In the 19th and 20th centuries, internal migration to productive locations was one of the biggest sources of upwards mobility and economic progress1. This could be countryside to city, densely populated small farms to the huge, mechanizable expanses of the American West, or from declining towns whose economic geography no longer made sense to the bustling metropolises of the future. This sort of internal migration is normal; people want to be richer, so they move to where their skills are most valuable. But at some point in the past 30 or so years, this pattern broke.
The world as a whole, and most countries throughout human history, follow the pattern seen in Italy, Japan, and Spain on this graph. People move from low-productivity to high productivity places to improve their standard of living. This is exactly what you’d expect from both common sense and basic economic theory. But Canada, Germany, the United States, and above all France and Britain show the opposite pattern. Instead of moving towards those places where their skills are at a premium, citizens of these countries flee them in droves.
At first glance, this is baffling, like watching water flow up a hill. But a look at the list of anomalous countries suggests an explanation: Third World immigration.
Evidence Immigration Drives This
So we have a hypothesis. Is it actually true? Evidence from the United States suggests yes:
Above, I have established three key empirical results: (i) in accounting terms, foreign inflows account for a vast share (40%) of local population adjustment; (ii) this share is largely attributable to spatial correlation between migrant enclaves and local employment conditions; and (iii) despite this, foreign inflows do not significantly accelerate local population adjustment, as they crowd out the contribution of internal mobility (Amior 2023).
This applies to both skilled and unskilled labor, and the effect is large enough to explain the oft-remarked upon decline in internal mobility within the United States.
This reduction in internal migration is also partly responsible for the surprisingly small drop in wages associated with mass immigration; when immigrants arrive, not only do some of their local competitors leave (lowering labor supply in an easily-measurable way), but fewer potential competitors arrive (lowering future labor supply in a way that is easily-missed). For instance, in the famous case-study of the Mariel Boatlift, “the evidence suggests that around 50% of the wage recovery over the 1980s in Miami, relative to a number of potential control locations, is explained by internal migration, with the rest explained by other factors, such as technology adoption (Monras 2020).”
The same happened in Germany, where “low-wage workers [were] more likely to leave or not enter the workforce in response to [Czech] immigration [emphasis mine]” (Dustmann, Schönberg, and Stuhler 2016).
If the immigration is stopped, does internal mobility in the expected direction resume? Yes. In 1924, when the United States cut off almost all immigration from outside Northwestern Europe and the Americas, with a corresponding collapse in the number of immigrants, “the loss of immigrant labor was replaced on a nearly one-for-one basis by new inflows of internal migrants, as well as immigration from unrestricted countries” in urban areas, especially in industry2 (Abramitzky et al 2021).
In short, actually-existing immigration does drive locals away, and in the place where it is best measured (the United States), this effect is large enough to fully explain otherwise-puzzling changes in internal migration since 1970. Since immigration to the other countries on the list is generally less selective, and space more at a premium (even in Canada, as the inhabited portion is small), we can reasonably assume the same applies to them.
Mechanisms
If people are moving in the opposite direction of the economic “pull” factors, there must be some “push” factors driving them. They’re trying to escape. But what exactly are they running away from?
Mechanism #1: Overpopulation
The first mechanism by which immigrants drive locals out of productive areas is through higher housing prices and other local overpopulation issues, such as congestion. This mechanism is approximately origin agnostic; immigration from anywhere increases demand for housing and the number of vehicles on the road. In the United States, an immigration inflow equivalent to 1% of a city’s population increases rents and housing prices by about 1% (Saiz 2007). Counterfactually, in the absence of immigration, housing prices in productive areas would still increase, but more of the people living there would be locals, and housing prices in the rest of the country would fall.
This can be alleviated to some extent by permitting reform and better infrastructure, but there are fundamental spatial limitations to both housing and transport infrastructure in dense cities3 and doing these things would improve overpopulation issues even more in the absence of immigration.
Mechanism #2: Behavioral Differences
However, influxes of Third World immigrants often lower housing prices, at least locally4 (Sá 2015). This implies many more locals leaving productive areas than would be explained by housing prices or other population issues alone. Here, local emigration is driven by the newcomer’s behavior. In the United States, this is often called white flight, but it is not limited to white people or America. This is often caused by concern over crime and physical safety, but it doesn’t have to be. Political differences, cultural differences, disagreements about appropriate use of public space (such as littering or noise levels), or simple dissatisfaction with minority status can all cause locals to leave in defiance of the economic incentives. For instance, even West Coast Asians, who are wealthy and commit very little crime, drive white flight by greatly increasing5 the grind associated education (Enjeti 2017).
Unlike housing prices, this mechanism is strongly sensitive to the race and national origin of immigrants. Ellis Islander immigration in the early 20th century did not produce anything like white flight; there was no absolute decline in native-born Americans even in the demographically transformed East Coast metropolises, unlike what has happened in modern day California, New York City, or London. The more different an immigrant group is from the locals, the more different behavioral norms and standards will lead to native outmigration.
Who Loses?
To the extent that locals are more skilled than immigrants, which is true in almost every wealthy country, the national economy as a whole is harmed. Skilled locals are driven out of the places where their skills can be best put to use by the poor behavior and overcrowding issues brought by unskilled (and often economically inactive) immigrants. As such, labor is misallocated and national productivity falls below the counterfactual.
Furthermore, locals miss out on upwards mobility. In high-immigration countries, the age-old life story of “ambitious country kid moves to the big city to better himself” is practically gone. Instead, it is the immigrants themselves that benefit from the upwards mobility provided by high-productivity cities6.
Distorting the Immigration Debate
Immigration driving internal migration away from productivity centers doesn’t just impact individuals or the economy at large. It also warps the larger immigration discussion by making immigration look much better than it really is.
First, many debates over the economic impacts of immigration compare the fiscal impact or income of immigrants with natives. That’s how you get charts like this one:
But since immigrants cluster in high-productivity areas7, and by their presence drive natives away from them, immigrant incomes are artificially inflated relative to natives. For instance, in Britain, ethnic minorities are overrepresented by a factor of 3 in London, which has 50% higher productivity than the rest of the country. Were it not for them driving white British out of London, white British incomes would be higher and ethnic minority incomes lower than in reality. As such, the fiscal impacts of immigration and economic performance of immigrants is significantly worse than it looks on paper.
Just as importantly, “native internal migration diffuses the impact of immigration from the affected local labor markets to the national economy” (Borjas and Edo 2022). This principle applies to housing prices as well; much-touted economics papers finding that immigration has little effect on local wages or housing prices are distorted by the fact that internal migration spreads these effects over the entire economy. If this effect is properly accounted for, the downwards pressure on wages and upwards pressure on housing prices that you’d expect from supply-and-demand reappear (Monras 2020).
Conclusions
Immigration has countless effects on locals. It is impossible to enumerate and accurately measure all of them. The obvious way to check if immigration is beneficial or harmful is to see whether locals move towards them or away from them—and revealed preference is that, in Western countries, locals dislike actually-existing immigration so much that they are willing to take large pay cuts to escape it. In immigration-heavy countries, people move away from economic opportunity! This harms the national economy (especially when it is low-skilled Third Worlders driving out locals via behavior rather than housing prices), distorts the immigration debate, and has effectively destroyed one of the biggest pathways for upwards mobility in the 19th and 20th centuries. Internal migration away from productive areas should be a huge screaming red flag signaling to policymakers that the time has come for immigration restriction.
This effect was so big that essentially all of the Soviet Union’s economic gains from 1917 to 1945 came from accelerating this shift from low-productivity countryside to high-productivity manufacturing; productivity within each sector barely budged.
In agriculture, the immigrants were instead replaced by capital, which helped make American agriculture by far the most modern and productive in the world during World War II.
With current technology. Flying cars and hundred-mile-high diamond arcologies would change things.
Those locals who leave don’t disappear, they move elsewhere, so national housing prices still increase.
And destructively. As education is mostly zero-sum signaling, Asians’ propensity to grind makes childhood worse for all Americans who aspire to middle-class status.
This would not be the case if it were the immigrants themselves that turned the cities they moved to into high-productivity areas. In theory, this isn’t impossible. German and Jewish immigrants in Eastern Europe and Huguenots in Germany both created new high-productivity centers. But in practice, all of the major high productivity centers attracting Third World immigration today were high productivity long before the current wave of immigrants showed up. New York, London, Paris, and Berlin have been economic centers for centuries, and even Silicon Valley, where high-productivity immigrants are incredibly thick on the ground, was already the world’s most important tech hub by 1970. As such, locals and immigrants are effectively competing for access to the same limited set of productivity centers created by locals long ago.
Partly because of path dependency—immigrants join previous enclaves, which happen to be concentrated in high-productivity areas—and partly because if you’re going to move to another country, you may as well go where your skills are most valuable.
I'm not sure whether education is worthless or not, so I'd be interested in hearing the evidence that education is mostly signalling.