America in the Gilded Age was a starkly unequal place, not just in terms of inequality between people but inequality between regions. Long-settled, fast-industrializing states in the Northeast were far richer than those of the West or the South, which had many fewer factories, railroads, and other kinds of capital goods that allowed for productive work and high wages. But around 1880 that began to change, and for 100 years, income gaps between states slowly converged at a rate of about 1.8 percent per year.
But since 1980, that process has began to slow, and over the past decade it’s essentially stopped entirely. Today, Massachusetts’s GDP per capita is about double what you find in Mississippi — roughly equivalent to the gap between Switzerland and Slovakia — and it’s not getting any narrower.
Phillip Longman of New America’s Open Markets program has been arguing for some years now that Reagan-era shifts in the federal government’s attitude toward corporate concentration are to blame. This is one of several arguments that’s helped inspire Democrats to start calling for a rethink of federal antitrust policy. But new empirical research from Peter Ganong of the University of Chicago and Daniel Shoag of Harvard’s Kennedy School of Government suggests the issue is more complicated than that. After all, even as the richest cities have gotten richer on a per capita basis, their share of aggregate national output has stagnated because their populations are growing slowly.
Ganong and Shoag argue that the slowing population growth in rich cities and the slowing of regional income convergence are intimately linked trends.
Less skilled workers used to move to rich states to increase their wages. That lowered average income in the rich states while raising it in the poor ones, as people’s natural tendency to move toward economic opportunity helped drive nationwide convergence of wages and incomes. But in the contemporary United States, zoning restrictions that prevent adequate levels of house building mean that much of the higher incomes earned in rich states simply pass through in the form of higher housing costs.
For skilled workers, this trade-off is worth it, but for the working class, it generally isn’t. Consequently, working-class people have begun to move out of the rich states and toward the cheap ones — throwing the pattern of convergence into reverse.
Two big shifts in migration and economics
This set of four charts in Ganong and Shoag’s paper tells the fundamental story — in the old days, there was a strong tendency for poor states’ per capita incomes to grow faster than those of rich ones and an equally strong tendency for people to move away from poor states to go live in rich ones. But in recent years, the income convergence trend has slowed and the migration pattern has reversed.
People move, of course, for non-economic reasons. You can see clearly on these charts that the warm weather of Nevada and Arizona causes those states to punch above their weight in terms of migration in both eras. But the overall pattern is striking. Lots of people used to move to rich places like California, Maryland, and the tri-state area around New York City. These days, very few people move there, even though the typical resident of the South or Midwest could earn more by moving to a rich city.
The reason is that these states are also more expensive, and for working-class people the higher costs are no longer worth the higher wages.
This chart shows that until 1990 or so, both skilled and unskilled workers could improve their standard of living, even considering housing costs, by moving to a high-income state. But the net gains for unskilled workers began to diminish sharply, and by 2010 a typical low-skill household was actually worse off in a high-income state due to the even higher housing costs.
Traditionally, in other words, both lawyers and janitors earned more in the New York City area than they did in the Deep South. Today, “lawyers continue to earn much more in the New York area in both nominal terms and net of housing costs, but janitors now earn less in the New York area after subtracting housing.”
The result is that less skilled workers now tend to eschew the highest-wage, highest-cost locations — creating a powerful counterpressure to other forces that would otherwise drive regional income convergence.
The paradox of regional inequality
This and other lines of recent research tend to indicate that the gains to increasing the housing supply (whether through zoning changes to allow more market-rate housing or through the direct construction of social housing) would produce large economic benefits. Regional inequality would be reduced, as the pattern of state-level income convergence restarted. Ganong and Shoag also believe that about 8 percent of the increase in individual-level inequality can be explained through this mechanism. Meanwhile, overall GDP would be about 9.5 percent higher, and the structural increase in the capital share of national income would be greatly reduced.
In short, with more elastic housing supply, the United States would be richer on average, and the gains would be disproportionately concentrated among poorer people and poorer states.
But there is a paradoxical aspect to this. The housing fix for regional inequality entails more rather than less concentration of economic activity in rich coastal metro areas. The mechanism is that with a greater supply of housing, the working-class share of the population of these metro areas would grow disproportionately — dragging per capita incomes down while pulling them up in poorer places. Sunbelt and Rust Belt cities would be richer but smaller, while coastal ones would be bigger.
This would leave almost everyone better off, but it’s not exactly the political solution to the problem of regional inequality that elected officials are looking for. To get that job done, politicians may need to look at more direct solutions like moving white-collar government work to cities that have suffered population decline or creating new universities in declining areas.