Bryan Caplan  

Mea Culpa: How I Succumbed to Anti-Foreign Bias

I'll Defend IQ... Amateur Bank Lending...

A high fraction of immigrants are young, low-skilled, Hispanic males. Given these demographics, I long assumed that immigrants would have relatively high crime rates. While I kept this problem in perspective, I took it for granted that increased crime was a genuine drawback of immigration.

I was wrong.

A fascinating NBER Working Paper (earlier, free version here) by Kristin Butcher and Anne Piehl shows that, despite their demographics, immigrants are drastically less criminal than native-born Americans. In fact, immigrants have one-fifth the incarceration rate of natives. Yes, natives are incarcerated at five times the rate of the foreign-born:

Using the 1980, 1990, and 2000 Censuses, we show that 18-40 year-old male immigrants have lower institutionalization rates than the native born in each year. The gap in these institutionalization rates widens over the decades, and by 2000 immigrants have institutionalization rates that are one-fifth of the native born.
Is this base statistical trickery? Not likely; these are raw numbers that anyone can double-check against the census. Could these results simply reflect the practice of deporting criminal aliens? Nope; our actual practice is to make immigrants serve their full sentence before expelling them.

But how can we reconcile the facts with the demographics? Butcher and Piehl show that given their demographics, we should expect immigrants to commit crimes at double the native rate. But for some reason(s), demographics yield a massive overprediction; immigrants commit crimes at one-tenth the expected rate given their demographics. Yes, if immigrants acted like otherwise similar natives, they were be ten times as criminal as they actually are.

Why would this be so? The main theories that Butcher and Piehl explore: (1) Immigrants self-select for law-abiding behavior; (2) Immigrants are more responsive to deterrence than natives. Personally, the most plausible story to me is that immigrants feel lucky to be in the U.S. and don't want to forfeit their "big break," so they try extra-hard to stay out of trouble.

In any case, given my repeated attacks on anti-foreign bias, I should have been more skeptical of popular claims about immigrants' criminal propensities. So to every immigrant out there, I'm sorry. I'll try not to let it happen again.

P.S. I won't be posting much during the next two weeks. Why? Well, let's just say that if you see me at Comic-Con, don't hesitate to say hi!

Comments and Sharing

TRACKBACKS (12 to date)
TrackBack URL:
The author at Trade Diversion in a related article titled Immigrant Criminals writes:
    A new NBER paper shreds an old stereotype, reports Bryan Caplan: Kristin Butcher and Anne Piehl shows that, despite their demographics, immigrants are drastically less criminal than native-born Americans... [G]iven their demographics, we should expect ... [Tracked on July 18, 2007 3:57 PM]
The author at MaxSpeak, You Listen! in a related article titled A GOOD WORD FOR BRYAN CAPLAN writes:
    For this post. (On a study finding immigrants much less likely to commit crimes than residents born in the U.S.A.)... [Tracked on July 19, 2007 10:29 AM]
The author at amcgltd in a related article titled Whodathunkit? writes:
    It would seem recent immigrants are actually ten times less likely to commit a crime than the native-born. This is so counter-intuitive I'm having a bit of trouble getting my head around it. However, Bryan is an economics professor.... [Tracked on July 19, 2007 2:56 PM]
The author at 40 Years in The Desert in a related article titled This is How You Apologize if You are Mensch writes:
    Brian Caplan offers an apology. ... Let's be clear, the error here is minor, as he is pro large scale immigration, he was arguing for this in spite of what he thought were crime issues, which is far more benign than what Lou Dobbs done. I te... [Tracked on July 21, 2007 10:28 PM]
COMMENTS (28 to date)
Matthew c writes:

I was wondering when you would get around to that study. . . Interesting stuff for sure.

caveat bettor writes:

Immigrants have 3 distinct disadvantages relative to native-born Americans: 1) they don't know the odds of getting caught as well, 2) they don't know the system well enough to get off on charges as easily, and 3) they don't get preferential treatment in the system because they don't have friends who have friends in high places

ed writes:

I think the reason people believe (wrongly) that Hispanic immigrants have high crime rates is that the children and grandchildren of Hispanic immigrants do have high crime rates.

That is also the reason why illegal immigration could increase overall crime, even though the original immigrants themselves are quite law abiding.

Biomed Tim writes:

Someone should forward this to Thomas Sowell.

randy writes:

maybe it is the way it is because these immigrants are so culturally secluded as to avoid getting caught. if your detectives can't interview people (due to a language barrier) to learn more about a crime and the perp, i would imagine it becomes quite difficult to apprehend the criminal.

maniakes writes:

Both hypotheses strike me as plausible. I'd like to add four more:

1. Immigrants who were middle-class in their (poor) country of origin likely have middle class values which predispose them to be more law abiding than native-born poor. (This assumes that part of the link between poverty and crime is partly because the same things that make someone likely to become poor (poor impulse control, high time preference, low value on hard work and personal responsibility, etc) also make one more disposed to crime).

2. Immigrants from poor countries feel affluent even at what in the US is a low relative income level, and thus act affluent. (This assumes that part of the link between poverty and crime is partly due to a sense of desparation, a feeling right or wrong that crime is the best avenue towards improving a bad personal situation).

3. Immigrants who are caught committing crimes are often deported instead of incarcerated. (This should be relatively easy to check, by adding deportation-for-criminal-action numbers to the incarceration rate).

4. Due to a well developed support structure to evade detection for violating immigration laws, it may be sustantially harder to catch criminals who are illegal immigrants. (This naturally only applies to illegal immigrants).

guy in the veal calf office writes:

How do the statistics deal with illegal immigrants? By definition 100% of them are committing a crime.

Karl Smith writes:

My guess it that the penalty is simply much higher.

If you are an immigrant and you are caught for even a small infraction I would think that you are deported. Moreover, I would suppose that risk having your buddies deported as well.

Therefore, there is both a private and a social cost to committing crime that natives do not have. When something is more expensive we tend to do less of it.

Squid writes:

Maybe the more important statistic to look at is the amount of crime that children of immigrants commit. This could be huge because these children are not taken care of while their parents are working long hours. Also, these children are likely to live in poor urban areas where crime is more common.

shecky writes:

Even the crime rates of immigrant children may be surprisingly low. It could have changed, but ISTR that crime rates here in Los Angeles, a place that's been a center of new immigrant population for a long time, are lower now than when I was a kid in the 70s.

Lord writes:

I don't doubt legal immigrants fear deportation. OTOH, 10% of arrests in Orange County are illegals, which doesn't strike me as low at all.

Leif writes:

ed, do you have some statistics about that? I'm not trying to be a nuisance I'm genuinely interested in seeing if their children commit crimes as rates sig. higher than the average. thanks.

Steve Sailer writes:

First of all, the paper is ignoring the obvious distinction between legal and illegal immigrants. Second, what it's telling us is that the illegal immigrants' sons and grandsons have very high crime rates. That is _not_ an argument in your favor!

ed writes:

I don't have any very good statistics on crime rates for children of immigrants. According to the Bureau of Justice Statistics

On June 30, 2006, an estimated 4.8% of black men were in prison or jail, compared to 1.9% of Hispanic men and 0.7% of white men.

That leads me to believe that second and third generation Hispanic immigrants have higher crime rates than whites. I too would like to see some better statistics.

TGGP writes:

ed, your suspicions are right. Crime increases from first to second generation and from second to third. Stats here.

Another major factor in why first generation has low rates is that crime is a young man's game, whereas the immigrants often already have families and go north for employment, both of which "grounds" them away from committing crimes or joining gangs.

Patri Friedman writes:

I won't be at comic-con, but I'll be in San Diego that weekend, if you have a social dinner let me know.

conchis writes:

Deporting immigrant offenders after they've served their sentences would still pull the numbers down. If a significant proportion of those incarcerated in the native population are recidivists, and the immigrant population would have had the same recidivism rate if offenders weren't deported, then you've just removed a significant chunk of those likely to offend, and thus reduce the incarceration rate, even if the underlying propensity to offend isn't that different.

Of course, this wouldn't change the fact that fewer immigrants are committing crimes. It simply illustrates that immigrant crime is in one sense easier to deal with than native crime.

P.S. If you're primarily worried about the effect on natives, then you'll probably want to break crimes down by their victims too. If most of the incarceration is, say, drug-use related, or immigrant-on-immigrant crime, then you probably shouldn't think of that as a big problem with immigration, except insofar as it ties up police resources.

Mr. Econotarian writes:

I'm not surprised by this at all.

El Salvador's experience was that after the civil war (which involved a lot of "official" violence), lots of people moved back to the country from the U.S. Unfortunately, their kids who grew up in the U.S. (in cities like LA) picked up on the gang culture and brought it back, spawning gangs like MS-13, which now has made it back into the U.S.

Today, almost as many people die in gang violence in El Salvador as did during the "official violence" of the civil war.

Many of the deported immigrants are youth and young adults who grew up in U.S. cities with hugely active gang cultures. "We're seeing the deportation of all these young men whose families fled to the U.S. during the wars or economic crises of the 1980s and 90s," notes Geoff Thale, senior associate at the Washington Office on Latin America. "They grew up in immigrant neighborhoods in L.A. and Chicago, in gangs like Salvatrucha and Diez y Ocho. These are U.S.–based gangs that are exported to Central America, where the men show up culturally disoriented but much more sophisticated in criminal activity."

Also see:

David in NY writes:

I think people should consider why the children of immigrants might have higher crime rates (assuming that to be true), and what this means with respect to the remedy for the situation. I believe it to be a result of housing and school segregation. Because of pervasive segregation of housing on racial and ethnic grounds, the children even of law-abiding immigrants are essentially forced to grow up in neighborhoods in which it is extremely easy (and there are social incentives) to learn crime. Econotarian's comment above supports this. So the children do turn to crime, more than their parents. [Basic notion lifted from James Q. Wilson's Thinking About Crime.]

Unfortunately, this country has absolutely no interest in integrated housing (on either class or racial bases), or, for that matter, in improving conditions in segregated neighborhoods that make crime attractive to youngsters (conditions like, say, low employment possibilities and inferior schools).

In any event, it would seem a shame to stop people coming here because some of their children might turn to crime because of conditions this country does nothing to ameliorate.

El Presidente writes:

David, actually there is a practice in Los Angeles called inclusionary zoning that attempts to integrate people by floorplan. The size and price of the dwellings in new construction have to be an amalgamation of sorts. It isn’t ubiquitous but it is growing.

The veil of ignorance concept is repugnant to capitalism, at the individual level, because if you had the choice between two people you could compare their characteristics to trends and make stereotyped judgments that might be effective in selecting the best candidate, at least in the short run. It is a way to prejudge an outcome and where society expresses a strong bias, a prudent businessman does also. This is where Sowell points out that the practice of segregation/slavery is economically disadvantageous to society as a whole though individuals might think it advantageous to them, personally.

For the sake of argument, California has a higher incarceration rate than other states. Does this mean Californians are inherently bad, or those in say Wisconsin are inherently good? Incidence of arrest/conviction does not tell us accurately about incidence of crime. Nor does it tell us what motivates criminal behavior. It has been mentioned that all illegal aliens are criminal by definition. This is a significant factor in their willingness to report crime. Since they are most likely to be victimized by someone they know, and the people they know most are similar to them, they are less likely to report each other. This is what happens in isolated ethnic communities and often between gangs. Retribution becomes their de facto law enforcement. I’m not sure you were entirely wrong about the crime in the first place. I just wouldn’t blame the immigrants so much, and us so little. Think cross-elasticity of the factors of production, cross-border wage differential, balance of trade, farm subsidies, cash crops, black markets, general equilibrium models. I think you get the idea.

The Bracero agricultural guest worker program ended following the unionization efforts of Chavez. In effect we said, “You can have all the rights you want, but you’ll never be able to enforce them.” If we intend to use people, to treat them as less than equal, then perhaps we share some responsibility for the tendency toward crime. Perhaps we do this to our own, and more effectively. I would ask those familiar with this particular research if immigrants’ children become more criminal (or at least get caught more) because they take on the characteristics similar to others who are native born and criminal. If so, why would they do that to themselves in this land of flowing milk and honey? Maybe they don’t do it to themselves and maybe there’s something else going on here.

David in NY writes:

What exactly is your point, El Presidente?

pjgoober writes:

David in NY writes:
"I think people should consider why the children of immigrants might have higher crime rates (assuming that to be true), and what this means with respect to the remedy for the situation. I believe it to be a result of housing and school segregation."

David, non-hispanic whites are 55% of US births and that percentage is steadily dropping (adding in asian kids with the white ones would only slighty change the percentage, and would slighty slow but not halt the percentage decline). The main driver of that change is post 1970 immigrants and thier descendants. That puts a huge damper on what the most comprehensive and complete school integration plan imaginable could concievably accomplish. Making sure that all kids go to majority white schools will soon be mathematically impossible.

If you want the children and grandchildren of immigrants to go to schools with a higher ratio of white kids, then mass immigration is working against that. If your argument is for school integration by economics (not race) then all of this still applies. The well-to-do have less kids than the poor, and mass immigration is helping to steadily increase the ratio of poor children to well-to-do children in the aggregate US school system. Simple arithmatic shows that massive unskilled immigration is utterly killing the integrationist dreams of progressives in the cradle and 99% of them don't even realize it.

This is one very important area in which the current era of mass immigration greatly differs from past eras that is often overlooked.

Biff Spaceman writes:

Illegal immigrants, primarily from Mexico, are self selected achievers who, following a rational weighing of economic facts, take the initiative to improve their situations by going to America, applying their skills and drive, improve their lot financially and provide the backbone for a subculture of like minded hard working people operating under the radar. Often they will live very frugally and send money back to family in Mexico. Mexican workers I worked with in the restaurant business began as illegals and advanced as they learned language and job skills. Some made great efforts to become legalized after the fact and created very clever schemes to achieve this without leaving the country. They paid into someone's social security, paid sales taxes, purchased goods and services, and were better than average citizens because they had a real disincentive.
American born lazies, of all ethnicities, in my experience in the workplace have a sense of entitlement and are able to rationalize away lots of their personal failings, usually blaming them on others. The funniest thing I ever heard (and I still recall today) was a white middle class kid washing dishes who stormed out on a busy night saying "My butt hurts, I'm wet and sticky and this is bullshit!" and out he went. No Mexican, legal or illegal, ever did anything like that.

David in NY writes:

pj goober -- You ignore housing segregation, which is my, and James Q. Wilson's, main point. And you ignore my two possible remedies: 1) actually take steps to end racial barriers to housing integration by making sure that successful immigrants can move out of bad neighborhoods (I know you will be shocked, shocked to be informed that such racial barriers still exist, but testing studies indicate that members of minority groups do not receive full information about available housing units about half the time) and 2) actually paying attention to improving conditions (employment, schooling, alternatives to crime, quality child care, health insurance) for young people in poor and lower-income segregated neighborhoods.

If you wanted to stop crime, and not simply immigration of brown people, you might actually be interested in this stuff. Knowing that it's not the immigration that causes the criminality -- but rather having kids grow up in the terrible neighborhoods we provide for them -- gives us a real tool for dealing with crime. Too bad you're focusing on the wrong problem.

El Presidente writes:

David in NY writes:
What exactly is your point, El Presidente?

My point, exactly, is that whether or not the numbers were objectively wrong, and I question whether we know, the underlying attitude is wrong and pervasive. To say people are good or bad, acceptable or not, on the basis of group statistics is not very valuable. I think that goes without saying. As to how to get people to understand the worth of their fellow human beings, you suggest housing integration. This is good. I suggest mental integration. Any amti-immigrant bias is as foolish as a pro-immigrant bias. It is a bias that defines success in terms of the status quo and us-them. The rationally self-interested person is generally seen as the foundation of economics; homo-economicus. Per game theory we know that so long as there is life left to live, there is no rational self interest that does not seek the well being of others. That would be egotistical self interest, not rational self interest. That perversion is what leads to the housing segregation and immigration biases, among other things. Etzioni writes about bounded rationality. Perhaps this is a polite way of saying we don't always know what's good for us, but chances are if it's bad for "them" it will become bad for "us" eventually. I'd really like to quote Lincoln here, but I'll let it go.

El Presidente writes:

I wonder if we have reliable statistics to tell us what the crime rates are of similar people who stay in their country of origin? Do they have higher or lower propensity toward crime/conviction than when they are here? That might be an interesting comparison.

Dan McIntosh writes:

Before we make a comparison, we should show that the populations are comparable. It's not just about education level or economic class. The people who make it past U.S. border barriers and make lives for themselves must be

*dissatisfied enough to recognize there is someplace better
*ambitious enough to want to want to be here
*brave enough to try to get here
*tough enough to endure the trip
*adaptable enough to learn from failure
*determined enough to try again
*smart enough to succeed

Think of it as natural selection. Most people don't even try. These people are the exceptions. We should expect them to be exceptional.

Dan McIntosh writes:

I should also add that the statistic on incarceration shouldn't be assumed to measure the actual crime rate. It may be telling us that a member of this group--if and when he turns to crime--is less likely to get caught.

Comments for this entry have been closed
Return to top