All posts by Eli Dourado

The Future of Immigration Economics

This is a guest post by Eli Dourado, a research fellow at the Mercatus Center at George Mason University and a Ph.D. candidate in Economics at George Mason University. He studies Internet governance and cybersecurity through the lens of political economy, and his Ph.D. thesis is on metapolitics. Eli Dourado maintains a blog-cum-website at

Economists like Lant Pritchett and Michael Clemens have estimated that the net economic benefits of immigration are enormous. A literature review by Clemens suggests that open borders would double global GDP. But how have the costs and benefits of immigration changed over time? And how are they likely to change in the future?

Production and Economic Institutions

In the distant past — say, from 1 million years ago to 1000 years ago — the world was much poorer than it is today. In particular, there was a lot less capital per worker. Consequently, production tended to be labor intensive and labor tended to be less specialized. Roughly speaking, one worker was as good as another. There weren’t large gains to efficiently matching workers with firms.

In addition, economic institutions were much more homogenous than they are today. Foraging (until 10,000 years ago) and farming (since) institutions are not all alike, but there was arguably less cross-sectional variation in economic institutions than there is today (here’s a review of foragers versus farmers). If you drop a forager into a foreign foraging community, he might be terribly confused about the cultural practices of the new group, but he would understand the general system of production quite well.

If this argument is correct, then the benefits of immigration in the past must have been lower than they are today. Moving a worker to a new country adds to the output of the new country and subtracts from the output of the old country. If the worker is equally productive in both places, on net, there is not much change in the total value produced due to the worker’s migration. Furthermore, the worker himself might be relatively indifferent between the two countries’ economic institutions. Indeed, it seems likely that the main incentive to migrate in the distant past might have been temporary factors like drought and pestilence, not factors like a desire to live under different economic institutions.

In contrast, as we get wealthier, we seem to be heading toward a much more capital-intensive future. This means that opportunities for specialization of labor are likely to multiply, increasing the importance of matching workers with the right capital. Workers no longer are perfectly substitutable, and they will continue to become less so. Finding exactly the right person to hire entails a global employee search, which means immigration is more economically important. And workers that invest in specialized skills will need to migrate to where those skills are needed. For example, the shale gas boom is in North Dakota today, but soon it could be in China.

Increasing wealth also means potentially more scope for variation in economic institutions. In a world where everyone is a forager or a farmer, it’s much more difficult to engage in economic regulation. For example, how do you establish occupational licensing if there is basically only one occupation? But with more labor specialization and greater inequality of productivity, it becomes possible to extract rents, redistribute wealth, tinker with incentives, control production, and so on.

Improvements in technology may extend the capabilities of authoritarian regimes. Modern transportation, communications, and organizational technology have increased the degree of control of the state. As Cowen puts it:

Assume that we had no cars, no trucks, no planes, no telephones, no TV or radio, and no rail network. Of course we would all be much poorer. But how large could government be? Government might take on more characteristics of a petty tyrant, but we would not expect to find the modern administrative state, commanding forty to fifty percent of gross domestic product in the developed nations, and reaching into the lives of every individual daily.

Likewise, new surveillance and information technologies could increase state control further, and that control could be abused in many circumstances. I have argued elsewhere that technological innovation will at the margin tend to combat this greater control — but in the mean time, the possibility of severe dystopia should not be casually dismissed.

The wider array of possible economic institutions in the future means that not only will production be more sensitive to matching considerations, but that matching people with the institutions they would like to live under will become more important. Without freer immigration, people could be stuck under economic institutions that they don’t like. These mismatches could be relatively benign — consider the wealthy American who prefers European-style capitalism. But they could also be severe. Immigration could play an increasingly important role in ensuring that political regimes match the preferences of their subjects.

Telecommunications and Culture

We are just now at the start of a telecommunications revolution, driven by a growing Internet and an explosive adoption of mobile phones. The revolution hasn’t hit all corners of the globe equally, but it doesn’t show any signs of slowing down.

As peoples become less isolated, I would expect more cultural convergence across territories. For example, because English is (for now) the dominant language on the Internet, more people in non-English-speaking countries are learning English. If some other language is dominant in the future, more people will learn that language. Either way, I expect long-run linguistic convergence to many fewer languages than are spoken today, and that at a minimum, most people will at least be proficient in a common tongue, even if they continue to speak traditional languages. The Internet is the anti-Babel.

Evidence of the decline in linguistic diversity is provided by the Endangered Languages Project, which warns that “about half of the world’s approximate 7,000 languages are at risk of disappearing in the next 100 years.” While the reduction in linguistic diversity has some costs, the greater standardization of interpersonal communication protocols means lower assimilation costs for immigrants and the societies that accept them.

Likewise, I expect that other elements of culture will also tend to converge globally as telecommunication improves around the world — at least territorially. We might see an increasing amount of cultural variation within territories even as we see less cultural variation across territories. To some extent this is the case already. I feel at home talking to libertarian economists whatever their countries of origin, but I feel somewhat confused talking to the average American. As more people come online and are better able to connect with their global communities, this trend will accelerate.

As language and culture become more territorially homogenous, the real and perceived costs of immigration will decline. It will become less obvious whether the guy moving in next door is a new immigrant or a descendant of a Founding Father. People who oppose immigration because the assimilation costs are high might have less of an argument in the future.

It’s possible that better telepresence technology will make immigration less important, simply by making it less necessary for people to be in the same territory in order to work or play together. However, the rich world today already has pretty good telepresence technology, and business executives still seem to find some benefit to physically visiting their overseas operations. I suspect that telepresence technology will still be used where necessary in the future, but that nothing will ever make physical presence obsolete.

Innovation and Fixed Costs

One way to think about the history of innovation is as a continual grabbing of lowest-hanging fruit. One dimension of low-hangingness is the (lack of) fixed costs associated with an innovation. Inventing the wheel did not have a lot of fixed costs associated with it; on the other hand, curing cancer is likely to turn out to require a large up-front investment.

Fixed costs are more economical when they can be divided over large populations. This means that the more we innovate, the more we will rely on globalized markets for further innovation. To get to the next tier of low-hanging fruit, we simply need more people involved both with the production and consumption of new innovations. We need free trade and free movement of people.

In particular, fixed costs drive much of the phenomenon of clustering of economic activity. Vibrant research communities tend to be localized. Silicon Valley has been the primary hub for information technology for 40 years, but as other high-technology areas commercialize, new hubs will develop for these areas. For example, Boston is increasingly dominant in the biomedical field.

Fixed costs make these hubs globally important. While there can be ancillary hubs in different countries for each field, it is probably only economically efficient for there to be one primary hub. To the extent that one hub is dominant, workers who specialize in the field will need to migrate there. To the extent that ancillary hubs are possible, there needs to be significant cross-pollination of ideas. An ancillary web startup community is likely to form in Buenos Aires only if it is composed of at least some people who have spent considerable time in Silicon Valley.

Innovation at a global scale can bear more fixed costs than innovation at smaller scales. Consequently, freer immigration will tend to produce greater benefits as we use up the ideas that do not require such large fixed investments.


All of these factors support net benefits of immigration that are not only large, as Pritchett and Clemens argue, but growing. In the future, it will be even more imperative that we adopt sensible policies with respect to immigration. Since the global benefits of immigration are already large, we should adopt pro-migration policies now. But the opportunity cost of restrictionist policies seems likely to grow for the foreseeable future.