forcefully, and in some cases, enforcement measures led to media and federal government inquiries about alleged civil rights violations and misappropriations of law enforcement resources. Among 287(g) participating localities, the largest reductions in foreign-born populations were seen in Dallas, Los Angeles, Riverside, and Phoenix. Aside from these four, no other 287(g) participating locality removed more unauthorized immigrants between 2007 and 2009 than any one of several so-called “sanctuary cities,” including New York, Chicago, Denver, and San Francisco, which did not (by definition) participate in 287(g). On a national level, 287(g) had little to no ability to control the outflow of undocumented immigrants. The recession, on the other hand, proved quite effective at reversing the growth trend of immigration into the United States. Yet even the observed decreases in foreign-born Mexican immigrants did nothing to improve the employment opportunities for low-skilled native residents.[14]

Policy Prescriptions and Areas for Further Research

As Emily Owens has noted, while it is not clear that policies like heightened policing or employer sanctions are effective at all in reducing immigration, it is also unknown how much loosening those restrictions would increase immigration. The less responsive immigration flows are to U.S. policy, the more policymakers should consider trying to reduce the negative social effects associated with undocumented immigrants directly, rather than waging a seemingly futile battle to reduce the number of people overstaying their visas or entering the country illegally.

Given the lack of observed economic benefits to native residents from “attrition through enforcement” policies, and their clear ineffectiveness at achieving desired immigration flows, even when applied rigorously for a sustained period of time, it is probably unwise that policymakers have trended in this direction, as evident in recent state laws like Arizona’s S.B. 1070 and Alabama’s H.B. 56, as well as the federally initiated Secure Communities program of 2008, which is a deportation program similar to 287(g). Part of this push for stricter enforcement and deportation has involved a renewed call to eliminate “sanctuary cities” and other areas that do not actively seek to identify and remove undocumented immigrants. However, the revitalized neighborhoods and small businesses formed and aided by immigrants have led many local governments in the country to resist such calls. Some states (i.e., Illinois, Massachusetts, and New York) have even attempted, unsuccessfully, to withdraw from the Secure Communities program.[15] These disagreements between different regions of the country illustrate just how divided government officials are in their understanding of the economic impact of immigration.

Another example of an “attrition through enforcement” program of questionable merit is the so-called “detention bed mandate.” In late 2013, NPR reported on a Congressional mandate issued to ICE that required the agency to fill “34,000 beds in some 250 facilities across the country, per day, with immigrant detainees.”[16] This program began in 2009 and costs taxpayers an estimated $2 billion annually. Immigrants in these facilities can be made to stay for as little as one day or for as long as a few years, again beckoning questions about potential civil rights violations. But in short, the practice of enforcing an arbitrary inmate quota likely has an insignificant effect, given the results noted above, on decreasing crime or improving native employment.

Moving forward, as the immigration debate continues, one of the chief roadblocks in studying illegal immigration will be the continued decentralized nature of social data collection in the U.S., especially compared to Europe, and particularly when it comes to criminal justice issues. This fact was on clear display recently in the case of the undocumented immigrant with prior criminal convictions who shot and killed a woman in San Francisco.[17] In the U.S., the best research is on tightening immigration reforms, but there is less research being done on how efforts to actively incorporate undocumented immigrants into society (e.g., sanctuary city ordinances, local IDs and drivers licenses, and in-state school tuition for child arrivals) affect criminal behavior and economic outcomes for natives, such as wages and employment. In other words, while past research has revealed insights about the potential unintended negative impacts of laws aimed at making the U.S. a less attractive place to live, less is known about the potential unintended positive impacts of laws aimed at inclusion. This knowledge could help to enlighten future debates over comprehensive reform.


Perceptions of how illegal immigration affects native residents have shaped policies, but these policies are potentially ineffective both in general (i.e., at stopping the tide of undocumented immigrants from entering the country, as well as at deporting them) and in their specific focuses, like in lowering crime and improving native employment. Policymakers should consider objectively the effects of past policies such as IRCA and 287(g), before instituting new, non-data driven mandates and legislation in response to public demand.


About the Authors

Emily Owens, PhD
Associate Professor of Criminology, University of Pennsylvania and Associate Professor of Business Economics and Public Policy, The Wharton School

Professor Owens studies a wide range of topics in the economics of crime, including policing, sentencing, and the impact of local public policies on criminal behavior. This includes studying how government policies affect the prevalence of criminal activity as well as how agents within the criminal justice system, particularly police, prosecutors, and judges, respond to policy changes. Professor Owens is currently engaged in an NIJ-funded field experiment evaluating a police training program, along with ongoing research projects on alcohol regulation, immigration policy, and local economic development programs.

Matthew Freedman, PhD
Associate Professor, School of Economics, Drexel University and Scholar, Penn Institute for Urban Research

Professor Freedman is a labor economist whose research revolves around job mobility and earnings dynamics within and between industries and geographic locations. One main area of current work focuses on the role of job accessibility and spatial mismatch in shaping observed patterns of job and income mobility among low-income workers, with a particular emphasis on how various policies and programs associated with transportation and housing can help to mitigate mismatch problems.

Sarah Bohn, PhD
Research Fellow, Public Policy Institute of California

A labor economist, Dr. Bohn focuses on how policy affects individual and family economic well-being, with particular attention to low-income and vulnerable populations. Her recent research analyzes the latest data on income inequality and economic mobility across California, and she is conducting an in-depth study of pathways to in-demand technical careers through California community colleges. She has published research on poverty, income inequality, the workforce skills gap, California’s community colleges, and the labor market impact of immigration policy.


 [1] “Broad Public Support for Undocumented Immigrants,” People Press, June 4, 2015.

 [2] “The Politics of Immigration and Crime,” by Mary C. Waters and Jessica T. Simes, (PDF).



 [5] To date, over 70 law enforcement agencies have entered into 287(g) agreements. 287(g) is now being phased out across the country, but there is much to learn from the results of past agreements.

 [6] Matthew Freedman, Emily Owens, and Sarah Bohn. 2015. “Immigration, Employment Opportunities, and Criminal Behavior.” 

 [7] Demographic and population research shows that immigrant neighborhoods tend to have higher poverty rates, more residents per housing unit, more people of Mexican descent, a higher fraction of adults who speak Spanish at home, and a higher fraction of foreign-born residents. In Bexar County, nearly half of neighborhood residents in the sample identified themselves as being of Mexican descent, about 9% of residents were foreign-born, and 6% were identified as non-citizens.

 [8] Sherrie Kossoudji and Deborah Cobb-Clark, “Coming out of the Shadows: Learning about Legal Status and Wages from the Legalized Population,” Journal of Labor Economics 20 (2002): 598-628.

 [9] Census block groups were the official measure of neighborhoods used by Owens and her coauthors.

 [10] Emilio A. Parrado, “Immigration Enforcement Policies, the Economic Recession, and the Size of Local Mexican Immigrant Populations,” The Annals of the American Academy of Political and Social Science 641 (2012): 16-37.

 [11] Randy Capps, Marc R. Rosenblum, Cristina Rodriguez, and Muzaffar Chishti, “Delegation and Divergence: A Study of 287(g) State and Local Immigration Enforcement,” Washington, DC: Migration Policy Institute, 2011. [Note: Many more deportations occurred outside of 287(g). For instance, the American Community Survey reveals that roughly 393,000 immigrant removals took place in 2009.]

 [12] Sarah Bohn and Robert Santillano. 2015. “Local Immigration Enforcement and Local Economies” (Unpublished Working Paper, obtained from Bohn).

 [13] Parrado (2012).

 [14] Ibid.

 [15] Ibid.

 [16] “Little Known Immigration Mandate Keeps Detention Beds Full,” NPR, November 19, 2013. 

 [17] “San Francisco Murder Case Exposes Lapses in Immigration Enforcement,” New York Times, July, 8, 2015.