Helping Workers Relocate

One of the strengths of the American economy is that our country is extremely large and diverse, and yet homogenous enough that a person can move from Missouri to Maryland or from Maine to Montana without facing extraordinary hurdles. Compared to European countries the United States offers more places a person can move to (a bigger country) and also a population with a higher propensity to move. That helps people adjust to the ups and downs of economic life. But as Enrico Moretti writes in the Wall Street Joural, our main public programs to help the unemployed don’t build on this strength and in some ways work against it since you get a flat check no matter where you go but you can stretch your dollar further in an economically depressed area where things are cheaper.

Moretti’s proposal is that “[u]nemployed individuals living in areas with above-average unemployment rates should receive part of their unemployment insurance check in the form of a relocation voucher” to help them with the cost of moving elsewhere.

That might help, although the politics of paying people to move away from where they live strike me as vicious potentially to the point of unworkability. It seems to me more important to actually work on the root causes here. Moretti’s choice of examples of a depressed-but-cheap labor market and a more expensive labor market are Detroit and Chicago and his basic scheme works for those cities. But Silicon Valley isn’t just expensive because unemployment is lower than in Detroit, it’s expensive because so much of its land is set aside exclusively for gigantic “residential estates” with little opportunity to build housing to accommodate an influx of working class residents. Similar, though less extreme, housing supply constraints pinch in the Seattle area and throughout the bulk of the Northeast Corridor. Nobody would have migrated west to settle the frontier if it had been illegal to build houses there, and the logic of urban relocation is much the same.

In terms of direct payments for relocation, I’m skeptical that this is as much of a win-win as Moretti says:

This policy would help even those who aren’t willing to move. If there are 1,000 unemployed workers looking for jobs in a city and only 100 job openings, the probability of each worker finding a job is one in 10. But if 500 of these unemployed workers relocate, the probability that each of the remaining workers finds a job is doubled. Unemployed workers who stay in a local labor market with high unemployment effectively impose a negative externality on everyone else in that market, while workers who move away generate a positive externality. A voucher is a way to shift this calculus in favor of mobility.

That makes sense if you think of the marginal worker as having a job on an assembly line or in a coal mine, but I don’t think it applies to a modern economy in which the vast majority of employment is in services and the marginal worker is likely to work at a restaurant or a retail store. Vast in-migration to Texas over the past 20 years hasn’t created a labor market glut, it’s created a lot of jobs in the “doing stuff for all the other people who live in Texas now” sector. A depopulating metro area could find itself facing spiraling job losses until you reach a new lower-level equilibrium with fewer service providers and less specialization.