This post expands on some points I made in a post to the Open Borders Action Group on Facebook. There, I expressed puzzlement at the emphasis people pay to using deportation of criminal non-citizens (and in particular those in violation of authorized immigration status) as a crime-fighting tool. That Facebook post, and this blog post, will focus on the United States, though many of the points made are general.
To many people, the idea that there exist foreign-born non-citizens, particularly illegal immigrants, who have criminal records and still roam the streets safely is an indicator that United States immigration enforcement is dysfunctional and broken. Thus, Donald Trump’s remarks about illegal immigrants and crime struck a chord with a lot of his audience. And the killing of Kate Steinle by illegal immigrant and repeat felon Juan Francisco Lopez-Sanchez (who has admitted to firing the killing shots but claims they were accidental) was viewed as evidence of a breakdown of law enforcement. The killing has led to a proposal for a new law called “Kate’s Law” that has led to a lot of discussion, see for instance here and here.
This post has two main goals:
- Critiquing the high-level view that deportation can be a key strategy for reducing crime, particularly in the United States. I don’t claim that deportation can never reduce crime, just that it’s not a proven strategy to do so and most of the data suggest its effect is negligible in magnitude, ambiguous in sign, and swamped by the other side-effects.
- Emphasizing the importance on the open borders side of not carelessly conceding too much to restrictionists regarding how to deal with criminals, due to concerns about slippery slopes, ratchet effects, and logical inconsistency. I don’t claim that it’s inappropriate to make any exceptions for criminals, just that some exceptions should be made after careful consideration of all the angles rather than as a generous carte blanche of “do what you want with the criminals.”
A couple of notes here. Since this post is largely conceptual, I’m going to conflate a few fairly different notions. There is a notion of birthplace (native-born versus foreign-born), a notion of citizenship (citizen versus non-citizen), and a notion of authorization for status (legal versus illegal immigrant). Also, in the United States, many non-citizens are on non-immigrant visas, i.e., their visa does not specify immigrant intent, although many of them later transition to a long-term immigrant status. A detailed analysis of the empirics of crime patterns would need to avoid conflating these issues, but since the purpose of this post is rather different I’ll be a little careless.
Table of contents
1.1. How does the law treat people convicted of crimes based on immigration status?
The short answer here is that, as far as the law goes, non-citizens convicted of crimes are no more a hazard to public safety than citizens convicted of the same crimes. First off, anybody who is not a US citizen, lawful permanent resident, or conditional permanent resident, and who has been convicted of an aggravated felony, a category of crime that includes both violent and non-violent crimes (some of the latter being victimless crimes) can be subject to a speedy removal process called administrative removal for aggravated felons, which means that the person can be removed simply through some paperwork and without getting a hearing before an immigration judge (more here).
Even so, as immigration.procon.org notes, in the United States, those convicted of violent crimes need to first finish their prison terms, and after that they may be deported to their home country. And the way administrative removal works, they are deported straight out of prison, so they don’t spend a single day free in the streets of the United States: it’s prison in the US and then back to freedom in their home country.
In contrast, citizens are required to finish their sentence in prison, and after that they are free — to roam around in the United States. Even if the deportation of criminal non-citizens is a flawed process with many people failing to get deported, or returning to the United States, it’s at worst as bad (from the public safety perspective within the United States) as the treatment of criminal citizens.
For those convicted of non-violent crimes, the person may be deported before the completion of his or her sentence. Since re-entering the United States seems a task of comparable or greater hardness to having one’s prison term shortened or getting out on parole as a U.S. citizen, it’s again unclear that non-citizens pose a greater risk to public safety than citizens. Of course, there’s a big question mark regarding whether people convicted of non-violent crimes are threats to public safety to begin with.
The United States is also a participant in the International Prisoner Transfer Program. A prisoner who is a citizen of another participating country may transfer from a United States prison to a prison in the home country, subject to approval by both countries. However, such transfer must be initiated through a request by the prisoner, and therefore does not concern us here.
A bit more about re-entering after having been deported for crimes. The United States has a summary removal procedure called reinstatement of removal. What this says is that somebody who re-enters the United States without authorization after having previously been deported, removed, or excluded can be removed again without any kind of hearing or process, simply by “reinstating” the previous order. This in particular applies to those who were subject to administrative removal for aggravated felonies, or otherwise deported or excluded based on criminal history. Of course, after the person gets re-deported, the person may re-enter yet again, and get deported yet again, and so on. But two things to note: first, insofar as this isn’t enough to keep the streets of the United States safe, the problem can’t really be solved by more deportations, but by more imprisonment (which is sort of what Kate’s Law was pointing to). Second, the same public safety challenge applies to citizens as well, except that in the case of citizens, there isn’t even an option to deport people, however temporarily.
The upshot of all this is that, for a citizen and a non-citizen who commit the same crime, the law enforcement response in the case of the non-citizen is equally or more protective of public safety (in the US) compared to the response in the case of the citizen. If the law enforcement apparatus of the United States is lenient enough that criminal non-citizens can roam the streets freely and with impunity, then the same is even more true of criminal citizens.
An old post by Nathan, titled Answer to Vipul’s question about enforcement, has some interesting thoughts on deportation that are relevant to this discussion. Basically, Nathan argues that deportation is rarely the appropriate response, even if there are cases where it is not an unjust response:
I’m not absolutely wedded to the idea that deportation is never permissible. However, I can’t think of any situations where it would be appropriate. There are certainly crimes for which deportation would not be an excessive punishment; but for those crimes it’s usually either inapt or insufficient. A man guilty of rape or murder shouldn’t be deported, but imprisoned. Maybe there are scenarios where deportation would be the right thing to do, but I can’t think of them. I have some sympathy for the Nicene council which banished Arius the heresiarch for his views when they temporarily had the emperor on their side– they had suffered much at the hands of the pagans, and would yet suffer much at the hands of the Arians, and mere banishment is impressively moderate under the circumstances– but it’s not a precedent to imitate today, when the principle of free thought has been firmly established.
1.2. Immigrant crime rates appear lower than, and definitely aren’t significantly higher than, native crime rates
In the United States, one of the main concerns surrounding crime is that of crime by Hispanic illegal immigrants. We have a page on the subject that links to many literature reviews, and you should also read Alex Nowrasteh’s recent summary of the research and my co-blogger Joel’s take on immigrants and crime.
The broad consensus of these reviews appears to be that the foreign-born are considerably less likely to engage in crime than the native-born, and that this effect holds in aggregate, within each ethnicity, and for every combination of ethnicity and high school graduation status. Admittedly, the threat of deportation for crime is believed to be one contributing factor to the lower crime rate, but scholars who have studied the issue believe it to only be a partial explanation. The a priori selection for greater future orientation is believed to be another driving factor in the lower crime rates, and this applies to both legal and illegal immigration, and to both the status quo and substantially more liberal migration policy.
On the other hand, Hispanics have crime rates somewhere between non-Hispanic whites and blacks, which is a contributing factor to the perception of high Hispanic crime. But a lot of this higher Hispanic crime doesn’t come from foreign-born Hispanics.
In addition to comparing overall crime rates, we can also look at specific research on the effect of deportation on crime rates. Alex’s recent summary of the research includes a discussion of two relevant pieces of research:
The phased rollout of the Secure Communities (S-COMM) immigration enforcement program provided a natural experiment. A recent paper by Thomas J. Miles and Adam B. Cox used the phased rollout to see how S-COMM affected crime rates per county. If immigrants were disproportionately criminal, then S-COMM would decrease the crime rates. They found that S-COMM “led to no meaningful reduction in the FBI index crime rate” including violent crimes. Relying on similar data with different specifications, Treyger et al. found that S-COMM did not decrease crime rates nor did it lead to an increase in discriminatory policing that some critics were worried about. According to both reports, the population of immigrants is either not correlated, or negatively correlated, with crime rates.
As far as long-run immigration policy is concerned, one could plausibly argue that, even if the foreign-born have lower crime rates than the native-born, allowing more immigration can still raise crime if the children of these foreign-born have higher criminal propensity. This line of reasoning is partly supported by evidence, both with respect to Hispanics in the United States, and with respect to other immigrant groups historically; this phenomenon has been discussed under the name of second-generation crime.
For the purposes of deportation policy, however, this doesn’t apply, because we are specifically talking about deporting non-citizens for crimes they have committed, rather than crimes we expect their children might commit. And native-born people in the United States are United States citizens (by birthright citizenship) so there are no official grounds for deporting them.
1.3. In absolute terms, crime by immigrants is a small fraction of overall crime
In the United States, the foreign-born constitute about 13% of the population. Given that their crime rates are somewhat lower than those of the native-born, they account for less than 13% of the overall crime in the United States. Targeting crime by immigrants therefore won’t make a huge dent in the overall crime problem.
Concretely, what this means is that if you believe the criminal justice system is too lenient against the foreign-born, and that this creates a significant crime risk for natives, then you should be far more concerned about the criminal justice system being too lenient overall. For every case of a criminal foreign-born non-citizen individual who was either acquitted or released after serving a prison term and then committed more crimes, you’ll probably find many more native-born citizens who do the same thing. Perhaps the relevant remedy here is to make prison terms longer for particular types of offenses, or to better identify those who may be repeat offenders. What the optimal remedy is, and how to balance the rights of former criminals with public safety needs, is not the topic of this post. But it behooves those concerned about crime levels to consider the problem in generality rather than find solutions for a subset of the population that contributes little to the overall problem.
Note that this definitely won’t hold under open borders: under open borders, the foreign-born will be a much larger share of the population, and are likely to contribute a significant share of overall crime. The question of what crime rates would be under open borders is open. It is plausible that the currently observed phenomenon of lower immigrant crime rates than native crime rates will break down under open borders, though I still don’t expect a massive overall increase in crime rate. I considered these questions in an earlier post, and we’ll hopefully have more coverage of the issue.
1.4. Are there unique challenges associated with domestic criminal law enforcement when applied to non-citizens?
One plausible argument for choosing deportation as a crime-fighting strategy for non-citizens is that domestic criminal law enforcement becomes particularly hard for these people. Is that true?
Ironically, it is, but largely because of immigration enforcement. Law enforcement officers have difficulty carrying out their job in immigrant communities partly because of the distrust in these communities of law enforcement, given their fear of deportation and harassment. This leads to a dynamic where police officers tend to avoid the area, leaving the policing of these areas to those prone to corruption and bullying, further worsening the interaction between police and residents. A similar phenomenon been observed for many black communities in the United States, where the relevant form of enforcement is not immigration enforcement but other laws such as drug enforcement and Broken Windows policing.
It is partly for this reason that many “sanctuary cities” have adopted explicit policies surrounding non-enforcement of federal immigration laws. In other words, police are instructed to focus on the goal of fighting crime, leaving the enforcement of federal immigration law to federal authorities. In other words, to the extent that unique challenges apply to domestic law enforcement for non-citizens, they point in the direction of separating law enforcement from immigration enforcement.
A small note here about crime in border towns specifically as a result of illicit border activity. Organized crime plays an important role in facilitating drug smuggling and migrant smuggling, and the clashes between different organized crime groups, and between them and law enforcement agencies, can be responsible for higher-than-usual violence levels in border towns. That being said, as an empirical matter, it appears that overall crime rates are lower in border towns than in comparably sized interior towns. One of the lowest-crime areas, El Paso, is a border town in Texas whose low crime rate has even been called a miracle. The oft-noted point that border towns account for a disproportionate rate of federal crimes (which include crimes related to smuggling) does not impugn their overall safety record.
2.0. Is this worth making an issue of? Can’t the treatment of violent criminals be a small concession that makes the open borders position much more widely palatable?
Criminals are one of the few categories for which many open borders advocates are willing to make exceptions to their general view that borders should be open to all. Thus, for instance, Bryan Caplan writes:
Hey Mr. Caplan,Do you think Israel should open their borders?
Yes. But I wouldn’t strongly object if they excluded people with violent criminal records or denied new-comers the vote. (Same goes for countries other than Israel, too).
It’s not clear to me if making a clear exclusion for criminals is philosophically consistent, but the argument for public safety being a valid concern in immigration law does carry some weight. In an earlier post in December 2012, I considered in detail the question of whether blanket denial of the right to migrate based on a criminal record is just (and also linked to many other people who had conceded an exception to open borders for violent criminals).
My purpose when I wrote that post, way back in 2012, was to simply explore the space of possibilities regarding how to trade off the right to migrate with public safety concerns in receiving countries. However, as I’ve thought more about this and looked more at the types of disputes and debates that arise in practice, a few other concerns have emerged in my mind.
2.1. Scope creep with criminality and immigration
The idea of keeping criminals out, and deporting those who commit crimes, is subject to significant scope creep. Once we start seeing immigration policy as a way to select and shape a better society, why stop at merely excluding violent criminals? Why not also aim to exclude people who have a 50%+ probability of being net fiscal drains, or who are more likely than not to vote the wrong way? And even within the realm of crime, why stop merely at those crimes that actually merit prison terms? Why not expand the scope to everything ranging from playing loud music to running a gambling house to downloading music in violation of copyright law?
In fact, this particular slippery slope is not merely hypothetical. It’s already happened. As already mentioned, United States immigration law can exclude and deport people for aggravated felonies, many of which are neither aggravated nor felonies. The Immigration Policy Center, an immigrant rights and legal advocacy group, has a good overview. Here’s how the IPC’s overview puts it:
As initially enacted in 1988, the term “aggravated felony” referred only to murder, federal drug trafficking, and illicit trafficking of certain firearms and destructive devices. Congress has since expanded the definition of “aggravated felony” on numerous occasions, but has never removed a crime from the list. Today, the definition of “aggravated felony” covers more than thirty types of offenses, including simple battery, theft, filing a false tax return, and failing to appear in court. Even offenses that sound serious, such as “sexual abuse of a minor,” can encompass conduct that some states classify as misdemeanors or do not criminalize at all, such as consensual intercourse between a 17-year-old and a 16-year-old.
While aggravated felonies are considered serious enough to allow for administrative removal for those who are not US citizens or permanent residents, there are also other, lesser, “crimes” that can be used to both exclude and deport people, albeit with more of a semblance of due process (i.e., they cannot be used as a basis for administrative removal, but they can still be used as evidence against the alien in a hearing before an immigration judge). Crimes that can be used to exclude and deport people are called crimes involving moral turpitude (aka crimes of moral turpitude, and abbreviated as CMT). This category includes aggravated felonies but also includes other crimes. NOLO has a good review.
The United States has also historically passed many laws restricting immigration based on one’s speech and political views, including the Immigration Act of 1903 (also known as the Anarchist Exclusion Act) and the Alien and Sedition Acts of 1798. This, despite the fact that freedom of speech is protected by the First Amendment to the United States Constitution. Once we grant that the public safety interest justifies special punishments for non-citizens (over and above the usual fines and prison terms), keeping the domain of application restricted to crime would be hard.
My co-blogger John Lee has done a great post linking to Pulitzer Prize-winning journalist Linda Greenhouse’s discussion of how migrants could be deported for minor offenses (also read John’s follow-up post discussing the resolution of one of the discussed cases).
2.2. Who has responsibility for shaping a criminal?
Personally, I reject the notion that state or national governments are morally responsible for the criminal actions of particular individuals who were born there or raised there. There could be exceptions where state propaganda or state action facilitates criminal activity, but state complicity in private crime needs to be positively established — not merely assumed. Therefore, I don’t believe, for instance, that just because a Chinese citizen came to the United States and committed crimes, the Chinese government, the Chinese nation, or the Chinese people as a whole are “responsible” for that crime and “deserve” to have the person back (this argument is a variant of the state responsibility thesis that has been cited by some philosophers as an argument against open borders).
Even if you believed in high-level national responsibility for the criminal actions of individuals, it’s not clear what nation gets the blame for immigrant crime. Is it the nation the person holds official citizenship of? Is it the nation the person grew up in? Is it the nation where the person first started on the path to crime? In the United States, DREAMers are likely to have had many of their formative experiences in the United States. Thus, we could reasonably argue in the case of DREAMers who commit crimes, any national responsibility for the crimes falls on the United States, rather than their birthplace. Even for those who migrate as adults and then commit crimes, their path of crime may well have begun in the United States. At best, the logic of responsibility can be used to deport criminals who committed their first deportation-worthy crime in their country of origin, in the same way as it could be used to deny initial entry.
2.3. Criminals can commit crimes elsewhere too
From a universalist perspective, deporting those with criminal proclivities, whom we believe could be repeat offenders, doesn’t really solve the problem: the person could commit crimes elsewhere too. There could be some cases where deportation might reduce criminality, for instance, deporting members of a gang could break up the criminal activity of the gang, and individual deported gang members may be unlikely to continue to engage in the relevant crimes (on the other hand, they may start new gangs). It’s unclear that the countries the criminals are being deported to would be more capable of dealing with the criminal activity — they may well be less able to handle it. Perhaps a cost-benefit analysis would still show that deportation reduces overall expected global crime, but such a claim needs careful argumentation.
Of course, citizenists and territorialists in any country would consider the reduction of crime within the country (and/or directed at citizens of the country) to be more important than reducing global crime. So it’s understandable that they accept deportation as a possible crime reduction strategy. But those of a more universalist bent should push back against this reasoning.
Co-blogger Joel pointed me to an article in The Atlantic that made the interesting claim that deporting gang members from the U.S. had actually increased organized criminal actiity both in the U.S. and in the countries the people were deported to. Here’s a key excerpt from the article:
MS-13 formed in the Rampart area of Los Angeles in 1988 or 1989. A civil war in El Salvador had displaced a fifth of that country’s population, and a small number of the roughly 300,000 Salvadorans living in L.A. banded together to form the gang. But MS-13 didn’t really take off until several years later, in El Salvador, after the U.S. adopted a get-tough policy on crime and immigration and began deporting first thousands, and then tens of thousands, of Central Americans each year, including many gang members.
Introduced into war-ravaged El Salvador, the gang spread quickly among demobilized soldiers and a younger generation accustomed to violence. Many deportees who had been only loosely affiliated with MS-13 in the U.S. became hard-core members after being stranded in a country they did not know, with only other gang members to rely on. As the gang proliferated and El Salvador tried to crack down on it, some deportees began finding their way back into the U.S.—in many cases bringing other, newly recruited gangsters with them. Deportation, incubation, and return: it’s a cycle we’ve been caught in ever since.
Today, MS-13 has perhaps 6,000 to 10,000 members in the United States. It has grown moderately in recent years and now has a presence in 43 states (up from 32 in 2003 and 15 in 1996). Most members of the gang are foreign-born. Since 2005, ICE has arrested about 2,000 of them; 13 percent have been deported before.
Salvadoran police report that 90 percent of deported gang members return to the U.S. After several spins through the deportation-and-return cycle, MS-13 members now control many of the “coyote” services that bring aliens up from Central America. Deportation—a free trip south—can be quite profitable for those gang members who bring others back with them upon their return.
While I don’t know enoughabout the specifics to endorse the claim of the article, this is the sort of ripple effect that people concerned about the long-run effects on global crime would have to account for. These kinds of effects are hard to predict, but a reasonable rule of thumb is that they’re likely to be less positive overall than the naive view of deportation as “taking criminals off the streets” suggests.
Much of the current discussion on immigration and crime comes from two angles: the use of anecdotes to justify strong immigration restriction and deportation policies against non-citizens accused of crimes, and the use of empirical data to study the relationship between migration status and crime. In addition, the defense of the civil and procedural rights of non-citizens accused of crimes is also a perspective that gets some airing. My post looks at the issue from a few slightly different angles. It focuses on whether deportation can or should be an important part of a crime-fighting strategy, and highlights some other relevant considerations about moral responsibility and effects that often get sidelined by the tug-of-war between the citizenist and due process-defending perspectives.
In addition to the many inline links in the post, the following might be of interest to readers:
- Secure the US-Mexico border: open it by John Lee, Open Borders: The Case, February 25, 2013.
- Open borders, crime and targeting the natives: the case of South Africa by Grieve Chelwa, Open Borders: The Case, May 3, 2013.
- Worried about Hispanic immigration to the US? Your worst fears have already come true by John Lee, Open Borders: The Case, February 19, 2014.
- Deporting ‘Felons, Not Families’ Obama’s immigration plan has no room for criminals. But what’s a criminal? by Christie Thompson, The Marshall Project, November 21, 2014.
The photograph featured at the top of this post depicts police personnel at a 2006 march for immigrant rights in Los Angeles, California. Photograph by Jonathan McIntosh and licensed under the Creative Commons Attribution licence.