Immigration, production, and the Rybczynski theorem

Two of the assumptions of some of the pro-immigration arguments, when combined together, strike me as a bit odd.  It is commonly claimed for instance that migration to the United States does not lower American wages much if any (I agree by the way, as do most economists).  It also is claimed that migration will help us boost our production of capital-intensive goods, such as tech products.  Therein lies the tension.

Enter the Rybczynski theorem.  The sequence here is as follows.  An influx of labor does not lower wages, but it does cause both labor and capital to flow out of the capital-intensive sector and into the labor-intensive sector.  Labor-intensive production will rise but capital-intensive production will fall.  (If you are wondering about the intuition, consider that prices and wages are remaining fixed.  All of the adjustment must take place in terms of quantities, and equalizing marginal product and wage, following the influx of new labor, requires more capital in the labor-intensive sector, thus draining some production from the capital-intensive sector.)

In other words, the core model for constant or near-constant wages, following immigration, also implies that capital-intensive production should fall, following an exogenous influx of labor.

You will note that immigration remains welfare-improving in this model.  Still, it is not exactly the deal which has been promised.  You also can imagine someone taking a more dynamic perspective and fearing the long-term growth consequences of losing output in the capital-intensive sector.  You also now have liberty to wonder if a negative wage effect might mean a stronger rather than weaker case for immigration, given that the capital-intensive sector in the U.S. arguably produces global public goods.

Here is one empirical study of the matter, by Slaughter and Hanson.  It supports the predictions of the Rybczynski theorem.  Ethan Lewis (pdf) considers the famous Cuban boat lift example, when the influx of immigrants did not lower wages, but he also finds it may have hurt capital goods production in the Miami area.  Those results are hardly dispositive, but they don’t exactly throw out the Rybczynski framework either.

To be sure, the Rybczynski theorem is far from self-evidently correct.  Prices and wages need not be fixed, especially for a country as large on the global stage as the United States.  2 x 2 x 2 models do not adequately capture the heterogeneity of labor.  Even defining “labor-intensive” and “capital-intensive” is fraught with ambiguity when countries do not share all of the same technologies, as illustrated by the debates over the Heckscher-Ohlin theorem (to measure labor-intensity, are we adding up the number of bodies or instead measuring their “effectiveness”?, in which case labor and capital blur together because capital makes labor more effective).

Still, models are useful in organizing our thoughts, and this is a model which I do not see getting much attention.  I typically have applied comparative advantage to this question (if more immigrants come, high-skilled citizens are freed up to produce more capital goods), but perhaps the Rybczynski model is also relevant.

I also note that I do not view the primary purpose of this blog as hammering home specific policy conclusions, I would rather put doubts and thought processes on the table.  Anyway, maybe this model provides some structure for a better understanding of the trade-offs involved with immigration policy.

I look forward to reading your comments on this issue.


Really depends whether we're talking about high-skilled or low-skilled immigrants, here. There's always the fact that Heckscher-Ohlin basically doesn't predict what actually happens at all (Leontief) – i.e. that USA has the most capital-per-worker but exports labor-intensive stuff and imports capital-intensive stuff (those Japanese cars are good...)

But maybe one way of explaining this (don't know the empirics) is considering human capital separately. We produce a hugely disproportionate number of PhDs, have a deep market for skilled labor, so forth. (Exporting Photoshop is labor-intensive, but not really). I mean defining these terms, as you noted, becomes really murky. We can add up the number of bodies, yes, but can we add up the "number of capitals" or worse the "number of human capitals"? This can almost only be measured in effectiveness or investment.

It's always informative to take an argument to its logical conclusion. If enough Mexicans and Cubans and whatever else "unskilled immigrants" come (say 100 million) our wages would *have* to decrease. Our comparative advantage would change immediately. I'm all for LSI because I think America has room for a lot more people and that our current population density is unfairly low. I think it's good economics, and I think there's nothing wrong with a higher GDP (a good way of counterbalancing China, as well). But no one can tell me that there's not a point after which LSI changes the fundamentals of capital (human or otherwise)-intensive America to farming and exporting bananas.

*(Note, unlike others I don't think this is per se a bad thing. But it is, and a lot of people don't want to accept this).*

Maybe we'd get back our good ol' manufacturing though.

Leamer proved 30 years ago that Leontief was wrong -- the conceptually correct calculation is comparing capital-intensity of U.S. net exports to U.S. consumption. The H-O model has held up well over time. See Leamer's latest defense of H-O here:

I'll look into this. I don't think it changes my argument which is that Tyler's question depends entirely on whether we're talking about LSI or HSI. And I'm inclined to agree that it's a matter of comparative advantage.

We sent a man on moon without 3rd world invaders. UNIX, C, DARPANET, Supercomputing, Genomics, all were invented by White People. After high-skilled 3rd world invaders like Indians, Chinese came, America is a cesspool of corruption, & failure. Indians smile at you, while stabbing in back. Indians, Chinese, deceive, lie, cheat, steal technology.

Japan, Korea, Israel are doing better without immigration. Boeing 787 is more Japanese than American. this is not an ideological battle of Socialism vs liberty. This is war against white people. Why do hostile globalist elite defend Israel as a Jewish ethnostate with Jewish only immigration, but ravage white majority Europe/North America into a multi-ethnic, multi-cultural Gulag with dystopian non-White colonization?

East Asia is 99% yellow. Africa is 90% Black. West Asia is 99% Brown. But 3rd world colonizers, Muslims, Sikhs, Hispanics are aggressively advancing their agenda by annihilating gullible Whites, just as China annihilates Tibet.

Why do gullible Whites cuckold for murderous anti-White elite, who confiscate our guns, infiltrate/subvert our banks & espionage agencies, indoctrinate White kids in academia/mass media, plunder White jobs/wages, & butcher White soldiers in bankrupting wars?

"Native" Americans are not native. They invaded from East Asia. Yellow & Brown races committed 10-times more genocide, slavery, imperialism than Whites. Since Old-Testament, Whites have been victims of Jewish/Crypto-Jewish, Turkic, Muslim, N.African imperialism, slavery, genocide.

Gullible Whites should reject subversive anti-White ideologies - libertarianism, feminism, liberalism, & reject hostile slanders of racism. Love to all humanity, but White people must organize to advance their interests, their fertility, their homelands. Spread this message. Reading list: , , ,

Ok dude, you don't have to convince me of all the amazing things American white men have done. But so many of the lab scientists in academia are all foreigners. Few others are willing to get paid peanuts for such hard work.

And Rajat Gupta was a real corrupt guy, but wait Bernie Madoff is white. Oh wait, he's Jew, you probably hate him.

Comments for this post are closed

Being paid peanuts for skilled work? That sounds like a feature not a bug. *rubs hands together menacingly*

Comments for this post are closed


I despise Rajat Gupta and Bernie Madoff because of what they did, not their race / ethnicity. Several people committed suicide after the Madoff scandal broke. In China, they would have hanged him.

Its absolutely true that a substantial fraction of lab scientists in the U.S. are immigrants. However, it's also true that immigration has driven Americans out of these fields. From my comments yesterday.

"The issue of native displacement from STEM fields is very real. A few years ago, I took my son on a tour (open house) of Harvey Mudd university. A very impressive school to say the least. More on that later. What shocked me, was the professors literally begging the parents and kids to consider a career in science and technology. I heard professors saying things like

“You will be able to get a job” “No, jobs for software engineers still exist in the U.S., lots actually” “Scientists and engineers have very low unemployment”

The skepticism of the prospective students and parents was palpable. The combined notions that your job will be outsourced and you will be replaced (sooner or later) by an H1B were commonplace. Of course, they were a material deterrent.

What made this so disturbing is that the Harvey Mudd professors were right. Harvey Mudd graduates have (probably even today) excellent job prospects. Harvey Mudd may well be the best science and technology school in the U.S., perhaps better than Caltech and MIT (though smaller). Decades ago, I obtained a science degree from a school with a good reputation, the University of Chicago. I was amazed at how much better the Harvey Mudd students were. The Harvey Mudd professors had/have considerable basis for their confidence, as to the employment prospects of their students. Yet even in Claremont, CA parents and students were apprehensive. Note that my visit was well before the crash of 2008."

Comments for this post are closed

@Peter Schaeffer:

I don't get your anecdote. You describe a parental fear that could very well be irrational. Wasn't it the 80's when we feared the Japs were going to own all America? Wasn't it 150 years ago that Californian labor violently opposed Asian immigrant railway gangs?

For arguments sake, let's assume these parents are right and the Professors wrong. What then? What better major would you suggest? Psychology? English Lit?

There's bound to be competition and let's face it. 50 years ago you could make a fair living making toys or stitching socks in some southern factory. You can't do that now (mostly). Do you support a large customs-duty on Chinese toys and Indonesian made clothes so that Americans can regain that lost paradise of factory labor?

Comments for this post are closed

Notice the rise of the death or glory model of being a computer programmer. Either you will be a Mark Zuckerberg type and start your own company and get rich, or Mark Zuckerberg will start a giant astroturf political movement to have you replaced by H-1B foreign male programmers to do all the grunt work for Zuck. American women who want a 9-5 middle class job compatible with having One of the reasons there are so few female computer programmers today is because women aren't as interested in death-or-glory careers.

Comments for this post are closed

@Steve Sailer

BLS statistics shows 34% of "Computer Systems Analysts" as women. Not too shabby. Your "death or glory" model is quite generic. You could as well use it to explain away lawyers (30% women), physicians (32%) etc. where H1B's is not a strong factor.

Your reasoning to connect this to H1Bs is tenuous.

Besides it's hardly career death when you consider that median 2010 pay was ~$75,000 for a programmer. The fact that women choose to be teachers, social workers and psychologists has not much to do with Zuckerian strategies at all.

Comments for this post are closed

Academics like to be buried in information so they don't have to make difficult decisions. As animals we have become numb to our instinct through over rationalization and are destroying our souls in the process. Embrace the irrational, we'll be better off for it in the long run.

Comments for this post are closed


In the specific case of Harvey Mudd, the parents and students are wrong. Getting a job will not be a problem for the typical Harvey Mudd graduate. However, the Harvey Mudd anecdote shows how pervasive the fears of offshoring and H1B displacement really are. Of course, these fears are completely reasonable. The elite commitment to outsourcing every portable job in the economy and replacing every native worker hasn't exactly diminished over time.

"Do you support a large customs-duty on Chinese toys and Indonesian made clothes so that Americans can regain that lost paradise of factory labor?"

Yes, of course I do. Toys and textiles made with otherwise unused (unemployed) domestic resources (labor) are "free" for the U.S. economy as a whole. Imports are not. A trivial economic result is that tariffs raise GDP in an economy operating below capacity. That would be the U.S. today.

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Good post. Though isn't much of capital-intensive industrial investment now a choice variable in the United States, depending on the choices of the federal government?

Though isn’t much of capital-intensive industrial investment now a choice variable in the United States, depending on the choices of the federal government?

Well it depends on your definition of much.

Comments for this post are closed

Comments for this post are closed

One contrary view against immigration that I recently read (i think some article pointed out by Tyler a while ago) is that 'national identity' is essential to have a sense of 'culture' and 'equity' and immigration distorts these values. I wonder whether there are any models to support this?

Apart from this, I'm not aware of any other sensible (if true) arguments against immigration.

If 5000 years of human history counts, then yes, there are many models that support this. The deciding factor is the form of government: if you have a strong authoritarian government, such as a monarchy, many different cultures and nationalities can be held together.You don't need national identity to hold a monarch together, obviously! And many monarchs were in fact foreigners in their own countries. As for democracy, it's a bit tougher. A house divided itself cannot stand, and all that jazz. If democracy/republicanism remains the dominant government form, then there will likely be many countries with new borders. If instead a monarch is crowned, or a totalitarian ideology takes over, the national borders can hold as group identity is not important (monarch) or suppressed (totalitarian). Military dictatorship isn't good enough unless the top general is the Emperor because the military would probably become an oligarchy and eventually splinter.

Do these models describe what made the United States different, given the relatively open immigration for much of its history?

Outcast German revolutionaries from 1848 made perfect cannon fodder for the Union only a decade and a half later.

Comments for this post are closed

"Do these models describe what made the United States different, given the relatively open immigration for much of its history?"

Actually, they tell you a great deal. Throughout most of America's history the immigrants coming to America were remarkably homogenous (white, protestant, Northern European, etc.). And as it turns out, immigration wasn't particularly open either. The cost of immigration was so high that only relatively skilled individuals from relatively rich countries (back then Northern Europe) could afford to immigrate.

As a consequence, immigrants to the United States were, on average, better educated than the natives. This started to change towards the end of the 19th century as shipping technology improved. Eventually it became possible for unskilled labor to immigrate to the U.S. on a large scale.

Not surprisingly, public opinion turned against mass immigration after 1890 and around WWI mass immigration was stopped.

If being relatively homogeneous is a virtue and Northern Europe is an archetype of this precept the World Wars were a strange aberration. Maybe they were, but even then describing 400 years of quite-bloody previous European history is somewhat hard.

Europe's homogeneity is a thin veneer and often illusory. Let's wait and watch Europeans bail each other out as this EU crisis unfolds.

Comments for this post are closed

As far as I know, white, Protestant, Northern European nations are still distinct countries with distinct cultures and languages, e.g., Great Britain, Germany, Netherlands. So, the question still remains whether Mr. 8's models would predict that immigrants from these countries in a democracy would lead to "many countries with new borders", say separate nations of Massachusetts, Pennsylvania, and New York. In Europe, of course, these people's homelands remain separate countries so, maybe, there is something different about the US relative to Europe, e.g., near 100% population of immigrant origin or emphasis on individualism over collectivism.

In any event, an assertion that Mexican immigrants will have just as much difficulty in assimilating as non-Protestant Irish and non-Protestant, non-Northern European Italians hardly seems like an argument against immigration, except maybe among those that were against Irish and Italian immigration a century ago.

Comments for this post are closed


As you know, my point is that historically America had neither "open" immigration (because of costs) nor "diverse" immigration (for other reasons). That makes claims that we can extrapolate from the past to the present, suspect. When improving shipping technology made "open" immigration more possible, America closed the door.

As for Northern Europe, most of the countries in question appear to be quite successful. Indeed, Northern Europeans are quite successful in the U.S. Let me quote from

"A Scandinavian economist once stated to Milton Friedman: "In Scandinavia we have no poverty." Milton Friedman replied, "That's interesting, because in America among Scandinavians, we have no poverty either." Indeed, the poverty rate for Americans with Swedish ancestry is only 6.7%, half the U.S average. Economists Geranda Notten and Chris de Neubourg have calculated the poverty rate in Sweden using the American poverty threshold, finding it to be an identical 6.7%."

A quick check of state level statistics supports this conclusion. The upper-Midwestern states are all characterized by good schools, low crime rates, high incomes, low unemployment, low uninsured populations, a relatively small underclass, etc.

A useful parallel is that (apparently) Japanese Americans have life expectancy greater than Japanese living in Japan (apparently).

Comments for this post are closed

Comments for this post are closed

I'm vaguely pro-immigration, so forgive me if this doesn't seem fully thought out.

If you read and buy into Judith Harris and The Nurture Assumption, it seems like the real question for assimilation is whether immigrants have critical mass to have children grow up in peer groups of the immigrant culture. If they do, the immigrant culture will be preserved, if they don't they will be assimilated.

So it seems like the key question for whether or not you'll see conversion to American culture or displacement of American culture is whether the rate of immigration is sufficiently high that immigrant communities form (which it clearly is), and whether technology changes anything about the critical mass necessary for a peer group to form (probably technology makes the necessary community size smaller, and lets them be more spread out and still stay interconnected). One might argue that, for example, Mexican immigrants today get to live with their immigrant friends, watch Mexican TV, experience Mexican Internet, and otherwise immerse themselves in their immigrant culture. This probably wasn't true of Irish immigrants 100 years ago. So you probably should see considerably less assimilation today and in the future than you saw in the past.

All true. The biggest immigration issues aren't really economic (or just economic), but what kind of country we want to have. In 50 years, do we want our country to be more like Brazil (racially divided, radical inequality, mass poverty, economically unsuccessful, generally dysfunctional, etc.) or more like Finland.

The best single essay on the subject was written by Samuel Huntington (Clash of Civilizations). See "The Hispanic Challenge" (,7). A few quotes, but read it all.

"The persistent inflow of Hispanic immigrants threatens to divide the United States into two peoples, two cultures, and two languages. Unlike past immigrant groups, Mexicans and other Latinos have not assimilated into mainstream U.S. culture, forming instead their own political and linguistic enclaves -- from Los Angeles to Miami -- and rejecting the Anglo-Protestant values that built the American dream. The United States ignores this challenge at its peril. "

"The persistence of Mexican immigration into the United States reduces the incentives for cultural assimilation. Mexican Americans no longer think of themselves as members of a small minority who must accommodate the dominant group and adopt its culture. As their numbers increase, they become more committed to their own ethnic identity and culture. Sustained numerical expansion promotes cultural consolidation and leads Mexican Americans not to minimize but to glory in the differences between their culture and U.S. culture. As the president of the National Council of La Raza said in 1995: "The biggest problem we have is a cultural clash, a clash between our values and the values in American society." He then went on to spell out the superiority of Hispanic values to American values. In similar fashion, Lionel Sosa, a successful Mexican-American businessman in Texas, in 1998 hailed the emerging Hispanic middle-class professionals who look like Anglos, but whose "values remain quite different from an Anglo's."


Castañeda cited differences in social and economic equality, the unpredictability of events, concepts of time epitomized in the mañana syndrome, the ability to achieve results quickly, and attitudes toward history, expressed in the "cliché that Mexicans are obsessed with history, Americans with the future." Sosa identifies several Hispanic traits (very different from Anglo-Protestant ones) that "hold us Latinos back": mistrust of people outside the family; lack of initiative, self-reliance, and ambition; little use for education; and acceptance of poverty as a virtue necessary for entrance into heaven. Author Robert Kaplan quotes Alex Villa, a third-generation Mexican American in Tucson, Arizona, as saying that he knows almost no one in the Mexican community of South Tucson who believes in "education and hard work" as the way to material prosperity and is thus willing to "buy into America." Profound cultural differences clearly separate Mexicans and Americans, and the high level of immigration from Mexico sustains and reinforces the prevalence of Mexican values among Mexican Americans.

Continuation of this large immigration (without improved assimilation) could divide the United States into a country of two languages and two cultures. A few stable, prosperous democracies -- such as Canada and Belgium -- fit this pattern. The differences in culture within these countries, however, do not approximate those between the United States and Mexico, and even in these countries language differences persist. Not many Anglo-Canadians are equally fluent in English and French, and the Canadian government has had to impose penalties to get its top civil servants to achieve dual fluency. Much the same lack of dual competence is true of Walloons and Flemings in Belgium. The transformation of the United States into a country like these would not necessarily be the end of the world; it would, however, be the end of the America we have known for more than three centuries. Americans should not let that change happen unless they are convinced that this new nation would be a better one.

Such a transformation would not only revolutionize the United States, but it would also have serious consequences for Hispanics, who will be in the United States but not of it. Sosa ends his book, The Americano Dream, with encouragement for aspiring Hispanic entrepreneurs. "The Americano dream?" he asks. "It exists, it is realistic, and it is there for all of us to share." Sosa is wrong. There is no Americano dream. There is only the American dream created by an Anglo-Protestant society. Mexican Americans will share in that dream and in that society only if they dream in English."

Comments for this post are closed

This seems to me a circular definition of "critical mass". When the Irish and Italians first immigrated and lived in their ethnic neighborhoods, one could say, "See, there are so many of them that they are self-segregated and not assimilating, unlike earlier immigrants." Later, after they got elected President or played center field for the Yankees and married Marilyn Monroe, then one could say, "See, there were few enough of these immigrants such that they were able to assimilate, unlike today's immigrants."

Saying earlier immigrant groups assimilated because there weren't enough of them to reach "critical mass" is like saying that a stimulus program didn't work because we didn't do enough of it.

Comments for this post are closed

By the way, here is Ben Franklin on white, Protestant, Northern European German immigrants reaching critical mass and not speaking English:

Comments for this post are closed

I'm just saying I think the rate matters to assimilation, and that technology acts to make it easier for immigrants to avoid assimilation. Critical mass ought to be a function of technology. Ease of approaching critical mass is a function of the rate. If you don't care about assimilation, this doesn't matter.

But your original question was effectively "what's different today?" Well, the rate of immigration is relatively, but maybe not unprecedentedly high, and technology is acting to make assimilation considerably slower. I imagine, say, Tibetan immigrants are assimilating just fine, because they can't achieve this kind of isolation for their kids even today.

Comments for this post are closed

@BC if you can point me to where people were doubting the assimilation prospects of the Italians in 1996, that would interest me.

I do not think that was very common then.

Comments for this post are closed

Actually, my original question was not, "What's different today?", although I would anticipate that many immigration opponents would argue that today is much different from yesterday in terms of US immigration. My question was simpler. Mr. 8 asserted that based on "5000 years of history", there are many models that suggest that a democratic/republican nation with many immigrant groups would quickly degenerate into "many countries with new borders". My question was why, under those models, that didn't happen to the United States?

To your point on technology though, I would anticipate that many mass media technologies, such as broadcast television, would actually accelerate rather than interfere with assimilation. (We can all obsess over American Idol together.) Also, since Mr. 8's models are based on "5000 years of history", it would seem that the alleged incompatibility of democracy and immigration would not depend on modern technology.

Comments for this post are closed

Considering how radically different the composition of the country is and how radically different our society has changed, perhaps immigration has had a larger effect on the changing of our society than one often realizes. How radically different would the nation be if we had closed borders in 1848?

Comments for this post are closed

For a given rate of immigration and a given technology, I think you would see more rapid assimilation with more diverse immigrants. I.e., the same number of people but with fewer large blocks. You might similarly be able to help assimilation if immigrants were incented to spread throughout the country rather than cluster in particular states, cities, and neighborhoods.

Comments for this post are closed


"Later, after they got elected President or played center field for the Yankees and married Marilyn Monroe, then one could say, “See, there were few enough of these immigrants such that they were able to assimilate, unlike today’s immigrants.”"

Two pretty obvious points. The U.S. ended mass immigration around WWI. It was universally recognized (back then) that the immigration cutoff greatly facilitated assimilation. Of course, the U.S. back then was vastly (and I do mean vastly) better able to promote assimilation than it is today. Back then we had functional schools, tight law enforcement, vast assimilationist pressure, overt disdain for “diversity”, booming labor markets, near zero welfare, intact families, a common language, well paid low-skill jobs (with unions and benefits), etc. None of that is true now.

The second point is the current immigration wave is stunningly less diverse than has ever been the case before. Historically, no one country or language dominated immigration. The historic diversity of the immigrants helped to ensure that no separatist claims would arise. Now we have strikingly non-diverse immigrants with irredentist claims against the U.S. and a well-developed oppositional racial identity (check out what NCLR stands for).

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

"if you have a strong authoritarian government, such as a monarchy, many different cultures and nationalities can be held together."

Ottoman empire, does that ring a bell to you?

Well the Ottoman empire was an empire not a kingdom and it didn't really have a particularly strong government. So it's probably a poor example.

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed


I real life their are many powerful arguments against immigration. However, they are all predicated on some idea of what defines "good" vs. "bad". Essentially all of the arguments in favor of immigration amount to

"Its good for immigrants and for (my) special interest group"

Essentially all of the arguments against immigration amount to

"Its bad for America and the vast majority of Americans"

Note the lack of overlap in the implied constituencies. The Open Borders crowd cares immigrants (and mostly notably themselves) but isn't particularly concerned with America as a whole or the vast majority of Americans. The exact opposite applies to the restrictionists.

So who do you care about? If you care about America and Americans, then the following notes should be important to you..

1. The 1997 National Academy of Sciences study found that each low-skilled immigrant costs $89,000 over the course of his/her lifetime.

“The NRC estimates indicated that the average immigrant without a high school education imposes a net fiscal burden on public coffers of $89,000 during the course of his or her lifetime. The average immigrant with only a high school education creates a lifetime fiscal burden of $31,000.”

2. There is little evidence that the children, grandchildren, and great-grandchildren of illegals will do much better. Samuel Huntington looked at this subject in his book, “Who Are We”. See Table 9.1 on page 234. The bottom line is that educational attainment rises from the first to the second generation and then plateaus at levels far below the national average. For example, even by the fourth generation only 9.6% of Mexican-Americans have a post-high school degree.

3. The Heritage foundation found that low-skill immigrant households impose huge tax costs on Americans. See “The Fiscal Cost of Low-Skill Immigrants to the U.S. Taxpayer”. The summary is

“In FY 2004, low-skill immigrant households received $30,160 per household in immediate benefits and services (direct benefits, means-tested benefits, education, and population-based services). In general, low-skill immigrant households received about $10,000 more in government benefits than did the average U.S. household, largely because of the higher level of means-tested welfare benefits received by low-skill immigrant households. In contrast, low-skill immigrant households pay less in taxes than do other households. On average, low-skill immigrant households paid only $10,573 in taxes in FY 2004. Thus, low-skill immigrant households received nearly three dollars in immediate benefits and services for each dollar in taxes paid. A household’s net fiscal deficit equals the cost of benefits and services received minus taxes paid. When the costs of direct and means-tested benefits, education, and population-based services are counted, the average low-skill household had a fiscal deficit of $19,588 (expenditures of $30,160 minus $10,573 in taxes).”

4. Heather MacDonald has written extensively on the bleak realities of mass unskilled immigration. I recommend “Seeing Today’s Immigrants Straight”. Key quote

“If someone proposed a program to boost the number of Americans who lack a high school diploma, have children out of wedlock, sell drugs, steal, or use welfare, he’d be deemed mad. Yet liberalized immigration rules would do just that. The illegitimacy rate among Hispanics is high and rising faster than that of other ethnic groups; their dropout rate is the highest in the country; Hispanic children are joining gangs at younger and younger ages. Academic achievement is abysmal.”

5. Edward P. Lazear’s (CEA / Harvard Economics) paper “Mexican Assimilation in the United States” has a wealth of statistics showing the raw deal from south of the border. Summary quote.

“By almost any measure, immigrants from Mexico have performed worse and become assimilated more slowly than immigrants from other countries. Still, Mexico is a huge country, with many high ability people who could fare very well in the United States. Why have Mexicans done so badly? The answer is primarily immigration policy.”

See also “Lazear on Immigration”. Money quote

“Immigrants from Mexico do far worse when they migrate to the United States than do immigrants from other countries. Those difficulties are more a reflection of U.S. immigration policy than they are of underlying cultural differences. The following facts from the 2000 U.S. Census reveal that Mexican immigrants do not move into mainstream American society as rapidly as do other immigrants.”

Comments for this post are closed

Comments for this post are closed

Are the jobs in the "tech" industry that are hiring immigrants really capital intensive? At a place like, say, Google there is a huge amount of accumulated software capital that developers use, but it can be duplicated at zero marginal cost. Perhaps you would call that organizational capital? After that there are the developer's desktop machines and all the servers that Google uses, but those cost a relatively small amount compared to the employee's salaries. Especially when compared to jobs I tend to think of as capital intensive, such as car manufacturing.

I don't buy Tyler's argument. He's just clubbing together way too much variety of effects into "labor".

Say, I'm a mining equipment manufacturer and suddenly immigration laws relax somewhat. I'm more likely to hire skilled talent liberally from global markets. Say, good CAD designers, good engineers, R&D scientists, poly-lingual marketing staff etc.

I doubt the firm would suddenly de-automate its assembly line and go back to minimum wage men painting equipment with brushes and pails.

Ergo, I can think of several cases where migration can indeed boost production of high-tech items.

Alternatively, Tyler is thinking at some margin of immigration relaxation which is totally unrealistic e.g. an influx so large that you' get Americans to work at $1 an hour. Maybe then painting with brushes and pails does get economical.

I doubt the firm would suddenly de-automate its assembly line and go back to minimum wage men painting equipment with brushes and pails.

That statement misses the point. What a manufacturing company will decide is what path they take for their next production line:
Scenario a) Highly automated production line $20M cost, staff: 4 workers per shift, 3 shifts per week ($25/hour workers)
Scenario b) Automated production line $8M cost, staff: 16 workers per shift, 3 shifts per week ($15/hour workers)
Scenario c) Conveyor line $2M cost, staff: 32 workers per shift, 3 shifts per week ($9/hour workers)


Even if they went from (a) to (c); so long as the immigration relaxation also resulted in them getting better engineers, designers etc. we end up with a higher production / sales of said mining equipment.

Ergo, immigration relaxation can indeed boost our production of capital-intensive goods!

LOL, but you don't need better engineers for option C. The skill set to set up a system of Conveyor belts is much less than the skill set to set up a highly automated production process. So why would the company pay for better engineers when choosing C allows for the use of worse engineers? After all, better engineers will always cost as least as much as lower quality engineers.

Comments for this post are closed

To clarify:

I believe that increasing the pool of high skilled immigrants would probably result in companies hiring more good CAD designers, good engineers, R&D scientists, poly-lingual marketing staff. However, increasing the pool of low skilled immigrants would probably result in companies expanding their usage of labor intensive manufacturing and reduce their hiring of more good CAD designers, good engineers, R&D scientists, poly-lingual marketing staff.

Comments for this post are closed


I suspect we are arguing at cross purposes. The firm uses the better engineers not to build its convener belt system but to design and sell better mining equipment!

Making broad statements like "More immigration won't help capital intensive products" is silly I think. It depends. What sort of immigrants? Engineers? High school dropouts? How many of them? etc.

My point is Tyler's painting with too broad strokes.

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

A good example to illustrate Tyler's theory is what happened when Cesar Chavez's United Farm Worker union forced up lemon pickers' wages in California in the 1960s. Chavez didn't get any traction until the bracero guest worker program was shut down in 1964, but then he had a good decade and a half until illegal immigration inundated his union.

When Chavez drove up wages, the lemon growers immediately junked the wood ladders and oaken buckets they furnished their employees and replaced them with lightweight aluminum ladders and


Another example is tomato picking. Up until the early 1960s all tomatoes in California were picked by hand using Braceros (temporary workers from Mexico). Eventually the Bracero program was ended because of numerous abuses. The growers stated that no tomatoes would be grown in California without Braceros (greedy crybabies then, greedy crybabies now).

In fact tomato picking machines were immediately introduced and production soared. Back then the U.S. was a world leader in agricultural technology. Now we are a definite laggard. That's what 30 years of Open Borders does to a country.

World leader in using? Or manufacturing?


"World leader in using? Or manufacturing?"

Back then, both. Now neither. Nothing new about this. From

"What the history of the electric dynamo teaches about the future of the computer"

"David showed that World War I, which led to immigration controls and choked off the supply of cheap but untrained immigrant workers, was one of the spurs to make these changes. U.S. productivity growth eventually leapt in the 1920s, four decades after the commercialization of electricity. Productivity growth rates in U.S. manufacturing in the 1920s were more than 5 percent per year, a rate that makes the "new economy" look laughable, at least for now."

I'm no farmer but let's take manufacturing leadership in Agricultural Technology that you mention.

John Deere, Vermeer, Case (now CNH), Agco, Hagie, Catterpillar come to mind. All are American. I bet there are great foreign companies too. To say we are a laggard in ag manufacturing seems a tad panicky.

If not us who? Name a nation with more ag equipment manufacturing prowess than the US?

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

I suppose another way of looking at it is that sectors we consider to be "tech" are those where capital isn't a constraint on production, where you need equal numbers of mill workers and lathes, but a multiplier. Your capital base in servers or silicon fabs or such produces a number of units of your product with the traditional labor constraint, but that labor is only a tiny fraction of the firm's consumption of labor. The vast majority of the labor consumed by a tech firm is devoted to increasing the value of each unit of production. Unless the immigrants are going to maintain the servers or work the fabs, there's no reason to think that Rybczynski should apply.

Comments for this post are closed

"It is commonly claimed for instance that migration to the United States does not lower American wages much if any (I agree by the way, as do most economists)."

Except for economists who actually specialize in the subject ...

For example, here is labor economist Philip Martin of UC Davis on the history of wages in the lemon fields of Ventura County:

"Growth and change in Ventura county agriculture between 1980 and 2000 have not improved conditions for farm workers. In the late 1970s and early 1980s, most lemon pickers were unionized and, when the minimum wage was $2.65 an hour in 1978, many had piece-rate earnings of $5 to $7 an hour and worked 800 to 1,200 hours a year. In 2002, with a state minimum wage of $6.75, many workers earn $7 to $10 an hour, and hours of work have fallen, reducing annual earnings."

... "However, between 1942 and 1964, the packing houses that controlled citrus harvesting switched to Bracero workers, and Ventura county citrus became dependent on Mexican guest workers who were housed in barracks style camps. ...

The mid-1960s were a “time of transition” for labor in Ventura county, as lemon growers responded to the end of Mexican guest workers by reducing quality standards, introducing a piece rate wage system that aimed to standardize worker earnings even as grove conditions changed, and increased worker productivity with new clippers, lighter synthetic bags, aluminum ladders, and larger bins. Worker benefits were introduced or increased, as employers offered health and pension benefits and UI benefits to stabilize the work force, or to keep the same workers returning year-after-year.

... After the Bracero program ended, CGA developed “modern personnel practices” to recruit, reward, and encourage the return of the best pickers. The result was a win-win situation: CGA pickers saw their average piece rate earnings rise from $1.77 an hour in 1965 to $5.63 an hour in 1978, and average annual earnings rise from $267 (151 hours) in 1965 to $3,430 in 1978 (609 hours).[4] The number of pickers employed at CGA (W-2 statements issued) fell from 8,517 in 1965 to 1,292 in 1978, as average productivity rose sharply—from 3.4 boxes an hour in 1965, to 8.4 boxes an hour in 1978 (Mamer and Rosedale, 1980); CGA in 1966 expanded from serving 3 to 7 packinghouses.

CGA was able to add benefits for the most productive pickers and keep costs of picking reasonably low. The direct or wage cost of picking each box (of lemons18 bags or boxes = 1 bin) rose from $0.53 to $0.66 between 1965 and 1978, and the total cost per box rose from $0.63 to $1.09—total costs rose with the introduction of benefits that ranged from paid vacation to health insurance.

Comments for this post are closed

"Ethan Lewis (pdf) considers the famous Cuban boat lift example, when the influx of immigrants did not lower wages ..."

The Mariel boatlift from Cuba to Miami of May 1980 was cited by Berkeley economist David Card as showing that an increase in labor did not drive down wages because wages in Miami over the next few years were not depressed relative to four other cities he considered. What Card forgot about (much like what Steven Levitt forgot the Crack Wars in concocting his abortion-cut-crime theory) was the impact of cocaine on turbocharging Miami's economy in the early 1980s, the years of "Scarface" and "Miami Vice."

For details see "George Borjas v. David Card's Unworldly Philosophy" at


The other factor is labor mobility. To use a simple analogy, draining water from one end of a bathtub doesn't cause that side of the bathtub to be lower than the other side. The water moves.

The definitive paper on the subject is

"The Labor Demand Curve is Downward Sloping: Reexamining the Impact of Immigration on the Labor Market"

The abstract reads

"Immigration is not evenly balanced across groups of workers that have the same education but differ in their work experience, and the nature of the supply imbalance changes over time. This paper develops a new approach for estimating the labor market impact of immigration by exploiting this variation in supply shifts across education-experience groups. I assume that similarly educated workers with different levels of experience participate in a national labor market and are not perfect substitutes. The analysis indicates that immigration lowers the wage of competing workers: a 10 percent increase in supply reduces wages by 3 to 4 percent."

Comments for this post are closed

Comments for this post are closed

The Rybczynski theorem, as TC describes it, does not seem to rely at all on whether the influx of labor comes from a different country or not, just that there is an increase in the supply of labor. Thus, wouldn't within-country migration provide many examples from which we could collect empirical data? The migration from the Rust Belt to the Sun Belt would be one example. What was the effect on wages and capital-intensive production in both the Rust Belt and the Sun Belt as a result of the positive "immigration" to the Sun Belt and negative immigration from the Rust Belt? Also, what happens to wages and capital-intensive production as population immigrates from rural areas to cities and from urban areas to suburban areas? If increases in labor supply itself caused wage declines, we would expect rural areas to have the highest wages, which doesn't seem to be the case. Has capital-intensive production flowed from cities to rural areas and from suburban areas to urban areas, in the opposite direction of the immigration?

Comments for this post are closed

I'm a physicist, not an economist, so maybe I'm missing something...

If you have a shift to the labor-intensive sector because of the influx of labor, what about the change in prices for capital goods produced in this sector? It seems to me that if capital goods decline in price, the capital-intensive sector can do the same amount of work with less investment. This may keep the output from that sector the same, or may even increase it, no?

Wouldn't the long-run effects end up better because of this in either case? Doesn't this make it less likely there's long-term negative effects?

I read this blog (and other economics blogs) to learn more. Feel free to educate me.

Bravo. If a shift toward labor-intensive sector comes as a result of the influx of labor, it must means that the high-tech capital goods they replaced can find better, more profitable (and hence more productive) use elsewhere. I mean, if those machines are not really needed elsewhere in the production chain, then why are they more expensive than the labors in the first place?

If there were "better, more profitable (and hence more productive) use elsewhere" why is the capital in the position you say it's in?

That is precisely my point isn't it? Capital goods only generate profit when they are in the right spot i.e. the area of industry that needed them most. If a company finds itself losing profit by investing in machines instead of labor, that means either 1) Those machines can be used more productively elsewhere or 2) The costs of producing those machines actually exceed the labor-saving benefits that can be derived from them.

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

The Rybczynski theorem does not talk about prices, but quantities. Refer to the Wikipedia article. See also this: Or to put it most provocatively: In an open economy with product prices set in global markets, the derived demand for labor is infinitely elastic.

Thus, open immigration for a country like the USA will cause no wage inflation even when the entire world wants to move to the USA (as it will, since employers will welcome them). But, when and if this happens, and as it happens, the quantity (not price) of capital goods intensive industries in the USA ("like car making") will fall. Logically then, unless the Japanese step into the breach, the prices of these capital goods will actually rise a bit. But we'll be better off since everybody will be a programmer, everybody will be a brain surgeon, everybody will be a hair dresser, everybody will be a world class economist, everybody will be a master craftsman, or everybody will know where to find these people, for the price of peanuts (the Production Possibility Frontier will have shifted outwards).

Comments for this post are closed

Comments for this post are closed

The factor-price equalization assumption behind the R. theorem also implies that immigrants receive no wage boost when they come to the U.S. An assumption clearly at variance with the facts.

Many models imply that wages won't fall with greater immigration so the empirical fact that wages don't fall doesn't imply the R. theorem is the explanation.

Also, given the theorem, it's correct to say that output from the capital intensive sector falls. So what? The total amount of capital does not fall, it just moves to the labor-intensive sector as that sector expands (you don't say otherwise but worth mentioning). I am not buying a vague story about externalities which doesn't make any sense at this level of aggregation.

It also is claimed that migration will help us boost our production of capital-intensive goods, such as tech products

Are Microsoft Office, AutoCAD or Matlab tech products in the Tylerian sense. Do we call those labor intensive or capital intensive.

How about something like an I-Phone whose American component is design, engineering, sales, services and marketing. Is that labor or capital intensive?

Maybe the association of tech-products as capital-intensive is what's the flawed premise.

See my comment. I think these are "labor-intense", but clearly that's not right. Human capital intensive is a better way of putting it. This is one of the more common explanations to Leontief.

What is the marginal labor input for 100,000 additional copies of Autocad sold? Very little.

What is the marginal capital input for 100,000 additional copies of AutoCAD sold?

Comments for this post are closed

Robots don't write the code, Derek.

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Comments for this post are closed

Every result from Heckscher-Ohlin, including Rybczynski, is a long run result. It depends on capital moving from sector to sector. The corresponding short run model is called the Specific Factors Model and it predicts a decline in real wages upon immigration. After capital moves, real wages are restored, but labor will have lost out until then.

Comments for this post are closed

In other words, the core model for constant or near-constant wages, following immigration, also implies that capital-intensive production should fall, following an exogenous influx of labor.

Except that total production should be going up (you are increasing the population of workers, after all) so why can't both capital-intensive and labor-intensive production rise, with labor-intensive production rising more?

Comments for this post are closed

Is this a concern only if you hold capital investment equal, along with a labour supply increase? When I am making (awfully small) investment decisions, I gravitate towards regions with at least a moderately growing population. Which is why the US looks good even now, compared to much of Europe. So it could be that more immigration pulls in more (foreign) capital along with it, meaning overall not necessarily any less capital in capital intensive industries.

Comments for this post are closed

From Slaughter and Hanson:

1."It is important to emphasize that (6) and (7) are sufficient, but not necessary, conditions for
FPE. If there are increasing returns to scale, regional differences in production technologies, or
externalities in production, then regional unit factor requirements may not be equalized, even if
there is regional FPE. Equal unit factor requirements across regions requires not just equal factor
prices, but also the absence of significant scale effects, externalities, or arbitrary cross-state
differences in production technologies. In testing for FPE using (7), we are forced to assume that
these additional effects are inconsequential for relative regional factor prices."

They go on to add that they can control for Hicks-neutral technology differences across states, so long as those differences are homogeneous within states. "We allow wages to be relatively high in California, for instance, as long as this is due to factors in California being uniformly more productive in all industries (for whatever reason)."

2. "Even under standard assumptions, OLS estimates of b in (8) will not be efficient.
Efficiency is of great concern since, for a true b that is close but not equal to one, standard errors
that are too large will cause us to fail to reject relative FPE when it is in fact false. Generalized
Least Squares (GLS) techniques, such as the Seemingly Unrelated Regression (SUR) framework,
are the standard approach to obtain efficient coefficient estimates in this context. One potential
problem with the SUR estimator is it may perform poorly in small samples, as in our case with 35
observations per factor and per state. Unreported results bear out this concern. For several
states, SUR estimates of b are much lower than OLS estimates."

Pretty much any econometric study, arguing on either side, runs into these issues. But still.

Comments for this post are closed

My intuition would be that the relative share of capital-intensive output would fall but not necessarily decline.

Comments for this post are closed

Comments for this post are closed