Bryan Caplan  

Why Is Illegal Immigration So Low?

Tim Smeeding on Inequality... Thaler and Caplan on Homo Econ...
Contrary to popular opinion, U.S. immigration enforcement is draconian.  The proof is in the prices illegal immigrants pay to human smugglers.  Even Mexicans, our closest Third World neighbors, pay 2-3 years' worth of wages to illegally cross the border.  Every other Third World nationality pays more.  Millions of immigrants come illegally not because the cost is low but because the benefits are high.

Even so, observed levels of immigration immigration are puzzlingly low.  Let's stick with Mexico for simplicity.  $3000 is a lot for a Mexican farm worker in Mexico, but not so much for a Mexican farm worker in the U.S.  Suppose annual farm pay in Mexico is $1500, versus $15,000 in the U.S.  That's a raise of $13,500 a year.  Once he arrives, a Mexican worker could recoup his smuggling fee in under 12 weeks.  Why then do most stay home?

True, these numbers ignore the horrors of the border crossing.  They are bad, but still seem modest compared to the magnitude of the gain.  A 0.1% risk of death is a pretty high estimate.  According to standard calculations, pampered Americans would only pay $7000 or so to avoid that risk.  For would-be Mexican immigrants, even $1000 seems high.

Another major flaw in these calculations is that they assume that paying a human smuggler gets you into the U.S. for sure.  In reality, the success rate is around 50%.  The expected cost of crossing into the United States is therefore roughly double what it seems.

But neither of these sensible adjustments do much to resolve the puzzle.  Add $1000 danger cost to the $3000 out of pocket cost.  Then divide the sum by the probability of successful crossing. It implies that, on average, Mexicans recoup their full smuggling costs in 31 weeks.  That's still a great return on investment - 167% per year.

What other factors are at work?  One story I've heard is that illegal immigrants' expected gain is much lower than it seems because they only earn American wages if employed.  But this seems a trivial factor.  Illegal immigrants' unemployment risk is only slightly higher than natives'.  And don't forget there's unemployment risk back in Mexico, too.

George Borjas suggests that people's strong attachment to their homeland deters migration.  He's right about the general phenomenon, but grossly exaggerates the magnitude.  In any case, feelings of attachment vary widely, and current immigration is far below potential.  So you would still expect the marginal illegal immigrant's attachment to Mexico to be mild, leaving the puzzle intact.

The key factor, in my view, is quite different: Illegal immigration is relatively low because would-be immigrants have crummy credit and insurance options.

1. Crummy credit options.  Most rural Mexicans can't just go to the bank to get an illegal immigration loan.  Nor can they pay a coyote with a credit card.  They start in a desperate situation with a lousy credit rating.  To cross the border, they have to save years' worth of their own income, borrow years' worth of income from similarly desperate relatives and friends, or pay frighteningly high black market rates.  When economists invoke such arguments to explain Americans' behavior, I'm skeptical.  For poor people in developing countries, though, credit constraints are clearly a big deal.

What behavior economists call "debt aversion" amplifies the credit problem.  Most human beings dislike "being in debt," even when the debt quickly pays for itself.

2. Crummy insurance options.  Rural Mexicans can't readily buy "illegal immigration insurance."  So when they finally accumulate their nest egg to cross the border, they're risking their life savings.  50% chance of crossing the border and entering the land of plenty, 50% of losing their nest egg and getting sent back to Mexico.  A terrifying gamble.  The black market could offer such insurance, of course, but would-be customers are right to worry they'll never get to collect.

Various choice-under-uncertainty anomalies probably amplify the insurance problem.

And that's the heart of my story.  In a free market, of course, poor Mexicans would have access to reputable international lenders and insurers.  But neither the U.S. nor the Mexican government would tolerate major First World corporations financing illegal migration.

I hope no one will mistake my analytical approach for lack of sympathy with illegal immigrants' plight.  In a just world, they could enter the land of opportunity for the price of a bus ticket.  My point is that relatively low levels of immigration do not show that the seemingly immense gains of border crossing are illusory.  The gains are genuine.  They're just hard for people in desperate poverty to realize with only the black market to help them.

HT: Partly inspired by a Facebook exchange with Bill Dickens, and Alex Nowrasteh's request.

COMMENTS (10 to date)
CMOT writes:

"In a free market, of course, poor Mexicans would have access to reputable international lenders and insurers."

Native born US citizens DO have that access, yet payday loan stores charging 400% (or even 500%) interest are seemingly everywhere.

Since illegals can melt away into the population in ways that citizens can not, they would likely be even worse credit risks than the current legal citizen payday loan customers, facing even higher interest rates.

John Thacker writes:

Concerns about money laundering (and federal laws and regulations that make banks responsible for their customers' actions) are causing large banks to close branches on the Mexican border, which certainly makes it harder for both immigrants as well as current citizens to get banking and credit.

Peter H writes:

Also keep in mind purchasing power parity adjustment. $3000 USD buys you more in Mexico than in the US. So the $15,000/yr income in the US, while bigger, isn't quite 5x bigger.

BC writes:

How do the gains from illegal immigration compare to the drug trade and other organized crime? If the gains are comparable, then why doesn't organized crime get in the business of financing illegal immigration (for a share of the gains, of course)? To some extent, the high prices paid to human smugglers reflects some of this value but, as Bryan's analysis shows, the smugglers do not capture all or even most of the gains. Would organized crime have difficulty collecting payments from illegal immigrants once they enter the country and start working?

It's a sad state when we have to rely on organized crime to fulfill valuable social functions like providing credit for would be immigrants, but I guess that was also the case during Prohibition.

Anon writes:

As Peter H implies, this post is worthless without purchasing power adjustments. Not only do you have to factor in cost of living to get to a "net" wage after expected costs, you have to also determine what that net amount can then purchase.

NZ writes:


Organized crime does finance illegal immigration, but they do it by force: they extort would-be immigrants (legal and illegal) into being drug mules. If the mules cross the border and deliver their payload without getting caught, they get to stay in the US and live. If they get caught, the organized criminals never heard of'em. If they refuse, they get shot.

I had a collected bunch of news articles about this back in the Google Reader days, but I don't know where those articles are now. I'll let you Google for it, but I promise this is a real thing.

Also, I would guess that people in the human smuggling business don't stick to humans exclusively. Drugs are a lot easier to smuggle than people.

_NL writes:

Bryan: I'd also comment that the 50% figure applies to crossings, but doesn't figure in the rate of being caught after crossing (e.g. ICE raids on your employer), or the risk that the coyote will either abandon you in the desert or not show up after being paid. My understanding is that coyotes are prone to run at the first sign of border patrol, leaving the authorities to scoop up the slower migrants ("I don't need to outrun the bear..."). It's a marginal difference but it's part of the calculation nonetheless - especially if people are prone to overestimate risks versus benefits.

BC - My understanding is that, according to law enforcement, some people do finance their immigration into Western countries by acting as drug couriers. That way, you only need one delivery rather than some system of payroll garnishment.

There is no legally secured loan for smuggling migrants, a Bryan said. Relying on a future stream of payments from the worker being smuggled generally requires having leverage over them - either keeping their movements controlled or keeping their family hostage. Neither one of those is terribly cheap and the criminal penalties for both can be higher than smuggling along (it also requires fiercer employees who are ruthless about keeping indentured servants or hostages).

Since the classic migrant worker model is "go anywhere, do anything" by waiting at a designated area for a farmer, a contractor, whatever, it's terrible for monitoring the physical location of your human assets. So this only works if the migrant is doing work in a business you operate and can therefore physically control. Classically, this is sex trafficking, where the smuggler either operates the brothel or sells the indentures to the brothel. It could work for other illegal operations, too, like a meth factory or something else in a contained area with unpleasant work to do.

Of course, workers are generally aware that an indenture with hardened organized criminals is likely to turn into a much longer and more unpleasant indenture than you imagined, and maybe even more like chattel slavery. So the migrants prefer to pay rather than sign up to live under the thumb of criminals for years.

The typical model is you save up and pay cash. This is a lot easier if US relatives can send you money. But it's also plausible that you pay for your smuggling at the time, by muling some illicit substances. The main problem for the criminals is that you're inexperienced and likely to fail, drop the drugs, or get caught. They also cannot monitor you easily, so they may need to send along their own people to supervise. The main problem for the migrants is that immigration violations are punished with detention and deportation, but drug crimes are sentenced with much longer incarceration followed by imprisonment. The cartels are also dangerous, so association with them may not be advisable.

As a result, with the lack of either easy credit or enforceable loans, migrants are likely to save up and pay cash down.

Mr. Econotarian writes:

Any adult individual who wants in to the US from Mexico can get here. The coyote takes you into the desert, 10 of you run in different directions, and border patrol can't get all of them. If they get you, you get deported. Then you try again, perhaps with the same coyote. In 3-4 tries, you are in the US. You get pretty beaten up from scurrying through the brush at night, but the risk of death is not too high.

The truth is that the current labor market in the US is pretty bad, and Mexico is becoming a less worse place than it was say 20 yeasts ago. Fewer laborers are coming to the US.

Besides the desire for higher pay, there is a real desire to have your kids be born in the US as citizens and enjoy 12 years of higher quality free education and have a chance at a life their parents never could have had. Our public schools may seem bad to us, but at least our teachers don't blockade or break into electoral offices to avoid reforms.

NZ writes:

@Mr. Econotarian:

Don't forget the free emergency care, mostly-corruption-free police protection, and myriad of free programs to help illegal immigrants get food, housing, and other services. Hey, your kids might even get a major boost in help with their college tuition and a major lowering of the admission bar if they write an essay about how their parents had to climb over a fence of oppression to have them here.

(Of course, many of those kids wind up being more anti-American and/or more gang-affiliated than their parents, but let's not notice things too closely. What counts is that if they vote, they're almost guaranteed to vote Democrat.)

Finch writes:

It would be nice to see this analysis done on domestic moves of US citizens. My understanding is that people move from California to Texas a lot less than we would expect based on economic reasoning as well. Maybe there is more to the attachment argument than you think.

But I'd guess it's not some patriotic love-of-country; I'd guess it's social capital that you have to abandon when you move and may never earn in your new digs. That's going to be very hard to value.

Comments for this entry have been closed
Return to top