Bryan Caplan  

Gochenour-Nowrasteh on the Political Externalities of Immigration

How to measure the influence o... The Gains from Getting Rid of ...
Does immigration expand the welfare state by increasing the share of voters that benefit from government programs?  Or does immigration contract the welfare state by undermining voters' sense of national identity?  Critics of the welfare state tend to think the former; fans tend to think the latter.  Who's right?

The latest Cato working paper from my co-author Zachary Gochenour and my former student Alex Nowrasteh concludes that - at least in the United States - both sides are wrong.  Yes, most of the European evidence supports the view that immigration paves the way for austerity:
[I]ncreasing diversity could be a concern for the political supporters of the welfare states of Europe: in 2000, over 50 percent of the European population was concerned about immigrant abuse of the welfare, with those living in nations with higher social expenditures convinced that immigrants are more likely to abuse welfare [7]. A Norwegian survey about the political feasibility of introducing a minimum income found that 66 percent of Norwegians initially favored the scheme. However, merely mentioning that non-Norwegians residing in the country would receive the same benefi ts reduced support for the program to only 45 percent [5]. In Sweden, increased immigrant population share led to less support for redistribution among native Swedes according to surveys conducted every election year by the Swedish National Election Studies Program [12]. Furthermore, some authors have suggested that immigration has given new life to political parties that bundle anti-welfare policies with xenophobic policies [18]. [see the paper for cites]
If you look at the fifty United States, however, immigration has no detectable effect on TANF/AFCD, K-12 education, or Medicaid spending.  This is true for both per capita and total spending:
These findings are consistent and robust. No e ffect was found for immigrant population share or diversity on TANF benefi ts levels available per family of bene ficiaries or the total spent on the program. Likewise, there was no e ffect for total or per-pupil K-12 spending or for total or per-capita Medicaid spending. Therefore, these findings lend no support to the idea that immigration or the resulting immigration driven increases in diversity is linked to higher public spending by any of these measures.
You might think the TANF result is driven by immigrants long eligibility lag, but that's not it:
Limited TANF eligibility for immigrants could help explain the lack of native reaction to increased immigration and diversity. However, data from the GSS suggests that the median American thinks the government provides too much assistance to immigrants. In reality, the programs have stricter eligibility requirements than before: the Personal Responsibility and Work Opportunity Reconciliation (welfare reform) and Illegal Immigrant Reform and Immigrant Responsibility Acts (IIRIRA) of 1996 restricted non-citizen eligibility for TANF. Prior to 1996, non-citizens were generally eligible for the same welfare benefi ts as citizens but welfare reform and the IIRIRA barred TANF for new immigrants for five years after their entry. After the five year bar, states were allowed discretion in allowing non-citizens access to TANF. As of 2010, 34 states and Washington, D.C. allowed lawful permanent residents who have been in the U.S. for more than five years to draw on TANF. Many states restored portions of the welfare bene fits limited by welfare reform. Because of these laws and a diminishing poverty rate among immigrants, 4 percent of immigrants were using TANF in 1995 but only 1 percent were in 2009 [36]. [see the paper for cites]
What's going on?  The simplest story is that immigration has two roughly offsetting political effects: Although immigrant voters are a little more pro-welfare state, their very presence makes native voters a little more anti-welfare state.  Gochenour-Nowrasteh explore other stories at the end of the paper.  Read the whole thing.

P.S. If your economics or public policy department wants to hire an excellent assistant professor, Gochenour is on the market this year.

COMMENTS (18 to date)
David Friedman writes:

It seems to me that you are missing a straightforward economic argument for why one might expect less restrictive immigration laws to reduce the welfare state. If immigration is relatively free, a high level of welfare payments will tend to attract the sort of immigrants likely to go on welfare. That is an undesirable consequence from the standpoint of those already here, hence a reason to vote for less welfare. This is a familiar argument in context of interstate competition within the U.S., and the reason why supporters of the welfare state prefer it to be at the federal level.

Gene writes:

Are immigrants really pro welfare state, or are they just Democrat leaning because Republicans want to deport them all? I have a hard time imagining an immigrant believing to be entitled to welfare. People who switch countries to find higher paying jobs tend not to think this way in my personal experience.

Tom Dougherty writes:

I just started read this working paper. Page 3 claims, "72.9 percent of U.S. foreign born are neither white nor black", which is quite a remarkable claim. So, I check their source [footnote 34] and found their claim is not correct. Actually, 43.8 percent of U.S. foreign born are neither white nor black. Is this a significant problem for their paper? I don’t know because I have just started reading it.

So, how did they go wrong? According to the 2010 ACS 1-year estimate, 47.9 percent of U.S. foreign born are white and 8.3 are black, which means 43.8 are neither white nor black. The authors of the paper mixed ethnicity and race together to come up with their misleading claim. They added white not of Hispanic ethnicity (18.8 percent) to black of any ethnicity (8.3 percent) to get 27.1 percent. They subtracted this from 100 percent to get 72.9 percent. I would suggest that Hispanic is not a race and to added blacks of any ethnicity to white of non Hispanic ethnicity is incorrect. Their first provocative claim doesn’t hold water.

Andrew_FL writes:

Couple things:

In the first place, you are basing your conclusion that totally unrestricted immigration would not have this effect, on the basis of empirical evidence from restricted immigration. This directly ties into the next problem with your conclusion, which is:

You are taking an effect within a domain where said group reflects a small minority of the population, and extrapolating it into a domain outside all experience, where the group would then represent a large plurality of the population. In that circumstance, it doesn't matter how strong a pushback against socialism comes from the people already here, there cannot conceivably be enough to offset the effect.

The polling data still speak for themselves. You can tell a self serving story (as a commenter has above) that if only those meany Republicans would just embrace amnesty these people would suddenly become big fans of capitalism. It's not going to make the story true. But it's better than trying to pretend it doesn't matter.

aretae writes:

I'm having trouble parsing, as an open borders advocate.

no detectable effect on per-capita or total spending? That's impossible, unless the amount of immigration is too small to detect.

If there is a measureable quantity of immigrants, then either the total spending increases, or the per-capita spending goes down. Which happened?

Gene writes:


Of course total government spending will increase. So will the tax revenue. What will happen per capita is an empirical question, and can be controlled through policy.

And why only consider the public sector? Why do people act like everyone we consume and produce goes through government?

Daniel Kuehn writes:

Tom Dougherty -
It's standard to combine race and ethnicity (and label it as such - "race/ethnicity"), and speak in terms of "white, non-hispanic" when referring to whites. Maybe they need to label better (I don't see notation of this anywhere), but it's neither provocative nor incorrect.

Daniel Kuehn writes:

My bigger concern is the selection issues. They cite Borjas, but don't seem to do anything with that insight. TANF and the rest, we would expect, is a RHS variable with immigration on the LHS.

If this is just a correlation mixing the two effects and we think there is any validity to welfare magnets (which seems reasonable), then it would imply the effect of immigration on welfare is more negative than indicated.

I find it a little odd too that the problem is framed as a "welfare state" issue and they tied K-12 and UI spending into that... but that's just a framing concern.

nl7 writes:

So is the theory that Europeans support welfare altruistically or selfishly? Does the antipathy towards immigrant transfer payments come from the selfish idea that welfare should support "people like me" who might need help and immigrants are not part of that group? Or does it come from the altruistic idea that welfare should help "people who need it" and immigrants are not part of that group? Either case is xenophobic and possibly racist, but the theory is different (though not mutually exclusive).

I suppose there's also the more forthrightly spiteful motivation that just says, regardless of motivations to support welfare, it's bad if immigrants get anything nice.

I'm rejecting out of hand the idea that immigrants shouldn't get welfare because they didn't "pay in." That's not a requirement for most forms of transfer payment (FUTA/SUTA and worker comp aside) and so there must be some reason why it would only apply to immigrants.

sam writes:

The solution is far simpler:

The tax-paying population has a certain tolerance for taxation for welfare spending. Where there are more welfare recipients, there will be less spending per recipient.

Thus, welfare-seeking immigrants will go to the states that at present have less welfare recipients per capita, as they have less native-born welfare recipients to compete with.

Tom Dougherty writes:

Daniel Kuehn,

You say, “it’s standard to combine race and ethnicity”. But that is not what they did. They did the opposite. They separated ethnicity from race for whites. But they did not separate ethnicity from race for blacks. And then they added the two together. This is not legitimate.

The Census Bureau, the data source for the CATO working paper, does not consider Hispanic a race. Hispanic can be of any race and Census notes that as such. So, when the source of your data does not consider Hispanic to be a race, and you report that the Census Bureau estimates that 72.8 of the U.S. foreign born is neither of the white race nor of the black race then that becomes a provocative and, as it turns out, incorrect statement.

Daniel Kuehn writes:

Tom -
I separate out ethnicity from black when I'm classifying race/ethnicity - I agree they should. And they could label it better when they introduce the data. But it's not provocative, and presumably the number won't change too much when they take out Hispanic black.

Hispanic of any race, white non-Hispanic, and black non-Hispanic are what people typically think of when they hear Hispanic, white, and black, and that's why researchers usually classify race/ethnicity in that way.

Tom Dougherty writes:


You keep saying they could have labeled something better. It is not a label. It is a statement in the text of the article. An incorrect statement about race that is not supported from the source data and there is no mention of ethnicity in the statement. Your obtuseness is not a valid defense for inaccuracy.

Daniel Kuehn writes:

Tom -
re: "You keep saying they could have labeled something better. It is not a label. It is a statement in the text of the article."

A statement that serves the task of labeling the variables - providing an identification of what the variables consist of.

I'm running out of ways to say this, buddy.

re: "An incorrect statement about race that is not supported from the source data and there is no mention of ethnicity in the statement."

It's not an incorrect statement so long as there IS mention of ethnicity in the statement which is PRECISELY what I've been saying they need to include/clarify. But given that they explain what they do (and clean up the black non-Hispanic bit), everything is entirely above board and the superior way of classifying race/ethnicity.

re: "Your obtuseness is not a valid defense for inaccuracy."

Better watch it - Laura Landsburg tells me she takes hostile comments very seriously.

Daniel Kuehn writes:

Let me make this clear, Tom.

They should - in their data section - say "we did X, Y, and Z, and "white" refers to white non-hispanics, etc. etc." They should also take out black Hispanics and put them in with Hispanics.

That would be ideal. There is nothing misleading, inaccurate, or provocative about combining race and ethnicity with Hispanic as a primary designation and then divvying up the non-Hispanics across other race groups. It's very common and it's very common for good reason.

Tom Dougherty writes:

Yes, Daniel. If they had done this thing, that thing, and the other thing and if they had included this, took out that, rearranged this table and done a back flip, then it would have been a completely accurate statement. Good grief.

Christopher Chang writes:

The signs of the effects appear correct to me.


(i) this is more relevant to countries like Australia and Canada, which have political elites that are still meaningfully responsive to public opinion on these issues. US public opinion only appears to have the power to slow down "permanent underclass" immigration and welfare state growth; all measures that would meaningfully *reverse* either have been kept off the table for decades. (E.g. an obvious near-Pareto-improvement over the status quo involves slashing Mexican-source immigration by >90% while allowing increased immigration from similarly poor groups which have been assimilating better. This would be more popular, and far more beneficial, than all "immigration reform" measures that have had a chance of making it through Congress lately.)

Until the current US political elite's deathgrip is broken, no deliberate concessions on either of these issues is plausibly in the public interest, because there is every reason to believe that the general public will be blocked from enjoying enough benefits from any deal for it to be a net positive for them; otherwise the deal would not be politically viable.

(ii) one would expect that it only takes a relatively small amount of immigration (especially if it is distributed "efficiently") to reap most of this type of "benefit", whereas the losses obviously do not drop off. So this is one more valid argument for not reducing immigration to zero anywhere, but I doubt it justifies increasing immigration over the status quo in the US.

Daniel Kuehn writes:

Tom -
I said change one thing and describe what they did better.

Back-flips completely optional.

You've argue yourself into a corner and are exaggerating the counterargument to get out of it.

Comments for this entry have been closed
Return to top