Tag Archives: George Borjas

How did we come to be so certain that closed borders are our salvation?

Editorial note, added December 26, 2014: Welcome, Hacker News readers! This website is devoted to discussing the case for open borders, including the moral arguments for it and the practical question of how to get there. To address concerns surrounding migration liberalization, we suggest keyhole soutions and slippery slopes to it. For more about the site, you might want to read our site FAQ. Another post that you might find particularly relevant is Nathan Smith’s post on Mark Zuckerberg and FWD.us.

One puzzling thing I notice about debating immigration is how certain people often are that strictly restricting immigration is the right policy. Almost any person, when prompted, can articulate almost immediately a tonne of reasons why restricting immigration makes sense:

  • National governments have carte blanche to exclude any foreigner from their territory as matter of moral right
  • Open borders would let terrorists into our country
  • Open borders would let foreigners steal jobs from our people
  • Open borders would allow a foreign people to invade and steal our country from us
  • Permitting immigration imposes foreign cultures on our people
  • Immigrants will abuse our welfare system
  • Immigrants will undermine our institutions and replace them with their inferior ones
  • Liberalising immigration won’t really help poor foreigners anyway
  • Too many immigrants will swamp our territory or society to the point that it cannot function any longer
  • Letting in low-IQ/-skilled immigrants harms our economy or polity

But for some reason, the same people eager to expound on the litany of catastrophic harms that would no doubt ensue under open borders are rarely able to cite any sort of academic literature that backs them up. Their best retort, in terms of academic prestige, is George Borjas’s work on immigration’s impact on American wages, and maybe Robert Putnam’s work suggesting that diversity reduces some theoretical measure of “social capital”. You can’t find any empirical estimates that seriously support the above hypotheses — at least not to the degree that has people so certain the only right immigration policy is building a better and higher prison wall.

Now, if you turn the above propositions around, on all of them, we are either certain that open borders is immensely beneficial, or we’re just unsure. We know for a fact that liberalising immigration immensely helps the poorest human beings alive. Hardly any serious restrictionist disputes this; the only ones I’ve encountered who do are basing their certainty on foundations of sand: the most memorable example was a person who suggested that estimates of the place premium are wrong, because when you adjust for purchasing power parity, people in poor countries have better living standards than people in the US — such an economically-illiterate claim that it doesn’t even merit a rebuttal here. Most restrictionists are happy to concede that immigrants are made better off — they just believe that the act of immigrating makes natives dramatically worse off.

But the propositions to do with crime and “job theft” are our runners up for certainty: in the empirical literature, it’s difficult to find any serious social scientist who believes immigration increases crime rates, especially in a significant manner. And among economists, Borjas alone sticks out like a sore thumb for producing estimates showing dramatic depression of native wages (“dramatic” being a short-run reduction of a few percentage points). If there are any serious peer-reviewed, published analyses showing immigration leads to a significant spike in crime, or any landmark studies besides Borjas’s contradicting the economic consensus, I’d love to see them, because they seem to have slipped the minds of the restrictionists I’ve met so far.

Still, for virtually all the other propositions above, the evidence is either limited, decidedly mixed, or both. The long-run institutional, political, and societal effects of immigration have not been thoroughly studied in an empirical manner. But assuming we place the most weight on these outcomes (and ignore the other findings on the economics and crime of immigration), this means we ought to be cautiously uncertain about what the right immigration policy is. It means that even if we favour restrictionist policies, we do so with great uncertainty.

Yet the spectre of open borders seems to produce a stout certainty on the part of many people, who even if they aren’t dedicated restrictionists, seem quite convinced that the status quo or something close to it is certainly the right and best policy, given what we know now. There is strong certainty that a more liberal immigration policy of any kind would be a horrible idea. Yet engaging with these pro-status quo or even pro-closed borders assertions, one finds them disappointingly devoid of empirical backing.

The best ace the restrictionists have in their back pocket is the nuanced argument that reducing the proportion of high-IQ people in an economy below a certain percentage, or raising the proportion of low-IQ people in an economy above a certain percentage, would lead to a slowdown in innovation or corrosion of successful institutions. But even this claim is problematic, since it is difficult to tell how far IQ and economic growth and innovation are causally linked. And if having low-IQ immigrants is so devastating, this effect should surely be easy to demonstrate through meaningful measures of harm: slower economic growth rates, fewer number of patents filed per capita, higher crime rate. If we can’t observe these harms at existing levels of immigration — and, it bears repeating, the overwhelming majority of the empirical literature cannot find any such meaningful harms — then right now we are simply worrying about IQ for the sake of worrying about IQ.

If this whole post seems wishy-washy, since I’m essentially conceding that we are uncertain about the effect of open borders on quite a few dimensions, you’re partly right. But it’s more accurate to say that we are just as equally quite uncertain about the impact of closed borders, and to the extent we know anything with certainty, it’s how devastating they are. We can’t even rule out that closed borders are incredibly harmful to us on a number of dimensions (a straightforward reading of the empirical literature suggests that if you want to cut crime rates, you should subsidise immigration). Worse still, given the consistency of the literature regarding the impact of closed borders on the world economy and global poverty, we are absolutely certain that closed borders keep millions of people in poverty of the worst kind. We know that on average, the effect of closed borders halves the world economy.

Even if you think that the status quo of closed borders is right, it is worrying how uncertain we are about this conclusion. In many cases, the issues at hand simply haven’t been studied enough, and we know virtually nothing (we certainly don’t know enough to support most common restrictionist assertions about immigration). We do know the incredible destruction that closed borders wreaks on the world economy and the people of the world, to the tune of halving world GDP and keeping millions in poverty. We ought to have our top men and women working on figuring out whether we can crack the borders open at all. The fact that we don’t means we are simply irrationally certain that closed borders is the right answer. And that irrationality strikes me as best summed up in this 1881 cartoon, depicting Irish immigrants to the US — men and women bringing terrorism, crime, and corrupt institutions to American shores, people whose only contribution was adding themselves to the welfare rolls:

Editorial note: If you’re interested in discussing the many issues related to open borders, check out the Open Borders Action Group on Facebbook.

Open borders and the impending apocalypse

A common approach rebutting open borders is to argue that the costs of liberal immigration policies outweigh the benefits to humanity. I’ve never actually seen this belief explicitly expressed in a universalist manner — the argument is usually focused on how immigration will destroy the wealthy economies and liberal societies of the world. But I think this argument is a serious one, and I give it serious credit.

This does not always seem to be the case; one may sometimes feel that open borders advocates are a tad glib in dismissing concerns that open borders might “kill the goose that lays the golden egg.” To be blunt, this is because there is no empirical evidence supporting this claim.ChineseExclusionActHandbill[1]

If we look at the past, the same concerns people have today about Latin American, African, Arab, or South Asian immigrants used to be directed at East Asian, Southern European, and Eastern European immigrants. The same people today who vocally embrace “high-IQ” or “high-skilled” immigration of Jews, Europeans, and East Asians, would find that these very same groups of people used to be the “low-IQ” and “low-skilled” immigrants who were not so long ago literally treated as vermin in their countries. Fears that the unintelligent, criminal, brute Catholic Irishman or Italian, or the conniving and unintelligent Jew, might ruin civilisation turned out to be unfounded.

If current levels of immigration were a harbinger of impending doom, it would be quite easy to prove this. It’s fairly easy to point to anecdotes — but surely laying one’s finger on the data would be easy too. You’d show skyrocketing rates of crime, environmental collapse, or economic depression and clearly link them to immigration in some fashion. Yet no credible academic study I’m aware of has been able to do this. Restrictionist memes blame immigrants for the impending collapse of civilisation in Western Europe or California, yet the actual academic backing for these views is hard to find.

It’s surely not because academics are afraid of voicing politically incorrect views. A vast conspiracy of intellectuals to open the borders and silence such a devastating finding would be quite difficult to keep secret. And yes, one can find credible empiricists skeptical of immigration. Yet the most famous academics whose works actually credibly show negative impacts from immigration — George Borjas and Robert Putnam — both do nothing but disappoint.

Borjas finds that immigration to the US slightly reduces the incomes of the poorest American citizens — something that could easily be addressed through keyhole solutions which redistribute some of the gains from migration to poor natives. Putnam finds that social diversity reduces a theoretical measure of “social capital“, but even his credible result has been challenging for other researchers to replicate. If this is the worst we have to fear from immigration, I say bring it on.

The truth is, we don’t know very well what a world with open borders would look like. We know it would double world GDP — studies of the effects of  greater immigration on world GDP are remarkably consistent in predicting a massive boost to world income, regardless of their theoretical specifications or empirical approach. But given that far too few academics are seriously studying the impacts of immigration in an empirical fashion, we don’t have enough data to say with certainty that much of what we currently know to be true about immigration would still hold true in a world with massively looser immigration policies than today’s. We couldn’t guarantee that immigration would continue to be more or less neutral with respect to native incomes, and have a neutral to positive impact on crime.

But the precautionary principle only militates against immediate open borders. There is nothing stopping us from experimenting with a little more immigration. As the world’s population grows, as humanity grows richer, it makes absolutely no sense that our visa policies are held hostage by the immigration quotas of decades ago.

Open borders advocates actually aren’t asking for much. We simply believe in making the presumption that all who seek to move may do so — a presumption that can be overriden by a clear and pressing need, such as, say, the actual risk that your civilisation might collapse if you don’t shoot the next prospective immigrant in the face. As philosopher Phillip Cole puts it:

In effect all I’m proposing is that immigration should be brought under the same international legal framework as emigration. Immigration controls would become the exception rather than the rule, and would need to meet stringent tests in terms of evidence of national catastrophe that threatens the life of the nation, and so would be subject to international standards of fairness and legality.

I and I think other open borders advocates take concerns about global catastrophe quite seriously. Given that we typically come from universalist and sometimes even nationalist or citizenist moral starting points, we have every reason to be concerned that open borders might mean the end of the world as we know it, in a horrible way. But search the evidence, and you find no actual reason to be concerned about current immigration levels, and every reason to believe that open borders would immensely benefit us all. Even if you don’t find the evidence sufficiently compelling to tear down the border checkpoints right this moment, it’s compelling enough to demand more thorough research and compelling enough to demand experimentation with ever more liberal immigration policies.

How Does Immigration Impact Wages?

This post was originally published on the Cato-at-Liberty blog here and is republished with the permission of the author.

Many Americans are curious about the impact of immigration on the wages of other Americans.  The best research on this focuses on the period between 1990 and 2006, when almost 17 million people immigrated to the U.S. lawfully and a net 12 million came unlawfully.  The first major study is by Borjas and Katz (B&K) and the second is by Ottaviano and Peri (O&P).  O&P borrowed much of B&K’s methodology.  Here are the long run findings:

B&K draw a more negative conclusion than O&P.  The main differences are that O&P assume capital adjusts quicker to increased labor abundance and immigrants are more complementary.  B&K’s paper reflects their assumptions about native-immigrant substitutability.  Since immigrants are more likely to have less than a high school degree and more likely to have a graduate or professional degree than natives, B&K’s model assumes natives in those categories are competing with immigrants for jobs and therefore experience wage declines. 

Both O&P and B&K found that increased immigration has a larger affect on immigrants than natives.  Depending on their level of education, longer settled immigrants experience greater wage declines and smaller wage gains from more recent immigration compared to natives:

Both sets of authors rightly assume that more recent waves of immigrants are most similar to immigrants from older waves, making the two arrival cohorts of immigrants substitutes in the workplace.  Recent papers by Ethan Lewis and Giovanni Peri and Sparber make convincing argument that language ability of recent immigrants makes them more similar and, thus, substitutable with previous waves of immigrants.  Language ability also makes immigrants complements to natives, partly explaining why O&P and B&K found wage increases for so many American workers as a result of immigration.

Here is a comparison of the long run wages effects on immigrants and natives from the O&P and B&K study:

These charts merely explain the results of previous waves of immigration on the American labor market.  If immigration increases in the future these numbers will likely be different but the past is always a useful guide for anticipating the effects of future policy changes.


Borjas, George, “The Labor Demand Curve Is Downward Sloping: Reexamining the Impact of Immigration on the Labor Market.”

Borjas, George and Lawrence Katz, “The Evolution of the Mexican-Born Workforce in the United States.”

Ottaviano, Gianmarco and Giovanni Peri, “Immigration and National Wages: Clarifying the Theory and the Empirics.”

Peri, Giovanni and Chard Sparber, “Task Specialization, Immigration, and Wages.”

Lewis, Ethan, “Immigrant-Native Substitutability: The Role of Language Ability.”