When we think about economic growth, we generally think about inventions and technology—from the combustion engine to the iPhone—and about the effects of capital accumulation, like big dams or highways. But there is another, less widely understood factor: labor mobility. During America’s “special century” of fast economic growth from 1870 to 1970, new inventions were matched with vast labor mobility, with workers moving to produce and enjoy the benefits of technological progress. Technology meant that farms did not need as many workers, and so people left and moved to cities and suburbs where new inventions meant that jobs and opportunities were available.
A decline in labor mobility—and, in particular, limits on the capacity of workers to move to hot labor markets and away from dying ones—has played a substantial role in the slow GDP growth we have seen since the 1970s. Getting America on the move again is crucial to realizing the benefits of the technological progress we have made and to increasing incomes for ordinary workers.
When there are new inventions or investments, people need to move around to get the most out of them. People need to move to the regions that have become successful with the rise of certain sectors of the economy (think Silicon Valley); and people need to leave places that no longer support as much economic activity as they once did due to technological changes or trade competition (think manufacturing cities in the Rust Belt). Economic growth does not automatically occur when there are new innovations, nor is it necessarily shared by large parts of the population. Instead, innovation creates the possibility of new economic activity and employment. But we need to change where and how we live in order to capture these gains.
But over the last 40 years, we have gotten increasingly worse at allowing and encouraging people to move to the places where they will do their best economically. Economists from the Federal Reserve and elsewhere have concluded that labor mobility has declined substantially since the 1970s. Even researchers like Scott Winship, who have found little or no change in overall mobility rates, agree that there has been a decline in mobility among poorer and less well educated citizens.
Even more pressing than general declines in mobility are specific declines in rich and poor cities. People are not leaving places that have suffered economic hits, and are not moving to places that are booming. David Autor, David Dorn, and co-authors, for instance, have found that cities that have economically declined due to increased trade competition from China have not seen matching declines in population, but instead have just had increases in unemployment. And boomtowns like Silicon Valley or New York have seen little or no increase in their share of the national population since the 1990s, and instead have just had increasing housing prices. Both of these trends—failing to leave declining places and not moving to economically vital regions—are felt most acutely among the poor and the less skilled. The internal market for labor in the United States has become sclerotic, creating both lost wealth and greater inequality.
Why? Entry limits like zoning in rich cities and regions and occupational licensing regimes have kept people out of the hottest job markets. Exit limits like public benefits that are not easily transferable among jurisdictions have meant that people—particularly the least fortunate—are less willing to leave economically struggling places. And what I’ll call “shrinkage” limits have meant that economically struggling cities like Detroit or Atlantic City have physical plants and governmental structures that do not fit their current economic situation, keeping people in place and encouraging costly bailouts.
The result is a misallocation of labor and capital, and lower output levels. Further, major parts of federal policy work less well as a result. Monetary policy is harder to execute well without mobility, as unemployment and inflation rates differ substantially across the country. And federal safety-net spending is less efficacious, as people in worse job markets require more of it. These, too, have effects on output levels, and may even worsen productivity growth indirectly.
Many proposals to increase economic growth are wildly speculative, depending on difficult-to-support assumptions about where new ideas come from, how people react to tax rates, or how extending the life of intellectual property protection affects inventors. In contrast, policies aimed at unclogging our sclerotic national labor market would almost certainly allow us to use the technologies we have more effectively, creating lots of new economic activity.
But simply because those policies are more likely to succeed if enacted does not mean they will be any easier to enact. Our internal markets are sclerotic for many reasons, including the fact that policies that curb labor mobility can also provide local benefits and have local support. But these mostly state and local policies have a negative effect on the national economy. Federal, state, and local officials, working in concert, will need to attack the problem of a sclerotic internal market on many fronts with many tools.
There are three specific reasons why decreased labor mobility is a problem, each of which affects the output we get from our existing technologies and the creation of new ideas and technology.
First, declining mobility makes forging a coherent monetary policy much, much harder. As theorists like Robert Mundell and Barry Eichengreen note, labor mobility makes having a single currency for a large expanse like the United States work. Without labor mobility, the Fed has difficulty meeting its dual mandate of reducing unemployment and keeping prices stable, needing to create one set of interest rates while trying to manage inflation in San Francisco and unemployment in Gary, Indiana. The result is that we are likely to have greater unemployment, as the Fed tries to keep inflation from rising in tight markets. Such unemployment constitutes lost capacity—people who could be working but are not—and potentially lost growth, as “hysteresis,” a loss of skills from the long-term unemployed, sets in.
Second, urban economists have increasingly shown that location is extremely important to output. Being near the right set of workers can generate “agglomeration economies” and higher wages. In the right location, workers gain from labor market depth—they can specialize more easily and are insured against the failure of one firm—from information spillovers, and by learning from nearby people. Programmers learn more when they live in Silicon Valley than they do elsewhere, and also know they can invest in specialized skills and human capital more generally with confidence that there will be a market for their skills. Enrico Moretti and Chang-Tai Hsieh estimate that the lost output from workers not being able to move to the hottest labor markets, which happens largely because of land-use restrictions, is equal to roughly 13.5 percent of GDP. The failure of people to move to hot markets constitutes lost output—people could earn more in wages than they currently do—and lost growth, as workers don’t learn from people nearby. Someone in Atlantic City may have the idea for the next great app, but unless they incubate in the human-capital-rich atmosphere of a tech hub, they may not be able to translate that vision into a new product.
Finally, federal welfare spending is less efficient when there is low labor mobility among the poor. We need to pay more for programs like Medicaid, food stamps, and the like when people earn less. If people are not moving to opportunity, they have more need for redistributive spending. Further, if restrictions on entry increase housing prices—as they surely have in places like New York, San Francisco, and elsewhere—federal aid to the poor who live in rich markets needs to be higher in order to compensate for higher costs. And if federal taxes have to be higher to pay for benefits, output suffers, as we have to tax productive behavior in order to fund these benefits.
Roughly speaking, the problem of declining mobility is (at least in part) a product of a huge number of decisions by state and local officials to pass policies that may make sense locally but that impose grave costs on the national economy. Some federal policies that curb mobility have a similar feature—the relevant policymakers and interest groups care about their specific program, but are not much concerned with questions of efficiency and growth of the overall economy.
Entry Limits: To start, rich cities and states have made it very hard for people to move in. The major limit has been land-use rules. At the local level, zoning, subdivision rules, and historic preservation rules have limited construction not only in rich suburbs, as they once did, but increasingly in big cities and even distant political subdivisions. State intervention in land-use theoretically could loosen local restrictions, but in reality, most state governmental interventions serve to create what economist William Fischel calls a “double veto”—just another mechanism for opponents of new development to defeat efforts at construction.
Land-use restrictions started to have a real effect on housing construction at the regional level in the 1970s and ’80s, the same period in which American economic growth began to slow down. Since then, economically booming regions from New York to Boston to San Francisco have seen little population growth but huge upswings in property prices. Edward Glaeser and others have estimated the “zoning tax,” the average effect of land-use rules on property prices, to be as high as 50 percent in some regions.
This has had big effects on who lives where. Peter Ganong and Daniel Shoag have shown that from 1880 to 1980 there was “convergence” in per-capita incomes between U.S. states, as people migrated from poorer states to richer ones. But this process slowed in the 1970s and ’80s, at least for states that have heavily regulated land-use regimes, and has recently stopped altogether. People just can’t move to the best labor markets and to places that are most supportive of human capital development. The result is foregone output—remember Moretti and Hsieh’s estimate of 13.5 percent of GDP!—and lost growth.
But land-use rules are not the only way places keep useful labor out. In the last 50 years, the number of jobs requiring occupational licenses has increased from 5 percent to 23 percent, including, depending where you live, jobs like animal masseuse, auctioneer, bartender, florist, interior decorator, hair braider, and scrap metal recycler. Faced with lobbying from each industry, states and localities have continually added to the list of jobs requiring licenses. As Morris Kleiner and the Council of Economic Advisers have shown, these requirements reduce supply, making prices go up and reducing employment (but increasing wages for those able to procure licenses and generating revenues for licensing institutions, thus the industry support for them). Licenses frequently do not transfer across states, making it harder for people to enter labor markets in cities and states where there is the most demand.
Exit limits: At least with respect to entry limits, there is an increasing tide of intellectual criticism. But states, localities, and the federal government also make leaving economically struggling places difficult, and this problem isn’t even on most people’s radars.
Public benefits are often not transferable across places. The last 30 years have seen states gain an increasing amount of control over federal welfare policies from Medicaid to Temporary Assistance to Needy Families. With this control, they set all sorts of different eligibility standards. Someone receiving Medicaid in one state may not move to a state with better opportunities because she would lose her health insurance. Even federal policies that are formally transferable, like housing subsidies, make moving with them so onerous that transfers rarely occur.
State and local public workers make up about 13 percent of the overall workforce. The structure of public pensions creates substantial residential lock-in. Because the pensions are largely defined-benefit and require considerable vesting periods before the employee becomes eligible, workers face real incentives not to move across states.
In cities facing substantial economic crises (think Detroit before the bankruptcy), these two effects tend to get worse over time. As private sector workers leave, both beneficiaries of public benefits and public workers remain. This gives them greater control over local politics, further entrenching their benefits and, almost necessarily, increasing taxes and/or debt to fund the programs.
There are other limits. Homeownership is heavily subsidized by the federal government, both through the tax code and through interventions in the mortgage market. High homeownership rates result in decreased mobility and increased local unemployment (although the exact reason why is somewhat disputed). This bites particularly hard in declining markets, as homeowners face declining prices and “house lock,” the inability to sell a house and pay off the mortgage. Federal intervention in the housing market seems to make communities more static.
There are many more such interventions. States and local governments want to make it harder for people to leave, or at the very least don’t care about making it easier. The federal government often makes policies through entities—think Fannie Mae or the Centers for Medicare and Medicaid Services—that do not pay much attention to those policies’ effect on, say, inflation and unemployment.
Shrinkage Limits: When a city faces economic decline, it is often left with both a physical plant—i.e. houses and offices—and a government that is too big for its current economic condition. Atlantic City, for instance, has an economy that is not coming back—gambling is now legal in too many places and it’s too easy to fly to warmer climates. But the city itself remains. A large supply of houses relative to demand means that their value drops dramatically, and workers who would leave for greener pastures end up staying because of low housing prices. And having too big a government means that poorer residents and public workers stay—they are the beneficiaries—but paying for this drives up taxes and pushes richer workers out.
We make the problems associated with the inability to shrink worse. Building codes ensure that houses are built to last and that’s a good thing. But it also makes the housing overhang problem—that is, having more houses than people who would ordinarily want to live in a city—worse when the economy dips. Mobile homes and modular buildings are often limited or banned through local regulations, reducing the numbers of people who can move at least some of their housing capital.
Our tools for shrinking local governments in economically declining cities are not great. Chapter 9 municipal bankruptcy, for instance, is modeled on corporate reorganization, not on liquidation, and includes sharp limits on the ability of a bankruptcy judge to order substantial changes in the ways local governments are structured.
There are two approaches we could use to overcome these limits and make our labor market more mobile.
The first would be to try to fix each problem directly. The challenge there, though, is that local reforms are complicated by issues of local politics—most of the existing rules that curb labor market mobility are popular with the local and state governments that enact them. And while there is a substantial case for federal involvement in each of the policy areas discussed above, there are also strong reasons to leave these powers to states, localities, and specialized parts of the federal government.
A federal government that wanted to spur labor market mobility would have to be very creative in developing interventions that vindicated the federal interest in labor mobility without displacing state and local governments. Policies could include: reduction of home ownership subsidies, which would both directly reduce a limit on exit and indirectly reduce a limit on entry, as homeowners are the major driver of zoning restrictions; greater Fair Housing Act enforcement against restrictive zoning that harms people of color; relaxing the state action immunity doctrine in antitrust law to allow federal regulators to crack down on more state-backed occupational licensing cartels; greater harmonization of eligibility rules for federal benefits; reform of Chapter 9 to give judges greater authority to encourage governmental reform and allow bankrupt cities to invest more in shrinking their physical plants; and many others. Reformers could also target local and state governments, trying to change the procedures through which these policies are made, and thus changing the outcomes. But this is hard, hard work.
Another approach would be to attempt to leapfrog the problem by directly subsidizing mobility. We could provide people who move with tax subsidies—perhaps a bonus earned-income tax credit for low-income interstate movers—which would both encourage people to move and encourage employers to seek out employees from other states. The federal government could provide cash subsidies to cities or states that relax zoning or make occupational licenses transferable across state lines, perhaps through a competitive Race to the Top style program. Or it could provide penalties, like removing the mortgage-interest deduction for homeowners in localities that do not permit enough housing construction, as Glaeser and Joseph Gyourko propose.
But before we can fix the problem, we need to acknowledge it exists. To have a booming economy, we have to have a functioning internal market for labor. The great genius of the American economy is its diversity—you never know what part is going to take off next. We need to make it possible for all Americans to move to opportunity. The estimates of the effects of mobility limits are huge—13.5 percent of the GDP is equivalent to adding an economy the size of California’s! These estimates don’t even include the new inventions that a better labor market might spur. Further, limits on mobility contribute substantially to economic inequality, both by decreasing wages for people who can’t move and by increasing the value of houses owned by rich people lucky enough to live in highly-demanded regions.
Our economy will not achieve its potential unless we get Americans moving again.