Bryan Caplan  

Of Infants and Immigrants

Reading, Not Recommended... Blameworthy: How Party Loyalty...
If an additional infant born in America has positive externalities, asks Adam Ozimek, shouldn't an additional immigrant who moves to America have positive externalities, too?
Since the average immigrant age is around 30, that means when they've arrived they are already past the stage when they just consume society's resources, and have begin producing externalities. Doesn't this suggest that positive externality of immigrants is even larger than that for natives? If you're going to claim that the average lower education level of immigrants reduces their externalities, keep in mind that immigrants are also more likely to be small business owners and PhDs than natives.
Another objection is that immigration creates positive externalities for the country of destination, but negative externalities for the country of origin.  But when you remember remittances, the objection falls flat.  Workers do a lot more for the Third World by leaving than staying.

COMMENTS (20 to date)
josh writes:

Are you joking?

It depends on the immigrant and it depends on the native; and yes, there are ways to predict which new addition will be a ((n) economic) boon and which will be a ((n) economic) burden.

Apparently you are also unaware that their exists this thing called culture, and that sometimes (all the time?) new cultures introduced into the midst of existing ones can cause conflict. Do you consider ethnic cleansing of urban neighborhoods a positive or negative externality. What do you think happened to all the blacks in Compton? That they moved on up to the East Side?

[Comment edited with permission--Econlib Ed.]

Prakhar Goel writes:

Josh is correct. Your post is beyond ludicrous. Bryan, we know that you like immigration but surely you are still above blatant propaganda.

Most people when talking about immigration distinguish between legal immigration much of which comes from Asia and illegal immigration much of which comes from Mexico. Shouldn't that be taken into account before you start giving out fractions getting PhDs? Besides that is just the beginning. Immigrants from Asia and immigrants from Mexico, legal or not, are two completely distinct populations. To come from Asia, somebody needs to pay for a plane ticket resulting in a highly self-selected set of immigrants pretty much guaranteed to have above average SES. In contrast, to come from Mexico, you usually have to know somebody who can cross borders undetected --- a vastly different skills requirements. This does not even account for the cultural differences between Mexico vs. China and India and the implications for the political system. What happened to this post?

Yancey Ward writes:

Even if remittances counter the objection, should this make a difference on whether you allow an immigrant in?

dave smith writes:

I would expect that the biggest externality of an immigrant would be that he expands the division of labor in the destination country.

I would not expect a symmetric impact (negative externality) in the departing country because if there were a symmetric impact the benefit to the person immigrating would be much less and the person would not immigrate.

JPIrving writes:

It depends on the immigrant.

North Italians have higher IQs than Southern Italians. Argentinians have higher IQs than Mexicans. see where I am going.

In modern America the talented pay the taxes and keep the ship going. .

Evan writes:

josh, Prakhar Goel, you have to remember that human are programmed by natural selection to underestimate the benefits of interactions with foreigners. For that reason, when we are given evidence of negative behavior by them, we give it far more weight than we should.

Fortunately there is a cognitive technique called "The Outside View" that helps to correct this. When using the Outside view to predict the impact of immigration, what you do is instead of looking at the characteristics of current immigrant groups and trying to predict from that, you look at what people have predicted about past immigrant groups and see how well their predictions have turned out. This let's you compensate for cognitive biases by giving less weight to information that is vulnerable to them.

Using the Outside View, I can tell you that in the past people have said the same thing about Irish, Italians, Poles, Chinese, Japanese, and other immigrant groups that todays critics are saying about Hispanics. In fact, I am an avid reader of old books, and I can tell you that some novels I read from the 20s and 30s make the exact same arguments about Italian and Eastern European immigrants that critics of Hispanic immigration are making today (including the ones about IQ). Critics of immigration were dead wrong about those groups in the past, America has benefitted greatly from their coming here. So it is quite likely critics of immigration are dead wrong in the present too.

A couple of other points:

Prakhar Goel, you mention that you are worried about the effects of Hispanics on the political system. The fact that you are worried indicates to me that you do not actually know what the voting participation rates of Hispanics are. I happened to hear a news story the other day that showed that Hispanics are the least politcally active group in the country. I hope that assuages your fears.

josh, gangs exist because they can feed off of protecting various illegal victimless crimes If you want to stop gangs you should campaign to legalize drugs, prostitution, and gambling, not throw the baby out with the bathwater by keeping out all Hispanics becaue 1% of them might be in gangs.

JPIrving, in modern America the less talented are forced to pay taxes to give the talented free money. The majority of social programs benefit the middle class, even though the tax dollars come from both the poor and the middle class.

Also, the majority of race and IQ differences can be explained as easily by culture as by genetics.

floccina writes:

Although I think that allowing immigration is the right thing to do I understand josh and Prakhar Goel's concerns. I also think that realistically there is little Government in the USA can do to slow Mexican and central American immigration. I do think that something is happening on its own that will slow Mexican and central American immigration that is per capita income growth in Mexican and Central American. I think when percapita incoem reaches some level well below the level of the USA people just as soon stay where they are. Consider Italy emigration from Italy to the USA slowed well before they reached the standard of living of the USA.

On thing that can mitigate a Mexican problem (I should be honest and call it what it is, a mestizo problem) would be to allow much more immigration from China, India and Poland etc..

I would also say to josh and Prakhar Goel's things change in surprising ways over time so I would not worry too much. I have learned to love immigration

[Broken link fixed. Please check your links, especially when they are to your own blog!--Econlib Ed.]

Two Things writes:

You're kidding, right? Rather than cite any data about immigrants, Omizek reasons from an opinion survey of bloggers (???!!) that immigrants must produce positive externalities. ROFL! What a joke!

The average (and modal) immigrant produces a stiff negative externality at the state/local level and at the federal level.

The last claim you quote from Ozimek is genuinely dishonest; you should feel shame for quoting it.

He writes: "If you're going to claim that the average lower education level of immigrants reduces their externalities, keep in mind that immigrants are also more likely to be small business owners and PhDs than natives."

("Reduces their externalities?" Yeah, reduces them below zero, as we have already seen!)

Possibly a greater proportion of immigrants than natives have PhD's (most are foreign students who stay in the US after earning their PhD's here). But hardly anyone-- native or immigrant-- has a PhD.

It much more relevant (and honest) to note that immigrants are more likely to be poor than natives. Most immigrants are low-skill, low-wage workers and their offspring are mostly dropouts and criminals. The children and grandchildren of Mexican immigrants have the lowest high-school completion rate of any measured group!

Neither you nor Omizek seem willing to embrace a rational immigration policy of admitting only high-skilled immigrants (i.e., those who really will produce positive externalities) and excluding low-skilled immigrants-- whom we know empirically produce severe negative externalities. Why not? If we're going to have some policy, why not choose one which maximizes our welfare?

Even if you're a couple of closet self-flagellators, why do you want to drag the rest of us down?

Two Things writes:


I got a good chuckle out of the way you chided josh and Prakhar Goel. You wrote:

"josh, Prakhar Goel, you have to remember that human are programmed by natural selection to underestimate the benefits of interactions with foreigners."

Your argument proves the exact opposite of what you want it to.

Natural selection is unlikely to program any creature to "underestimate the benefits" of anything. If there were benefits to importing foreigners to the local ecosystem, those who reaped them would have left more offspring than those who didn't and humans would be mostly xenophilic now.

If humans are "naturally" xenophobic now, that strongly suggests that xenophobes reproduce more efficiently than xenophiles.

Which means, by your logic, that immigration opponents are right.

(In case you care, I don't think the average American today is especially xenophobic. I think natural selection has left humanity with a much more complex set of emotional and rational mechanisms for dealing with other humans than you seem to suggest.)

Homework assignment: Was immigration from Europe good for the pre-Columbian inhabitants of the Americas? Discuss.

Peter Finch writes:

Two Things, as was pointed out the last time this came up, those estimates are cost-to-government, not cost-to-society.

If the households in question earn more than their subsidy costs to provide, then they are net contributors. Summing the federal, state, and local numbers you provided, that's $11K in outlays, which presumably costs somewhere north of $11k to provide.

So if the households in question are making $20k, it's pretty certainly a win, if they are making $15k, it's close, and if they are making less than $11k, it's probably a loss, assuming the numbers you linked to are correct.

Two Things writes:

Peter Finch:

Your arithmetic is flawed. The typical low-skilled worker earns a wage close to his marginal product.*

You have to compare his marginal product (which is only slightly above his wage) to the SUM of his wage and taxpayer-provided subsidies.

So if his wage is $20K, his marginal product is perhaps $24K. Wage + subsidies = $31K. Net loss to "society" (taxpaying citizens) $7K.

And that's leaving out harms other than government subsidies, such as costs of crime-- and ignoring transaction costs (for the government to give $1 to an immigrant, taxpayers must give something like $1.50 to the government-- the excess goes for bureaucracy plus waste, fraud, and abuse).

*For a given business, on average employees' marginal products can only exceed their wages by the overall profit of the business, commonly less than 15%. Government workers are paid far above their marginal products. Some investment bankers and other successful rent-seekers are paid above their marginal products as well.

RP writes:

@Two Things,
Your argument about the efficiency of natural selection depends (among other things) on the evolutionary heritage being a good approximation of modern life. But human genes are designed for a zero-sum hunter-gather environment (or negative-sum due to violence), whereas we live in a positive-sum production environment with large gains from trade. That suggests that natural selection produces impulses that are systematically too xenophobic (whether or not the impulses are in fact xenophobic).

I should also note that people who argue that low-wage immigration implies (pecuniary [aka "fake"]) externalities in the US seem to have their reasoning backwards. Standard economic theory says that, in the short run, per capita native income increases precisely when immigrants come with relatively less skill, not when they have the same average skill as the native population. (In the long run, of course, there is no effect either way unless you're living in an increasing-returns-to-scale economy: immigration pushes up returns to physical & human capital, encouraging their acquisition, until in a per capita sense there is no long run effect.) This is straight-up Borjas, btw.

Peter Finch writes:

I think your new formulation is closer to correct.

I tried to describe the cost of subsidy being greater than the actual amount of subsidy in my original post.

I do note that there are positive non-monetary benefits as well, such as, for example, happier citizen relatives.

One should also question how long the immigrant household will stay low-skill and low-wage. Will his kids make $20K forever?

And we are both valuing the immigrant household's utility at zero, which not everyone will agree with.

But doing the calculation this simplified way means it's a slam-dunk positive for the engineer from India, and close, but probably negative for the guy who's going to wind up as a day-laborer and brings his kids. If we organized our immigration policy around this calculation, we'd presumably have a lot more immigration, although of a different composition. The tax and subsidy scheme certainly throws a monkey-wrench in naive analysis.

Peter Finch writes:

I misread Two Things citations. The total cost is $6,500 per household, not $11k, according to those sources.

That changes the result, but not the accounting.

Steve Sailer writes:

"If an additional infant born in America has positive externalities, asks Adam Ozimek, shouldn't an additional immigrant who moves to America have positive externalities, too?"

That's what's known as engaging in Conspicuous Assumption.

Kurbla writes:

You finally addressed real problem with emigration - the influence on the source country. But you're too easy on claim that "workers do much more for Third world by leaving than staying." It requires more calculation. Remittances are small; $1500 per cap/year is less than externalities (PPP) average immigrants generate at home in all except the poorest countries. Definitely less than externalities generated by, say, average Mexican worker in Mexico. Actually, it might be even that remittances have negative externalities:

Scot writes:

Can we except this as evidence that babies, immigrants, and guys on mushrooms all produce externalities that are roughly comparable?

shecky writes:

Why the concern over a source country? People in a failing country should not be sentenced to remain any more than you should be forced to remain in a town where you cannot earn a living. True concern over a source country might actually be to encourage migration of it's labor to places more able to utilizing it.

Kurbla writes:

Why the concern over a source country? Because if some people leave source country, others who live there might be worse off. For (extreme) example, if you allow single medical doctor from Tanzania to immigrate, outcome could be - death of 1000 children in Tanzania. You have to take that into consideration.

MikeP writes:

Because not only must we enforce laws that abrogate individuals' rights to migrate and consign them to poverty due to a condition of their birth, but we must enslave them to the populace of their place of birth too.

Comments for this entry have been closed
Return to top