Bryan Caplan  

Immigration vs. the Self-Interested Voter Hypothesis

Credentials and Teacher Qualit... Cesar Chavez on Immigration...
Who loses the most from additional immigration?  The data is clear: The biggest losers are immigrants who are already here.  This is hardly surprising: recent and new arrivals are in close competition because they supply nearly identical skills.  Ottaviano and Peri estimate that immigration from 1990-2004 raised the average native's wages by 1.8%, but slashed the average foreign-born worker's wages by 19.8%.

If self-interest had a big effect on policy preferences, current immigrants would be the staunchest opponents of immigration.  Are they?  I decided to check with the General Social Survey.  The GSS's best measure of attitudes about immigration is LETIN1.  It reads:
Do you think the number of immigrants to America nowadays should be...

increased a lot (=1); increased a little (=2); remain the same as it is (=3); reduced a little (=4); reduced a lot (=5)
The average answer for the whole sample is 3.7, with a median of 4.

I then defined the variable IMMIGRANT, which equals 0 if both your parents were born in the U.S., and 1 otherwise.  (As far as I can tell, the GSS doesn't have a good measure of whether you were born in this country, but my measure is fine for our purposes).  Survey says: Contrary to the self-interested voter hypothesis, people with at least one foreign-born parent are much less hostile to immigration: a full .8 points.

How big is this effect?  It is well-known that the educated and liberal are less hostile to immigration.  How does the effect of my immigrant measure compare?  Here's the multiple regression:


Education is measured in years; POLVIEWS runs from 1 (very liberal) to 7 (very conservative).  The estimated effect of having a foreign-born parent is therefore roughly equal to 14 years of education or 8 steps on a left-right scale that only has 6 steps to begin with.

It's one thing to say that self-interest has little or no effect on policy preferences.  The children of the foreign-born go far beyond this.  Immigrants hurt them the most, but they oppose immigration the least.  How is this possible? 

The best explanation is that the children of the foreign-born, like many other groups, are group-interested voters.  They're concerned about the well-being of people they identify with, people "like them."  The children of immigrants know what it's like to be an immigrant from first-hand experience.  They know the misery of the Old Country, and the hardships of the New.  And when they ponder immigration policy, their first thought isn't their wages.  Their first thought is that the law is denying someone like their parents, cousins, or neighbors a chance to work for a better life.

COMMENTS (14 to date)
Eric T writes:

I mostly agree with your explanation, but I think voter ignorance is probably a much bigger factor. How many immigrants do you expect understand the economics behind this? I mean, even the natives believe that immigration hurts them the most...

Faré writes:

The identification is not necessarily irrational. The recent immigrants know that if laws against new migrants come to pass, they are next in line to be discriminated against somehow. Also, inasmuch as new migrants include their friend, family, people they are more likely to marry, etc., it is rational for them to seek to increase the size of that portion of the market.

In other words, I think that you are looking only at a small part of the picture in your cost/benefit analysis of further immigration to recent migrants, and then jump to the wrong conclusion. Sometimes, small visible monetary gains are not the most important factors. You should know!

8 writes:

It's about group politics, which is what those who oppose mass immigration are concerned about.

Kevin H writes:

Does anyone actually believe that economic factors are the basis of anti-immigration feelings? At best they're a rationalization for the real basis which is more primal and tribal in nature.

I would be interested to see a study comparing white american attitudes toward Canadian/British/Irish immigration (legal or otherwise) to their opinions about Latin American immigrants.

People fear for their cultural norms/values far more so than their (imagined) economic interests.

Eelco Hoogendoorn writes:

Real effect of expressing such an opinion, either at the voting booth or polling station: zero.

Signalling value of percieved consistency and standing up for your group: nonzero.

Even if they would understand the economics of the matter (not sure to what extent this is the case), its still doesnt quite seem like it would be in their rational self interest to agitate against immigration.

Occam writes:


If self-interest had a big effect on policy preferences, current immigrants would be the staunchest opponents of immigration. Are they? I decided to check with the General Social Survey. The GSS's best measure of attitudes about immigration is LETIN1. It reads:

Are you not making the unwarranted assumption that current immigrants are able to correctly identify those policies which are in their self interest? It seems most probable that immigrants simply incorrectly identify a more open immigration policy as being in their self interest, easily explaining why they would support it.

A much better test of the effect of self interest on policy preference would be examining those immigrants who do identify more lax immigration policy as being "against their self interest", who still support said policy.

Occam writes:

Apologies for the double comment, but to follow up:

I believe you are utilizing a much too narrow conception of self interest in your analysis. You seem to only be focusing on economic self interest, forgetting that most immigrant workers have families they have left behind in their home countries. A more lax immigration policy would make it easier for these families to be reunited. Do you think it is permissible to attempt to contrast the psychic profit reaped by immigrants as a result of a more lax immigration policy with the economic costs which they would incur as a result of said policy being adopted? How can you possibly know whether or not a more lax immigration policy is in line with the self interest of an immigrant, without first possessing an extraordinary amount of knowledge about that particular individual's status and preferences?

kyle8 writes:

My hypothesis is that 1) the average immigrant is unaware of this economic argument. or 2) the average immigrant wants more people of their own culture to move into their new home.

I agree to some extent with what Kevin H wrote, but not entirely. Living on a border state I was once very pro immigrant, but I have seen first hand the rather large strains that huge amounts of immigration can have on a localized area.

I am certainly not a fan of unrestricted, unregistered immigration. Again this is a change in my previous libertarian beliefs that was unconcerned with the legality of recent immigrants.

Ted writes:

I think your hypothesis is plausible, but I think you make a bit of a leap in your inference. In order to uncover the effects of immigration on wages Ottaviano and Peri used a fairly sophisticated heterogenous agent, general equilibrium model and then calibrated it. I'm doubtful if most immigrants would intuitively understand the general equilibrium effects on their wages from further immigration. They may be able to observe their own wage reductions, but that they could tie that into the general equilibrium effects from immigration seems much more doubtful. Your hypothesis implies they understand this fact. I think it's entirely possible that if they were aware of the Ottaviano-Peri work, they may become harsh anti-immigration advocates and throw that whole in-group identity thing out the window.

Foobarista writes:

Immigrants are in favor of more of their kind of immigrants for the same reason that non-immigrants aren't: what you call "anti-foreign bias".

If you've ever been around immigrants, they _do_ refer to people not of their kind as "foreigners", particularly in their own languages.

Jason Malloy writes:

(As far as I can tell, the GSS doesn't have a good measure of whether you were born in this country, but my measure is fine for our purposes)

For future reference, Dr. C, this variable exists in the GSS; I use it a lot. The code you're looking for is BORN ("Was R born in this country").

Additionally there is PARBORN ("Were R's parents born in this country") and GRANBORN ("How many grandparent's born outside U.S.").

Ted Craig writes:

It's not somebody like their parents, cousins, etc. It IS their parents and cousins. Have more members of your extended clan in a country is beneficial anywhere you go. So, you confuse their self-interest.

Douglass Holmes writes:

Recent immigrants from Mexico and China know, first hand, how limited their opportunities are in those countries. Even if it is not in their self interest to support additional immigration, they can hardly justify keeping more immigrants out. They know what keeping them out means to the people we keep out. On the other hand, many citizens of the US who are not recent immigrants, or who immigrated from Canada don't really know first-hand what keeping these people out means to the would-be immigrants.

Dr. Liberty writes:

"The best explanation is that the children of the foreign-born, like many other groups, are group-interested voters"

Another way of saying this is that they are value-interested voters. They hold open immigration as a value because it would have benefited them in the past. And they want to treat others as they themselves would like to be treated. In this sense, it is self-interested, but using a wider view of "self" to include one's general values.

Comments for this entry have been closed
Return to top