While theorists have long suspected—based largely on anecdotes and a handful of city-level studies of a narrow sample of residents—that immigrant communities report fewer crimes to the police, two researchers have now provided some evidence to back it up.
Min Xie and Eric P. Baumer, academic researchers from the University of Maryland and Pennsylvania State University, respectively, found that rates of violence reported in traditional, long-anchored immigrant neighborhoods are similar to those observed elsewhere.
But, diving into 19 years of county-level data from the U.S. Census Bureau’s National Crime Victim Survey (NCVS) and the American Community Survey (ACS), they concluded that, “within newer immigrant destinations, we observe much lower rates of violence reporting in neighborhoods with a large concentration of immigrants.”
“Our study findings reveal comparable patterns for whites,blacks and Latinos,” they added.
Their conclusions, published in the journal Criminology, represent the first to attempt a county-level picture of the national landscape, without relying heavily on police-generated data.
That’s to say, crucially, that counties who have developed a larger immigrant population than the national average between 1996 and 2014 tend to chronically under-report rape, assault, and other forms of violent crime to the police.
Particularly, they found that when the share of residents who are immigrants in a county reaches around 35 percent, there is a marked decrease in willingness to report violent crime to the police. This only applies to counties who have exceeded the national average for immigrant populations over the past 20 years—and not for counties with longstanding large immigrant communities.
The researchers argued that this finding could be interpreted through the lens of legal cynicism, a social theory which explains that immigrant communities can grow skeptical of public institutions due to barriers including language, culture, and negative experiences with authorities such as the police.
This explanation has some prima facie appeal in a post-9/11 era of aggressive federal immigration enforcement, in which questions of immigration and identity are roiling the nation’s cultural battlefield.
The Trump administration is even pushing to include a question about citizenship on the 2020 census.
They also argued that under-reporting is likely a problem of access to services and public officials.
[O]ur findings […] emphasize the greater barriers that residents of immigrant neighborhoods may encounter in seeking police help. Within nontraditional destinations (less-established immigrant communities), all three groups considered in our study—Whites, Blacks, and Latinos—share a tendency to have considerably lower probabilities of reporting crime victimizations to the police when the level of immigrant concentration in their neighborhood reaches a high level.
They also suggest that the differences in crime reporting between newer and longstanding immigrant communities could be due to the long-term development of deeper institutional ties and public resources tailored to well-established immigrant communities’ needs.
Yet lurking underneath the topline findings and their cautious interpretation is a much more muddled picture of the questions the researchers set out to answer.
An Empirical House of Cards
The ultimate value of Xie and Baumer’s study might be in the novelty of the approach rather than the strength of its findings.
For one, crime reporting in predominantly immigrant communities remains an extremely opaque topic for researchers. That is partially because self-reported data collected by government agencies might carry inherent biases, as the researchers rightly pointed out.
For example, an undocumented woman suffering from domestic abuse or human trafficking might be less likely to reach out for help, despite certain legal protections and visa programs that ideally should protect her from government retaliation relating to her immigration status. When it comes time to report information about crimes she survived, her perception of insecurity might lead her to hesitate about offering information to government surveys.
In other words, there is no guarantee that U.S. Census Bureau data is even a clear window into true levels of violent crime.
The researchers did not include immigration status as a relevant variable but indicated that future research should examine the question.
Further, the National Crime Victim Survey is only administered in English and Spanish, thus privileging the responses and experiences of those speakers while under-representing those of others. Those who do not speak either of these languages might be even more unlikely to respond or report violent crime.
That said, their analysis accounted for many other factors that might relate to people’s willingness to report violent crime and otherwise confound their results, ranging from: victim-offender characteristics, such as the type or severity of the crime, relationship between the parties, and whether it was committed in public; and neighborhood characteristics such as racial/ethnic make-up, residential stability, population density, socioeconomic status, and police force characteristics.
Yet the validity of this research hinges on whether the percentage of a community that was born in another country is a strong enough stand-in variable for the many complex, multilayered issues that affect immigrants’ relationships to their local police forces.
Rather than clarifying the statistical picture by weeding out these many factors, it might be argued that the researchers are oversimplifying the question of immigrant crime reporting by attempting to boil it down to the sheer proportion of immigrants in a given community.
In that case, their findings could be interpreted as little more than an exercise in hypothesis testing.
That’s to say that some problems, such as under-reporting of crime in immigrant communities, truly sprout from numerous social issues that might not be sliced and diced or reduced into a percentage of residents born in another country, without biasing the data.
While this research could simply be evidence of social scientists’ preference for clean, simple narratives over cloudy, ambiguous realities, this emerging research is a worthy attempt to hear what can survey data and parabola curves have to say about trust in institutions and the way we build shared identity.
The full article is available online.
Roman Gressier is a TCR contributor and news intern.