Bryan Caplan  

Immigration: Seeing is Understanding

CoA Alert... GMU Gratitude...

Survey USA ran a neat survey about immigration late last year. They asked people in all 50 states:

Which of these 2 statements do you agree with more:

One: Immigrants take jobs away from Americans.

Two: Immigrants do jobs that Americans don't want.

One of the main findings took me off guard. I naturally assumed that states with a lot of immigrants would be anti-immigrant. After all, whenever I visit L.A., the complaints about immigration never stop. But it looks like I'm smack in the middle of a biased sample of elderly Angelenos. On average, high-immigration states like California are unusually PRO-immigrant.

To get a little more quantitative, let us define the Immigration Optimism Score to equal the % who gave answer #2 minus the % who gave answer #1. Then regress this Optimism Score on Immigrants as a % of the Population. The result: An extra percentage-point of immigrants increases the Immigration Optimism Score by 1.7 points. That's a lot, and you don't need fancy statistics to see it:

Table 1: Immigration Optimism Score as f(Immigrants/Population)


The simplest interpretation of this result is that people who rarely see an immigrant can easily scapegoat them for everything wrong in the world. Personal experience doesn't get in the way of fantasy. But people who actually see immigrants have trouble escaping the fact that immigrants do hard, dirty jobs that few Americans want - at a realistic wage, anyway.

But there's an obvious objection: Maybe what drives the results is the trivial fact that immigrants are pro-immigration. To address this possibility, I re-calculated the Immigration Optimism Score for non-immigrants, making the extreme assumptions that (a) 100% of all immigrants are pro-immigration (I personally know some counter-examples), and (b) immigrants were as likely to be surveyed as natives. The result: In states with lots of immigrants, even native-born Americans are more pro-immigration. An extra percentage-point of immigrants increases natives' Immigration Optimism Score by .8 points. See for yourself:

Table 2: Immigration Optimism Score of Native-Born Americans as f(Immigrants/Population)


Are there other interpretations? Sure. Maybe more native-born Americans in states with lots of immigrants hide their true opinions. But I doubt that effect is very large. The simplest interpretation of the data is also the best: Direct observation of immigrants leads to more reasonable beliefs about the effects of immigration.

Comments and Sharing

TRACKBACKS (5 to date)
TrackBack URL:
The author at Daniel W. Drezner in a related article titled Immigration round-up writes:
    Matthew Yglesias has some interesting posts and links up on the immigration question. This post takes down Robert Samuelson's recent Newsweek essay on whether Mexican immigrates will assimilate into the United States -- it echoes some of what I wrote... [Tracked on May 19, 2006 9:44 AM]
The author at Pajamas Media in a related article titled Seeing is Understanding writes:
    EconLog comments on a survey showing that pro-immigration favorable opinion is perceptibly higher in states with lots of immigrants.... [Tracked on May 19, 2006 10:09 AM]
The author at The Pink Flamingo Bar Grill in a related article titled Is the assumption that because I want the border controlled that I don't like immigrants? writes:
    What a ridiculous survey. It presumes that those of us who vote that immigrants are a net gain also agree with idea that we should allow illegal immigration to continue. [Tracked on May 19, 2006 1:12 PM]
The author at The National Chronicle in a related article titled Bigotry An Issue In Immigration War writes:
    Study finds that the more exposure to immigrants Americans have, the less likely they are to be anti-immigrant. [Econlog]... [Tracked on May 19, 2006 4:11 PM]
The author at Mike Linksvayer in a related article titled See cosmopolitan, think cosmopolitan writes:
    Brad Templeton has an excellent immigration rant. Following an anecdote about immigrant entrepreneurialism: Being anti-immigrant reminds me of racism, to use an inflamatory term. Racism is the belief that the broad circumstances of a person’s ancestr... [Tracked on May 24, 2006 9:20 PM]
COMMENTS (30 to date)
Taggert Brooks writes:

I absolutely think it is wrong to infer that those who say immigrants do jobs others do not want are therefore pro-immigration.

I have friends and relatives in Phoenix and LA. They mostly tend to the liberal side of things. And they see immigrants doing work they prefer not to do (landscaping, roofing, etc), yet they do not favor immigration.

The reason is they see the demands placed on social services and they worry about decreased access for natives and if they are honest with themselves the higher taxes.

So seeing is understanding really means they understand different reasons for opposing immigration. While I favor increased immigration becuase I understand they do not "steal" jobs I'm also not shouldered with the same social costs of education, health care and the like. The benefits of living in Wisconsin.

Randy S writes:

What you are describing is what occurs when “some individuals move from one country to another.” A phenomenon that may be “controlled politically, restricted, encouraged, planned, or accepted. What the US is experiencing, on the other hand, is not immigration but Migration. Migration is a “natural phenomenon: it happens, and no one can control it.” Migration is an extreme catastrophe, where instead of assimilating into the culture into which a people moves, (as what happens with immigration) an entire population moves into an area and changes the political, cultural, and economic make up of a country or area. This phenomenon has happened many times throughout history and, and it is at work all over he Western hemisphere today.

Independently of what it is called; a country in which 25 to 30% of the population identifies with another country, votes and participate in another countries election and remits most of their savings to another economy cannot be called the a sovereign nation. The immediate economic result of such massive migration would be an erosion of the quality of life, life expectancy, literacy rate and infant mortality of our population among others. Notwithstanding that some poor as a whole may benefit from being poor in an environment where poverty is richness as compared as the areas where they originate. In the long run openness to migration results in a disincentive to the needed structural change in the countries were the migrants come from.

Teller writes:

* Bryan Caplan’s most important contribution to the sicence of economics so far: Showing that people have too little incentives to be well informed opinions, and thus tend to have irrational political opinion, especially when there are emotional/ideological aspects involved.

* Bryan Caplans latest post:

"People say in opinion polls immigration is good, so it must be true!"

The economic intellectual meltdown over immigration continues.

I did what Caplan did on his last graph, but simple adjusted for per capita state income ( The significance of the share of immigrants completely disappears
(p 0.24). R square goes from 0.16 to 0.27. The average state income is however significantly correlated people with being positive towards immigration.

If you want to do this more serious there is plenty of data in the links. 100% opposing immigration is too extreme; the data does however includes detailed opinion by race and ideology. (for example 79% of Hispanics in California believe immigrants do jobs Americans would not).

Dave Milovich writes:

It'd be nice to be able to control for the possibility that disproportionately many people with relatively negative views of immigrants have moved from states with more immigrants to states with fewer immigrants.

Ken Hirsch writes:

Serious question for Bryan: If people in states with lots of immigrants are so positive about them, why do so many of them leave?

"The four states -- New York, California, New Jersey and Illinois -- that lost the most domestic migrants to other states were also four of the six states where two-thirds of the nation's foreign-born reside. Altogether, between 1995 and 2000, those four states gained 2.7 million new arrivals from abroad even as they suffered a net loss of some 2 million residents to other states, with many earlier immigrants now joining the out-migration." (source)

Lots more info at William H. Frey's web site.

I'd like someone to explain to me (with a straight face) how immigration was good for the residents of those states.

Scottynx writes:

Maybe native born americans in high immigration states are more likely to be of post recent (post 1965) immigrant decent, and thus be more likely to hold an opinion favorable to thier immigrating co-ethnics.

John Thacker writes:

Ugh. Pretty bad graph and statistics there, Bryan.

It's especially subject to the outlier effect, what with one point out there at 28%.

What is the correlation coefficient, anyway? Surely you should at least give that.

I'm pro-immigration, but I'm also against shoddy statistics presentation.

Matt writes:

I agree with John Thacker. This result needs better analysis.

It is no doubt true that packing all the North American squirrels into a few North American mega cities results in more efficient production of acorns.

What we see in the immigration debate is that a lot of the North American squirrels pay a disproportiante share of the fixed and variable expenses associated with squirrel migration.

What is the change in my net return vs your return as we adjust the population of squirrels? If Waltmart executives pay 30% of the government expenses, but garner 80% of the accelerated net, but Wal Mat clerks pay 50% of the expenses but garner only 20% of the accelerated net, then we have a problem.

25-30% of the fixed and variables costs associated with squirrel migration are covered by government taxes. If this squirrel migration results in a shift of corporate expenses to government, then we would want the hold the differential changes resulting from this shift to be equal across all income sectors.

William writes:


---others. Notwithstanding that some poor as a whole may benefit from being poor in an environment where poverty is richness as compared as the areas where they originate. In the long run openness to migration results in a disincentive to the needed structural change in the countries were the migrants come from. ----

Well said. We forgot to mention what fiscal havoc (double digit inflation) all of the remittances to Mexico create. Vincente Fox is complaining about proposed restrictions of Mexican migration to the US.

Could anyone imagine, if President Bush complained to the Canadian government if Americans flooded over the US/Canadian border, about proposed border controls.


Jason Malloy writes:

This a trite analysis for a number of seemingly obvious reasons, but let's even assume it's true:

Nobody wants to live in high crime neighborhoods, or send their kids to dangerous schools with low test scores. This is why we see "white flight". People who advocate more high-crime/low test-score immigration into America but flee or avoid the resulting neighborhoods and schools in their personal lives do not want this kind of immigration, regardless of what they say or think they want.

This is called a "revealed preference", correct Dr. Caplan? The revealed preference is for law-abiding, economically capable co-citizens, which the controversial illegal aliens being alluded to are objectively not (by definition even!). And most likely won't be for at least several more generations (though more likely never).

Now let's create another survey asking Americans if they want a new, huge, permanent, ethnically distinct, economically dependent, crime-prone, ever-expanding, aggrieved, entitled, socialist underclass. I wonder what they'd say? Because that's exactly what they're getting, but a lot of foolish, short-sighted economists seem eager to usher it in.

Jason Malloy writes:

NB: those are statistical attributes compared to majority norms, btw. Dependency and success are not limited to any group.

Al writes:

I think Bryan Caplan has the causation going in the wrong direction.

People don't become pro-immigration because they see lots of immigrants. Instead, lots of immigrants go to places where people are pro-immigration.

To me, that stands to reason - if you are an immigrant and could choose to go to either (A) a place with lots of people who would welcome you or (B) a place where there are lots of hostile people, where would you go? I would go to (A). Hence, there being a lot more immigrants in (A) than in (B).

[Your EconLog posting privileges have been restricted because you have supplied a non-functioning email address. Email us at if you would like us to consider reinstating your access to EconLog.]

rvman writes:

>Now let's create another survey asking Americans if
>they want a new, huge, permanent, ethnically
>distinct, economically dependent, crime-prone,
>ever-expanding, aggrieved, entitled, socialist

We've already got one of those, they are the Americans not doing the jobs Americans don't want that the illegals are doing. Maybe the next question should be "Do you want 50% higher food prices and 20% higher housing prices?" That's just as loaded a question, just as one sided, and is assuming into evidence as much or less than yours.

Texan here - I've spent most of my life around one or the other of our two 'underclasses'. The illegals, at least, want to be here, and work hard to make a life for themselves. Drive around a Hispanic neighborhood - it is full of immigrant-owned businesses. Then drive around a 'native' poor neighborhood - black or white - it is full of IMMIGRANT-owned businesses. That is capitalist, not socialist. And it is the first step to assimilation. So the signs are in Spanish. They still say "Toma Bud Light". My grocery store has a large Mexican section, to serve immigrants. On it sits 'Mexican' Coca-cola, made with real sugar, alongside the less familiar mango and guava drinks.

We will change, too. Or did you think the Pilgrims celebrated the first Thanksgiving with spaghetti and a Budweiser? We have, however, been quite successful at exporting our own culture and government worldwide - Hollywood, Coca-Cola and Democracy are probably the three most well known 'brands' in the world, and we 'own' all three. Don't lose faith in the durability of 'the American Way' now.

KH writes:

There's also a Southern regional effect.

Teller writes:

1. “work hard to make a life for themselves”

The Mexican employment rate is 56% according to the 2000 Census. 31% of Mexican immigrants use one form of public welfare.

2. “That is capitalist, not socialist.”

60% of Mexicans want to expand government, compared to 35% of whites (Pew Latino survey, 2002).

3. “Or did you think the Pilgrims celebrated the first Thanksgiving with spaghetti and a Budweiser”

That proves nothing. When the Pilgrims came there was no welfare state, no liberal ideology of entitlement based on minority group rights. Up until 1960 80% of Americans were from a few culturally similar countries in northwest Europe.

The evidence so far shows that Mexican immigrants do not assimilate to American levels even after 3-4 generations. Wages after 3 generations is only somewhat above African American ones.

Steve Sailer writes:

If you want to understand why high rates of immigration causes a state to move to the Left, I've been studying it quantitatively for years.

I would recommend you look at my summary article on "Affordable Family Formation," of which Mickey Kaus said:

"Steve Sailer has boiled down the explanation for why some states become red and others become blue to three simple words. ("God" is not one of them.) ... His equation sure works for San Francisco."

You can read it at:

CS writes:

Isn't this exactly what we would expect when:

1. States receiving high numbers of immigrants have high native-born Hispanic populations.
2. Hispanic (Mexican) Americans disproportionately support Hispanic (Mexican) immigration, which is the type at issue in the illegal immigrant debate.
3. Whites have been migrating out of border states like California at enormous rates, increasing the Hispanic proportion of the population further while selectively removing those who are most discomfited by immigration.

Why would it be at all surprising that support for immigration is higher in these states? Why postulate a learning effect before controlling for something as basic as the Hispanic population share?

Teller writes:

Cs Caplan did control for the share of immigrants, in his second graph (perhaps even overly strongly, assuming all immigrants are positive to immigration). The effect is weakened but remains.

But if you also control for state income the effect goes away. Rich feel-good states can afford to be mentally generous to immigrants, while poorer states feel more squeezed.

Regarding the cost of family formation:

I put more weight into Sailers and Huntington’s argument that immigration from Latin nations threatens the foundation of limited government that has made the US rich.

It is amazing to me that Caplan of all people doesn’t not get it. The greatest “natural” resource of the US is the individualist and pro-market Anglo-Saxon ethos of its population. They are less rationally irrational than other nations (ideological consumption of socialism is smaller), have withstood the forces that demand expanded government better, and are now therefore the richest nation on earth.

Free enterprise Classical Liberalism is a Anglo-Saxon invention, and has never gained any strong ideological foothold among the majority anywhere else. The most free economies of the world are ALL culturally Anglo-Saxon (USA, UK, Hong-Kong, Ireland, Canada, Australia, New Zeeland). The only exception really is Singapore, and until a few years ago Switzerland, that is now quickly converging to the rest of Europe.

Leftwing populist Latinos are not assimilating in this regard. They may oppose abortion, but they love big government. By making the US increasingly less WASP, culturally and NOT racially, you are digging the grave of libertarianism.

Less rational voters = bad.

Andrew Leigh writes:

But of course, location is endogenous, so it might just be that people who don't like immigrants move from California to North Dakota. For a neat attempt to circumvent this problem, see Dustmann & Preston in the EJ.

CS writes:

Teller, that's why I said 'NATIVE-BORN Hispanic Populations,' which Bryan did NOT control for, and should reduce the effect even further.

Teller writes:

OK sorry.

There is data in his link with whites, blacks and hispanics seperate, maybe you can use that.

mike writes:

In addition to the fact that this proposition doesn't control for something as simple as the Hispanic-born native population, there is another consideration I've been thinking about: immigration stifles honest discourse on the subject altogether.

Having lived both in the Pacific Northwest and in a heavily Hispanic populated area of Texas for portions of my high school years, I can tell you that you cannot honestly have any sort of public discourse on the consquences of immigration in an area with a high Hispanic population.

This has consequences beyond simply forcing individuals to conceal any anti-immigrant attitudes they may have; it also has the effect of preventing the sort of discourse that would lead people to question the effects of immigration on society. Simply put, there is far less critical thinking of the benefits versus the drawbacks of immigration in areas with high numbers of Hispanics. People receive much more indoctrination from education system, for instance, to accept immigration as an unquestionably positive force in areas with high Hispanic populations and it is very socially unacceptable to argue (or even think) otherwise.

Steve Sailer writes:

Mike is exactly right: more diversity means less freedom of speech, as was seen in the Danish cartoon crisis and a host of other examples:

mike writes:

I think the fact that this debate is shifting to how people perceive immigration rather than what the data actually suggests about the effects on immigration is, in itself, very revealing.

Foobarista writes:

Why is this discussion not using the word illegal? Frankly, it is an intellectual perversion to discuss illegal aliens and their effects as "immigration", and the poll is dishonest if it used the word "immigration" without the dreaded word illegal.

Dezakin writes:

'Why is this discussion not using the word illegal?'

Because legality is a political fiction.

Peter Schaeffer writes:

The results are far less impressive than they look. Why? Immigration drives out Americans whose lives are devastated by Open Borders. Those who profit from Open Borders remain. The fact that native born Americans are net leaving states such as California should show that immigration is definitely an economic minus, not a plus. This has been true for quite some time. More than 100% of the population growth of California is due to immigration.

Any number of studies have found no overall material economic gains from immigration. The best known is the 1997 National Academy of Sciences review. It found possible gains ranging from 0.01% - 0.1% of GDP. It also found much, much larger tax and income redistribution costs. Given the grave risks associated with importing a foreign language population with irredentist claims on our territory and well known problems (50%+ dropout rates, 48% illegitimacy), this is a bad deal for our nation. Notably a detailed study in the Netherlands reached similar conclusions. Money quote "For all entry ages, however, immigrants turn out to be a burden to the public budget if their social and economic characteristics are equal to those of the present non-Western resident".

Of course, "Robin Hood in Reverse" income redistribution driven by unskilled labor obviously profits some people. Any wonder that high IQ types such Tyler Cowen and Alex Tabarrok find it so appealing? Elites always what cheap servants, the nation be dammed.

James Withrow writes:

Legality is not a political fiction for immigration for most occupations. True, in landscaping or food service or janitorial services, illegal hiring is rarely punished and goes unabated. But, for better-paying jobs in this country, immigrants almost always have to go thru proper channels to obtain employment. So, immigration legality in a practical sense is class-based.

While some of the comments above help to explain why the native-born in high-immigration states are more approving of immigration-- and self-selection is probably the best of those explanations-- there's one other thing you should consider.

The native-born in California are used to seeing landscaping jobs go to people they assume are immigrants and probably illegal. When you see an entire industry go to illegal hiring, I think it's natural to assume that "Americans" don't want those jobs.

But the native-born in low-immigrant states are in the middle of the process of landscaping going illegal. They still see plenty of "Americans" doing those jobs, so are less likely to buy this racist accusation that those are jobs "Americans" won't do.

mike writes:

But, for better-paying jobs in this country, immigrants almost always have to go thru proper channels to obtain employment.

The vast majority of illegal immigrants are not qualified for those 'better-paying' jobs you refer to.

harry e. writes:

I think that the mass migration of peoples of Mexico and central American countries will eventually cause a serious problem to all persons who are citizens of the United States. I do the same kind of work as these immigrates do and do well. I do find a problem in obtaining work. I have to do heavy, heavy advertising to compete with the floods of people entering into the United States. I am very competitive with their prices in labor and often below what they ask for in obtaining jobs. My wife often gets mad at me for the prices I offer to customers, but in order to compete with this labor force, you have to undercut their prices. I see as the children of these immigrates graduate from colleges the same will be true. There are fewer jobs and a much larger supply of labor. The country as a whole, will lose, in this process.

Comments for this entry have been closed
Return to top