Stuck! The Law and Economics of Residential Stagnation
abstract. America has become a nation of homebodies. Rates of interstate mobility, by most estimates, have been falling for decades. Interstate mobility rates are particularly low and stagnant among disadvantaged groups—despite a growing connection between mobility and economic opportunity. Perhaps most importantly, mobility is declining in regions where it is needed most. Americans are not leaving places hit by economic crises, resulting in unemployment rates and low wages that linger in these areas for decades. And people are not moving to rich regions where the highest wages are available.
This Article advances two central claims. First, declining interstate mobility rates create problems for federal macroeconomic policymaking. Low rates of interstate mobility make it harder for the Federal Reserve to meet both sides of its “dual mandate”: ensuring both stable prices and maximum employment. Low interstate mobility rates also impair the efficacy and affordability of federal safety net programs that rely on state and local participation, and reduce wealth and growth by inhibiting agglomeration economies. While determining an optimal rate of interstate mobility is difficult, policies that unnaturally inhibit interstate moves worsen national economic problems.
Second, the Article argues that governments, mostly at the state and local levels, have created a huge number of legal barriers to interstate mobility. Land-use laws and occupational licensing regimes limit entry into local and state labor markets. Different eligibility standards for public benefits, public employee pension policies, homeownership subsidies, state and local tax regimes, and even basic property law rules inhibit exit from low-opportunity states and cities. Furthermore, building codes, mobile home bans, federal location-based subsidies, legal constraints on knocking down houses, and the problematic structure of Chapter 9 municipal bankruptcy all limit the capacity of failing cities to “shrink” gracefully, directly reducing exit among some populations and increasing the economic and social costs of entry limits elsewhere.
Combining these two insights, the Article shows that big questions of macroeconomic policy and performance turn on the content of state and local policies usually analyzed using microeconomic tools. Many of the legal barriers to interstate mobility emerged or became stricter during the period in which interstate mobility declined. While causation is difficult to determine, public policies developed by state and local governments more interested in guaranteeing local population stability than ensuring successful macroeconomic conditions either generated or failed to stymie falling mobility rates. The Article concludes by suggesting how the federal government could address stagnation in interstate mobility.
author. Professor of Law, Yale Law School. Thanks to Bruce Ackerman, Ian Ayres, Bob Ellickson, Chris Elmendorf, Heather Gerken, Alvin Klevorick, Daphna Renan, Ganesh Sitaraman, David Super, and Yair Listokin for insightful comments on early drafts. Thanks to Molly Alarcon, Leslie Arffa, Aurelia Chaudhury, Daniel Sheehan, Garrett West, and Katie Wynbrandt for terrific research assistance. Thanks also to the editors of the Yale Law Journal, especially Sophia Chua-Rubenfeld. All errors are my own.
Leaving one’s home in search of a better life is, perhaps, the most classic of all American stories. We fled England in pursuit of freedom from religious persecution by the British. We moved north in the Great Migration.1 “Go West, young man,” we were told, “and grow with the country.”2 Federal programs allow us to “move to opportunity.”3
But today, the number of Americans who leave home for new opportunities is in decline. A series of studies shows that the interstate4 migration rate has fallen substantially since the 1980s.5 Americans now move less often than Canadians, and no more than Finns or Danes.6 Even the most prominent study finding no general decline in mobility does observe that mobility rates are lower among disadvantaged groups and that mobility has not increased despite becoming “more important” to individual economic advancement.7
More troubling still, Americans are no longer moving from poor regions to rich ones. This observation captures two trends in declining mobility. First, fewer Americans are moving away from geographic areas of low economic opportunity. David Autor, David Dorn, and their colleagues have studied declining regions that lost manufacturing jobs due to shocks created by Chinese import competition. Traditionally, such shocks would be expected to generate temporary spikes in unemployment rates, which would then subside as unemployed people left the area to find new jobs. But these studies found that unemployment rates and average wage reductions persisted over time.8 Americans, especially those who are non-college educated,9 are choosing to stay in areas hit by negative economic shocks. There is a long history of localized shocks generating interstate mobility in the United States; today, however, economists at the International Monetary Fund note that “following the same negative shock to labor demand, affected workers have more and more tended to either drop out of the labor force or remain unemployed instead of relocating.”10
Second, lower-skilled workers are not moving to high-wage cities and regions. Bankers and technologists continue to move from Mississippi or Arkansas to New York or Silicon Valley, but few janitors make similar moves, despite the higher nominal wages on offer in rich regions for all types of jobs. As a result, local economic booms no longer create boomtowns. Economically successful regions like Silicon Valley, San Francisco, New York, and Boston have seen only slow population growth over the last twenty-five years.11 Inequality between states has become entrenched. Peter Ganong and Daniel Shoag have shown that a hundred-year trend of “convergence” between the richest and poorest states in per-capita state Gross Domestic Product (GDP) slowed in the 1980s and now has effectively come to a halt.12
This Article will make two claims. First, it will argue that the stickiness of America’s internal labor market is a fundamental macroeconomic problem that influences the quality of monetary policy, overall economic output and growth, and the efficacy of federal safety net spending. While there is no way to determine an optimal rate of interstate migration, important federal policies—like the use of a single currency and cooperative federalist social welfare programs—rely on a substantial amount of interstate labor mobility to function.13 Further, empirical estimates show that increasing interstate migration rates, and particularly moves to rich regions, would substantially increase economic activity and welfare.14
Second, the Article will show that state and local (and a few federal) laws and policies have created substantial barriers to interstate mobility, particularly for lower-income Americans. Land-use laws and occupational licensing regimes limit entry into local and state labor markets. Differing eligibility standards for public benefits, public employee pensions, homeownership tax subsidies, state and local tax laws, and even basic property law doctrines inhibit exit from low-opportunity states and cities. Building codes, mobile home bans, location-based subsidies, legal constraints on knocking down houses, and the problematic structure of Chapter 9 municipal bankruptcy all limit the capacity of failing cities to shrink gracefully, directly reducing exit among some populations and increasing the economic and social costs of entry limits elsewhere. The effect on mobility of a few of these policies, like land-use restrictions, is already understood in the literature. But this Article is the first to recognize the pervasiveness and variety of state and local policies that limit mobility.
A number of these policies changed substantially in ways that made populations stickier during the period when mobility fell. It is not clear whether these legal changes caused declines in mobility, or simply failed to push back against “natural” changes that reduced mobility—such as an aging population, declining churn in employment, and decreasing diversity of employers by region due to the increasing economic dominance of the service sector.15 But state and local policies in part dictate where people move, particularly by keeping people out of the richest metropolitan areas and best job markets.16 Whether as a direct cause or as mere bystanders, state, local, and federal laws therefore bear some responsibility for declining interstate mobility.
Abstracting from these two claims, a key takeaway of this Article is that the success of macroeconomic policy turns in large part on state and local government interventions traditionally analyzed using microeconomic tools.17 In aggregate, these local and state policies play a substantial role in creating or failing to combat the central macroeconomic problems of our time: slow growth rates, increasing inequality of wealth and income, and the difficulties of balancing inflation and unemployment.
However, state and local policies must answer to state and local needs, which are often in tension with broader national interests. In particular, population stability—the very opposite of mobility—can be beneficial to existing residents of a local or state government. Areas with stable populations are less risky and more attractive to investors.18 Families, too, may prefer population-stable areas, where grandchildren are more likely to live near their grandparents. State and local governments that want to promote investment and the interests of local families and homeowners may thus place a premium on stability. In fact, as I (and others) have argued elsewhere, the structure and process of state and local government decisionmaking often overrepresents the voices of those local residents who care the most about stability and the least about growth.19
State and local governments have few incentives to consider broader national economic implications when writing zoning codes or establishing public pension rules. My previous work has shown how the very existence of local governments encourages people to move away from economically superior locations in order to receive a preferred package of government services.20 Where local or state governments have the power to limit entry or reduce exit, the harm to agglomerative efficiency, and thus national economic output, is substantially increased. Surprisingly, even many federal policies also fail to take national macroeconomic issues into account.21 Most of the federal policies discussed in this Article—from Medicaid waivers to Chapter 9 bankruptcy for municipalities—are developed in silos without much input from institutions like the Federal Reserve or from other officials concerned with broader questions of unemployment and inflation.
Recognizing these tensions, this Article seeks to chart a course that will help law and policymakers at every level better promote federal macroeconomic policy. As this Article will establish, the entity best situated to protect federal interests is the federal government. Ideally, the federal government should develop tools that force decisionmakers at all levels of government to consider the macroeconomic implications of their interventions. As a second-best solution, the federal government could counteract laws that reduce mobility by creating direct financial incentives for individuals to move. Of course, neither strategy will solve all the problems of residential stagnation overnight—some people simply will not want or be able to move. But the strategies recommended in this Article are an important first step in reorienting federal, state, and local law to better reflect national economic needs.
The rest of the Article proceeds as follows: Part I discusses why mobility is important on a national macroeconomic level. Part II discusses how law obstructs entry, exit, and the graceful decline of dying cities. The Article concludes by sketching a legal and policy agenda to increase interstate mobility in America.
As a preliminary matter, this Article does not purport to span the entire field of mobility studies. A rich literature addresses issues of international labor mobility, particularly the immigration law regime.22 Scholars have also extensively discussed intrastate residential mobility, especially how residents and workers in a single labor market choose between competing local governments when deciding where to live.23 International and intrastate mobility are important fields, but to focus exclusively on these two areas is to ignore a crucial aspect of American migration patterns.
This Article instead seeks to push the discussion of interstate mobility to the forefront.24 As this Part will demonstrate, interstate migration has important macroeconomic implications that deserve serious scholarly attention. The following Sections discuss the relationship between mobility and (1) monetary policy, (2) agglomeration economies, and (3) federal safety net programs. This Part will show that levels of economic activity and growth turn on the capacity of labor to move to opportunity, and that central pieces of our national economic architecture—the use of the dollar as a single national currency and the organization of our social welfare policies—rely on movement by the unemployed and poor to locations where there are both jobs and a healthy tax base.
This Article does not paint on a blank canvas. A substantial empirical literature in economics discusses the evidence and causes of declining interstate mobility.25 Yet this literature focuses largely on factors external to federal, state, or local policy—factors like the aging population or the transition to a service-based economy. That said, there is a live debate over the role of land-use restrictions in causing and shaping the decline in mobility, one which the Article will turn to in Section II.A.1. This paper does not seek to adjudicate or debunk these competing causal theories for declining mobility. Rather, it seeks to supplement conventional accounts of interstate mobility by showing how policies, mostly at the state and local level, impede moves into hot job markets and out of particularly bad ones. The Article does not attempt to draw a causal link between these policies and the decline in mobility. Whatever the cause of the decline, policies that impose additional limits on mobility are a national macroeconomic problem because, as this Part will show, the central legal and policy institutions governing the national economy rely on interstate mobility. Thus, although often overlooked by national-level figures and analysts, these local policies encouraging stagnation or distortion of interstate mobility have serious macroeconomic consequences.
By law, the only legal currency in the United States is the dollar.26 While we take this for granted today, there is no economic reason why the United States must constitute a single currency area. The United States, Mexico, and Canada could theoretically emulate the adoption of the Euro and embrace a common currency, the hypothesized “Amero.”27 Likewise, if the Constitution did not forbid it, U.S. states could create their own currencies.28
The fact that the dollar is the sole currency of the United States limits the possibilities for federal monetary policy. The Federal Reserve must attempt to match the dollar’s value and supply to the needs of the American economy as a whole, in order to achieve both full employment and price stability.29 It cannot develop different monetary policies for different regions of the country. For this reason, laws that inhibit interstate mobility diminish the capacity of the federal government to manage the economy because they make regions of the American economy less similar. State and local policies can thus make it harder for the Federal Reserve to achieve its “dual mandate” of employment and price stability.30
A vast economic literature addresses the question of when it makes sense for countries or regions to adopt the same currency.31 Research on Optimum Currency Areas (OCAs) has been extremely influential on policymaking, particularly in the debate over the creation, management, and potential dissolution of the Euro.32
But impressive and extensive as it is, the OCA literature barely addresses the way laws affect the operation of currency areas. Though the literature discusses some government outputs apart from monetary policy—particularly the role of federal taxes and spending in providing transfers across areas in times of distress—it rarely considers the many ways that laws affect labor mobility, price setting, or any of the other factors that drive the optimality (or lack thereof) of currency areas. But laws, often created at subnational levels, clearly do affect these factors. Of particular interest for this paper, legal regimes that hamper labor mobility subvert the efficacy of monetary policy.
Research on OCAs began with Robert Mundell’s pioneering 1961 piece, A Theory of Optimum Currency Areas.33 European scholars had already begun to debate moving to a common currency, but the idea was viewed as a political impossibility.34 Mundell thus felt the need to defend his novel insight—that there is nothing necessary or necessarily attractive about having a currency that matches the size of a nation-state—as more than “purely academic.”35
Mundell’s work engaged with a broader literature about the costs and benefits of floating currencies.36 Floating currencies—those allowed to appreciate or depreciate freely rather than remaining pegged to other currencies at specific exchange rates—are attractive because they allow central banks to respond to demand shocks.37 Imagine two countries, Country A and Country B, with floating currencies that trade with one another.38 If Country A faces a negative shock—say, it is a major copper producer and copper is discovered elsewhere, reducing the price of copper and the size of Country A’s overall economy—it can increase the supply of money and thereby depreciate the value of its currency against that of Country B. Circulating more money devalues all goods and prices in Country A at once, bringing prices across the economy into line with the new economic reality created by the shock. Country A no longer would have contracts priced at levels based on its previous level of wealth. This change encourages economic activity and improves the trade balance with Country B. The result: full employment in Country A.
But what if Countries A and B have tied the value of their currencies to one another? Doing so makes transactions between the countries easier, increasing trade. But if it has fixed the value of its currency and it faces a negative demand shock, Country A cannot use monetary policy to devalue all goods and wages simultaneously and maintain full employment (because doing so would threaten the fixed relationship of the currencies).39 Instead, if Country A hopes to return to full employment following a negative shock, it has to engage in “internal devaluation,” absorbing the demand shock by reducing wages and the prices of goods across the economy, contract by contract.40 But prices are sticky and contracts take time to renegotiate, meaning the return to full employment can be lengthy and difficult (think Greece over the past few years).41 The advantage of letting currencies float, then, is that doing so helps countries weather shifts in demand by adjusting currency prices rather than by forcing individuals and firms to go through the painful process of adjusting the prices of all goods and services individually.42
Mundell’s brilliant move was to apply this same logic within nation-states. A national currency (like the U.S. dollar) is just like a set of currencies pegged to one another.43 The value of the “New York dollar,” for example, is pegged to the “Colorado dollar,” the “Florida dollar,” and the “California dollar.” Having a common currency across the country has clear benefits. It reduces transaction costs by obviating the need to change money, and it decreases the financial risk associated with fluctuating currency prices.44
But the benefits of having a common currency may not outweigh the costs of forgoing an internally floating currency regime if a country’s economy is sufficiently heterogeneous. Falling oil prices harm the economies of Alaska, Louisiana, Oklahoma, and Texas, but benefit the economies of California and New York. When the Federal Reserve sees falling oil prices, it faces a dilemma: should it increase the money supply? If it does, it can alleviate unemployment in the regions that have been harmed, but only at the cost of spurring inflation elsewhere.45 But if it does not increase the money supply, economic dislocation in oil-producing states will get worse. Having a single currency only makes sense if the benefit of reduced transaction costs outweighs the cost of not having tools to respond to asymmetric shocks.46
How do we know when this will be the case? Asymmetric shocks will be less of a problem, Mundell argued, if there is enough “internal factor mobility”—the movement of labor and other factors of production from one part of the currency area to the other.47 If the unemployed can move from the harmed region to the unharmed one, monetary policy interventions become less necessary. “But,” Mundell noted, “if labor and capital are insufficiently mobile within a country . . . one could expect varying rates of unemployment or inflation in the different regions.”48 In short, mobility of both labor and capital is the factor that determines whether a single currency makes sense for a country or set of countries.
Mundell’s work generated a massive literature asking under what circumstances economies should or should not be combined into currency areas. Ronald McKinnon, for example, argued that smaller, trade-dependent economies benefit less from flexible exchange rates.49 Because small, open economies import most goods, they have less control over internal prices, and currency devaluations are therefore less effective. Openness to trade thus makes fixed or single currencies work better.50 Peter Kenen focused on the kinds of industries within countries.51 He argued that diverse economies absorb asynchronous shocks better because the losses and gains of various industries offset one another.52 These jack-of-all-trades economies therefore make better candidates for fixed exchange rates or common currencies with other economies. Paul Krugman, Barry Eichengreen, and others argued that currency zones tend to get worse over time as places specialize and thus become more likely to face asynchronous shocks.53 Mundell’s later work focused on the degree to which currency unions spurred financial integration.54 Because people and banks from, say, California own stock in and lend money to businesses in Detroit (and vice versa), the two regions end up sharing the costs of asymmetric losses and spread risk.55
The big fish in the OCA literature ultimately disagreed about its most important application: the Euro.56 Mundell and McKinnon, arguing in favor of the Euro, focused on the benefits of integration for encouraging the diversification of risk.57 Eichengreen, on the other hand, argued that the Euro would face difficulties—pointing to labor immobility, asynchronous shocks, and the limited capacity for E.U.-directed redistribution.58
With the important exception of Marcus Fleming, the classic work in the literature largely ignored the extensive role of government beyond setting monetary policy.59 The quality of monetary policy turns on factors in the real economy, and those factors respond to changes in local, state, and federal policy. The optimality of a currency area depends upon labor mobility, integration of financial markets, the diversity of regional economies, and the differences in price setting across regions. The literature has incorporated levers of macroeconomic policy, like taxes and transfers across regions. But the literature pays little attention to the fact that the degree and type of integration between economies depends on a variety of laws and policies.60
As applied here, the upshot is that state and local laws that limit labor mobility clearly reduce the degree to which the United States is an OCA.61 These policies can prevent monetary policy from simultaneously matching the needs of tight markets with rising prices (like San Francisco) and slack ones with falling prices and high unemployment (like Atlantic City).62 Indeed, there is suggestive evidence that declining mobility, among other factors, can help explain some of the challenges the Federal Reserve has faced in setting monetary policy over the last forty years. The famed destabilization of the “Phillips Curve”—the concept that there exists an inverse relationship between inflation and unemployment—since the 1970s does not show up in regional data. In other words, regional economies have continued to see relatively stable tradeoffs between inflation and unemployment—when unemployment goes up, inflation goes down—even as the national economy does not always exhibit this tradeoff. 63 This suggests that the reason monetary policy has become a less successful tool for balancing inflation and unemployment is that regional economies have diverged from one another in important ways.
Another major strand of economic theory, “agglomeration economics,” shows why interstate labor mobility is essential to the economy.64 The basic insight of this work is simple: location matters. When people and capital congregate in particular cities and regions, they learn and trade more easily, and this creates wealth and generates economic growth.65 Restrictions on mobility limit the capacity of people and capital to combine, which in turn results in fewer gains from agglomeration.66
While the history of this idea goes back quite far—at least to Adam Smith—the classic articulation of the theory of agglomeration economics comes from Alfred Marshall.67 Marshall articulated three distinct gains from co-location: reduction in shipping costs for goods, the advantages of deep markets, and information spillovers between neighbors.68 These gains offset the costs of density, referred to as “congestion” in the literature, which include higher rents, traffic, pollution, and crime.69
During the early and middle parts of the twentieth century, a great deal of academic literature argued that difficulties in matching people and jobs could be a major impediment to national economic growth.70 Most notably, in the 1950s and 1960s, a group of economists and demographers, including Richard Easterlin, Simon Kuznets, and Dorothy Thomas, published a massive three-volume study demonstrating that a lack of interstate migration could impede growth and employment.71 This literature, for all of its merits, did not provide much of a microfoundation for how migration patterns worked or why jobs did not simply move to workers rather than the other way around. And for most of the twentieth century, only a small group of urban economists studied agglomeration economies.
But the field blossomed in the 1980s and 1990s when scholars used it to better understand the economics of international trade.72 Neoclassical economics could not easily explain how and when industries would cluster in a particular city.73 Masahisa Fujita, Paul Krugman, and Anthony Venables developed a model to answer just this question.74 They showed that changes in shipping costs for intermediate goods can explain where firms locate. According to their model, firms spread out completely when shipping costs are either infinite or zero.75 In the first case, there can be no trade; in the second, cost-free shipping allows firms to take advantage of the cheapest land rents, no matter how far away from the consumer. But if shipping costs are substantial but not infinite—as in most cases—firms have an incentive to cluster close to one another. Producers of intermediate goods (like companies that manufacture axles for cars) want to move near producers of final goods (like car manufacturers). That way, only the final producers pay shipping costs.
If shipping costs begin to fall, the incentive for firms to co-locate decreases.76 For a period of time, firms will remain clustered because the cluster itself provides gains. But if shipping costs fall far enough, firms will spread out again.
The important insight from this work is that the relationship between firm location and shipping costs is nonlinear. If there are clusters of firms (i.e., cities) but shipping costs then fall enough to cause firms to spread out, they will remain spread out even if shipping costs rise again.77 Consider manufacturing agglomeration in Cleveland, Ohio. Metal working and chemical companies formed in and moved to Cleveland to be close to its dominant steel, iron, and petroleum-refining industries, and then to each other.78 As shipping costs fell, manufacturers dispersed around the country and the world, and there was no longer as large a concentration of firms in Cleveland to attract new entrants.79 So even if shipping costs later increased back to their original high levels, there would not be much reason for departed manufacturers to return back to Cleveland in their previous numbers. They would likely cluster again, if conditions were right, but there is no reason to believe they would cluster in Cleveland. All the king’s horses and all the king’s men can’t quite put Cleveland back together again.80
The history of urban location in the United States follows this model quite closely. As Edward Glaeser and Janet Kohlhase show, the biggest cities in America sprouted up on shipping routes—either near ports, railroad hubs, or both.81 Manufacturing firms moved near these hubs to reach the national market, and the makers of intermediate goods followed close behind. The precipitous decline in shipping costs over the twentieth century—brought about by inventions like the internal combustion engine and the shipping container—meant that manufacturing firms could spread out, leading to the decline of the modern manufacturing city.
The cities and metropolitan areas that have thrived in the last forty years have been those with advantages rooted in other agglomeration economies, particularly deep labor markets in high-end service industries.82 A primary agglomeration benefit of deeper labor markets is that they simultaneously allow greater specialization and reduce firm-specific risk.83 For example, actors in Los Angeles have all sorts of advantages over actors in Milwaukee. A Los Angeles actor can increase his wages by specializing in, say, the role of a zombie or mafia henchman. By contrast, a Milwaukee actor has to play whatever roles he can come by. A Los Angeles actor can safely invest in human capital knowing that there will be jobs he can access without moving. The Los Angeles actor has “insurance” that a firm-specific failure—the owner of a theater investing with Bernie Madoff, for example—will not leave him jobless and forced to move. On the other hand, a similar firm-specific failure at one of the relatively few theaters in Milwaukee could end an actor’s career in that city.
Beyond the benefits of specialization and risk reduction, deep markets also allow for the capture of information spillovers that increase wealth and drive growth.84 In Silicon Valley, for example, software developers and venture capitalists learn just by having coffee with friends. A tech savant in Jacksonville, Florida would have no such opportunity to learn from peers. Lobbyists in D.C. learn from one another over dinner in Capitol Hill, becoming better at their jobs with each bit of gossip or scrap of insight into legislative procedure.85 Wall Street types learn about how to structure deals over steaks and cocktails at The 21 Club. And so forth. As Robert Lucas noted, “New York City’s garment district, financial district, diamond district, advertising district and many more are as much intellectual centers as is Columbia or New York University.”86
Glaeser and Giacomo Ponzetto have shown that cities with these advantages have seen growth despite decreasing shipping costs.87 Wages have soared in San Francisco, Boston, and New York as high-skilled workers have flocked to take advantage of these agglomeration economies.88 In a world where the costs of shipping goods are low but the opportunity costs of transporting people remain high, cities built on service economies have thrived.89 The nature of the economy determines the type of city that succeeds, and this dynamic can change rapidly.
This brief review of the agglomeration-economics literature provides three important takeaways about interstate mobility. First, you can’t go home again. Once cities decline, they will not come back in the same form. Car companies and parts suppliers are not likely to return to Detroit in the same numbers. A city may return to economic health, but it will not be for the same reasons. As a result, it has been detrimental for Detroit to have an infrastructure, population, and government tailored to the existence of an automobile industry that is never going to return. Cities must have the capacity to shrink their physical forms, their populations, and their governments when economic maladies, or simply economic changes, hit.
Second, individuals must be mobile to capture the gains from agglomeration. Agglomeration models, self-consciously, are premised on simple concepts of locational choice. In other words, these models assume that there are costs each time goods are shipped, but that firms and people can cheaply and easily move their base of location.90 In theory, this easy movement creates a “spatial equilibrium,” a condition in which there are no rents to be had from simply living in a region.91 When there is local economic growth, people move in. This spreads the gains from that local growth among a larger and larger group of people, until there is no incentive for the marginal resident to move to the city. As a result of the population moving to capture agglomeration gains, there is no greater economic advantage in these models from living in one region or city than there is from living in another.
But in practice, people are not moving, and localized economic booms have not produced “boomtowns.” As discussed both above and below, New York, San Francisco, Silicon Valley, and a few other cities have seen massive wage growth over the last twenty years, but not massive inflows of population.92 Economists like Glaeser and Krugman have turned to land-use restrictions to explain this fact.93 Though demand in these regions has predictably increased, restrictive land-use laws limit building and thus inward migration. As I will discuss below, these land-use restrictions are just one element of the set of legal limitations on interstate mobility.94
Third, when people and firms fail to move to agglomeration economies, the overall economy loses both output and growth. Actors are more productive in Los Angeles because of its deep labor market.95 Tech firms and tech workers are more innovative and better compensated in Silicon Valley. Wage growth is higher in dense areas.96 Further, deep labor markets drive investment in human capital and specialized education, as workers can be confident that there will be job openings that demand their newly gained skills.97 These effects are growing. In the last five years, a huge majority of jobs with high wages emerged in just a few metropolitan areas.98 Wage differentials between cities for similar jobs have increased over the past forty years; wages for all types of jobs have increased much more quickly in areas with high levels of human capital.99 In short, if people cannot move to booming areas and take advantage of agglomeration benefits, the whole economy suffers.
Economists use estimates of these lost labor market gains—that is, the higher wages people would earn in more productive regions—to estimate the cost of land-use restrictions. In a blockbuster study, economists Chang-Tai Hsieh and Enrico Moretti found that land-use restriction lowered aggregate U.S. growth by more than 50% from 1964 to 2009. Overall output would have been 8.9% higher if just New York, San Francisco, and Silicon Valley lowered regulatory constraints.100
Though some studies find smaller effects,101 even Hsieh and Moretti understate the true effects of limits on entry in one important way. They focus only on the loss of higher wages that workers would receive in places like San Francisco or New York. Limits on entry may have an even more concerning effect: the loss of innovation. As noted, increased density and better fit between people and places lead to the generation of new ideas.102 More frequent interactions between people can lead to new ideas, and these ideas drive economic growth. For example, in patent applications, inventors cite other local inventors at far higher rates than they cite nonlocal inventors.103 Jane Jacobs has argued that interactions between people working for different kinds of firms lead to cross-fertilization and new ideas.104
While we don’t really know where new ideas come from, limiting locational choice likely stifles innovation. An Atlantic City resident might be capable of developing an incredible new iPhone app, but it would prove difficult for her to make it a reality without the creative ferment and potential business partners with complementary skills that she could find in Silicon Valley. As a result, policies that inhibit interstate mobility harm not only the creation of wealth, but also innovation and growth. And stunted growth rates compound over time, thus significantly harming welfare.
One final branch of social science research that has addressed the interplay between mobility and economic well-being is the large literature on “neighborhood effects” in poverty relief.105 Scholars in this area, however, have primarily considered whether and how moving from one neighborhood to another within the same region influences labor market, criminal, health, or educational outcomes. To date, the study of neighborhood effects has not focused on the impact of moving to a completely different city or state.106
For example, the famous “Moving to Opportunity” program, launched by the Department of Housing and Urban Development in 1994, created a lottery system in five cities that gave vouchers to some public housing residents that they could use to move into low-poverty neighborhoods.107 The resulting data spawned countless studies and continues to yield insights into the role neighborhoods play in determining labor and education outcomes.108 Similarly, a number of prominent studies resulted from the remedial program that followed the Supreme Court’s decision in Hills v. Gautreaux, which resulted in the relocation of thousands of Chicago families in public housing to other neighborhoods in and around Chicago.109
Policy recommendations in the wake of such studies naturally considered the relationship between anti-poverty spending and locational decisions inside a metropolitan area. Yet this literature, as impressive and important as it is, does not address the role of interregional moves in reducing poverty. In failing to do so, it misses a key ingredient in the functioning of a proper welfare system.
First, employment rates can differ far more substantially across states or regions than across a metropolitan area.110 If workers in a Rust Belt city where the manufacturing industry has collapsed were able to move not merely to a nearby suburb, but across the country to high-wage, low-unemployment regions like Boston or the Bay Area, they would see greater labor market opportunities.111 Such increases in opportunity would lessen the need for federal benefits, with increased labor-market income replacing after-tax redistribution.
Second, state and local governments fund a great deal of welfare policy. Federal programs like Medicaid and Temporary Assistance for Needy Families (TANF) require matching funds from states.112 Further, despite fear of incentivizing exit, localities also engage in a substantial amount of economic redistribution.113 Addressing homelessness, for instance, is almost always a local responsibility.114 Most notably, public education has a huge redistributive element—although not always downward—for which states and localities provide about ninety percent of the funding.115
The crucial role played by state and local governments has major implications for federal welfare policy. If states and localities are responsible for funding a large part of social safety net spending, it is critical that poor people move to places where there are wealthier people and businesses to support those welfare programs. Under ordinary conditions, this mobility proceeds naturally—rich and poor alike are drawn to places that provide economic opportunity.116 But policies that interfere with mobility will lead to an increasing mismatch between state and local fiscal capacity and the needs for services and redistribution. For the federal government, alleviating poverty will simply be more costly if poor people are largely concentrated in states and localities that have fewer resources to offer.117
Decreased mobility across states and regions also raises intergenerational concerns. Raj Chetty and others used the Moving to Opportunity data to analyze the impact of neighborhoods on educational outcomes for students.118 Their findings suggest that, regardless of whether mobility inside a metropolitan area affects adult labor market outcomes, geography determines which schools students can attend and is thus a key factor in educational outcomes.119 School quality differs widely not just across neighborhoods within a metropolitan region, but also across the country. Limits on access to regions with high-quality schools and neighborhoods thus can negatively influence the educational outcomes of future generations.120
Intergenerational socioeconomic mobility also varies substantially across different regions in the United States.121 For example, as Chetty and others have found, “the probability that a child reaches the top quintile of the national income distribution starting from a family in the bottom quintile is 4.4% in Charlotte but 12.9% in San Jose.”122 Areas with residential segregation and concentrated poverty score worst on this socioeconomic mobility test.123 Social safety net policies are expressly aimed at encouraging mobility up the socioeconomic ladder. Encouraging interstate mobility—that is, facilitating migration from low-opportunity places like Charlotte to more promising cities like San Jose—is thus central to that mission.
D. When Is Labor Mobility a Problem? Stasis as a Good, Although One Well Understood by State and Local Governments
While the previous Sections have explored the benefits of mobility, it is also important to acknowledge the benefits of staying in place. People form deep attachments to the places in which they choose to live and may not want to leave behind a particular job, school, or church. Nothing in this Article suggests we should force people to leave places, or even provide net incentives for them to move.124 Rather, the Article argues that we should not impede mobility, instead allowing people to choose communities and labor markets as freely as possible.
But there are serious arguments that impeding mobility creates benefits, too. Many of these claims are encapsulated in the well-worked-out debate over homeownership subsidies, which will be discussed in depth later on. Supporters of the home mortgage interest deduction or government-sponsored mortgage subsidies stress a number of different benefits of homeownership for buyers, including the utility of a home as a savings device.125 But the central plank of their argument is that homeowners move less and care more about their communities.126
Supporters of homeownership subsidies argue that residential stability and personal investment lead to communal benefits. Homeowners are more likely to invest in their communities and in their own homes than are fly-by-night renters, it is claimed, contributing to the creation of local amenities.127 Because it is costly to move and because the value of their biggest asset (their home) is tied to the quality of their neighborhood, homeowners will participate in parent-teacher association meetings,128 vote,129 and so on.130 Brian McCabe has shown that tenure in a community is more important to most types of local participation (except for voting) than homeownership on its own.131 Long tenure in a community allows residents to overcome collective action problems to create public goods, leading to both financial and social benefits for families and their communities.
Stable populations may also generate greater public or even business investments in a community. Building a subway or light rail line through a city or town only makes economic sense if the population that it will serve remains in the area (or moves to the investment and then stays in place).132 Businesses that see stable populations face reduced risk when investing in a community, and this reduction in risk should generally encourage investment.133
Beyond generating wealth through increased investment, geographic stability may support the development of beneficial social values and relationships.134 Stable communities can strengthen intergenerational bonds, with grandparents living near grandchildren.135 Long-term friendships may be more likely to endure in stable communities than in transient ones.136 In such communities, social capital can more easily develop.
While there are trenchant critiques of these arguments, this Article does not seek to litigate them.137 Whatever the benefits of population stability, they are already championed by local and state governments—to the point where they eclipse national interests. State and local governments constantly seek investments from businesses and want to ensure that their own public investments are not boondoggles. Homeowners are the biggest players in local politics already, and by the very logic of the argument in favor of stability, care very much about others becoming homeowners.138 And incumbent politicians likely prefer an immobile electorate because it means the voters who supported them remain in place.
This is not to say there are no national-level gains to neighborhood stability. Perhaps stable communities spur greater levels of overall business investment, due to reduced risk. Perhaps, as Naomi Schoenbaum argues, interstate mobility has doleful distributional consequences inside families, with trailing spouses bearing a greater deal of the costs associated with moving, since moving across states often entails a decline in familial support for child-rearing and reduced social options.139 Perhaps local civic participation encouraged by stable communities serves as a schoolhouse of democracy, making for better citizens at the national level.140 And there is some evidence that the act of moving has negative health and psychological effects for children, particularly if the move happens between ages twelve and fourteen.141
But these gains are merely speculative, and the national harm of limiting interstate mobility is grave. The large majority of gains from population stability are captured by the communities in which people stay put and not by the rest of the nation.
II. why is interstate mobility difficult? federal, state, and local laws and our sticky internal labor market
Direct limits by states on individuals or capital traveling across state lines are often unconstitutional under the “right to travel” doctrine or the dormant Commerce Clause.142 But many constitutional local and state policies nevertheless indirectly impede labor and capital mobility.143 These policies have substantial negative effects on the efficiency of the national economy.
However, for reasons explained in Section I.D, state and local governments have little incentive to care about these costs and strong incentives to promote stability. These governments by design represent local interests, often at the expense of the national interest. Furthermore, local governments are dominated by pro-stability, anti-mobility stakeholders—in particular, homeowners and older voters who turn out in low-information local elections. The federal government makes policies that impede mobility as well, often without much input from figures concerned with macroeconomic performance.
This Part is divided into three Sections. Section II.A discusses two primary tools through which state and local governments limit entry into labor markets: land-use laws that restrict housing construction and occupational licensing rules that directly bar participation in state labor markets. Section II.A explains how these laws directly affect the macroeconomic forces discussed in Part I. Land-use rules that limit access to rich regions reduce agglomeration gains and result in lower wages, as suggested by Hsieh and Moretti’s work.144 These laws likely also reduce growth rates, as they limit the capture of information spillovers. Further, land-use rules create rents for landowners and reduce labor income for potential entrants, contributing to wealth and income inequality. Similar stories can be told about occupational licensing, although because these limits are usually set at the state level, the effects are less direct. But the difficulty in transferring licenses across states makes moving to opportunity harder, harming the national economy.
Barriers to entry into rich markets also weaken the capacity of the Federal Reserve to generate monetary policy, or for Congress to use fiscal policy, in a way that produces full employment and steady prices. Price inflation in July/August 2016 in Silicon Valley was 3.1% while it was -0.3% in Cleveland-Akron.145 There is even more variation in employment between such regions. In 2010, unemployment rates among the fifty biggest cities ranged from 5% in Omaha to 25% in Detroit.146 Even in 2016, a year of low unemployment, rates of unemployment varied by more than ten percentage points just between counties in California.147 The existence of such heterogeneity in prices and unemployment poses a significant obstacle to the Federal Reserve and Congress.
Section II.B discusses the tools through which state and local governments impose barriers to exit, similarly reducing labor mobility. Conditions on public pensions, variation in the standards for receiving public benefits, homeownership subsidies, and general property rules all operate to make leaving harder. State and local governments often enact these policies for reasons external to mobility or stability. To the extent these policies have the collateral effect of limiting mobility, state and local policymakers are often favorable toward increased stability and rarely concerned with the macroeconomic ramifications.
When the federal government and state governments collectively structure Medicaid benefits so that the unemployed in Detroit will not move to Houston, we see greater variation in regional unemployment rates. These variations complicate the task of the Federal Reserve in setting interest rates and of Congress in deciding whether to run deficits.148 When teachers cannot move between states because their defined-benefit pensions have not vested, or when homeowners cannot easily sell their homes, we see fewer regions maximizing agglomeration economies. Together, these exit limits create substantial macroeconomic harms.
Section II.C homes in on a set of barriers to exit that particularly afflict declining cities: laws that make it harder to shrink governments and physical structures to match the current economic condition of cities. State and local governments in declining regions often seek to maintain their size, hoping for an economic revival. But when declining cities fail to reduce their housing stocks or governmental operations, the broader economy suffers. Housing policies that encourage homeownership and preservation of old buildings cause workers who might thrive in more functional parts of the economy to stay in declining areas. By contrast, policies favoring temporary housing and demolition could weaken people’s ties to those areas and encourage mobility. Similarly, when declining cities maintain too-large governments, benefit recipients and public workers stick around even when there are better opportunities in other places. Further, because taxes must increase to fund these services, what remains of the private sector in these declining cities suffers, encouraging exit by key contributors to what is left of the local economy. Taken together, such policies constitute a systematic bias against mobility—to the detriment of the national economy.
Land use and occupational licensing make it hard for workers to enter better job markets. These two regimes became substantially more restrictive during the same period in which interstate mobility fell.
Of limits on mobility, the best understood in the legal and economic literatures are land-use regulations.149 Before the 1970s, land-use restrictions (zoning laws, subdivision regulations, historic preservation, and so on) limited access to some towns or communities, usually rich suburbs.150 They did not, however, cap housing construction in entire metropolitan regions. Builders could always construct new housing, either in downtowns or on the urban fringe.
Something dramatic happened to land-use regulation in the 1970s and 1980s: it became much, much stricter.151 Importantly, while this phenomenon affected all types of municipalities—from urban downtowns to inner-ring suburbs to exurbs—it only occurred in particular regions of the country. In particular, coastal metropolitan regions like San Francisco, New York, and Boston restricted construction in cities, suburbs, and exurbs.152 Because these popular regions restricted new housing, demand for living space outpaced supply. Housing prices soared, but population growth did not.
In contrast to these coastal regions, Southern and Southwestern metropolitan areas like Houston, Phoenix, and Atlanta continued to impose minimal land-use restrictions.153 Though demand to live in these regions grew as well, this demand led to increased housing construction and population, rather than substantially higher housing prices.154
Because the most restrictive regions tend to be the nation’s richest, their lethargic population growth has reduced levels of wealth in the United States as a whole. As previously described, Hsieh and Moretti estimate that GDP would be 8.9% higher if land-use restrictions were reduced in three restrictive regions: Silicon Valley, San Francisco, and New York.155 And once again, because barriers to mobility reduce the capture of information spillovers, land-use restrictions may indirectly impede growth as well.156
Land-use restrictions also contribute to economic inequality. Because these restrictions raise the cost of housing, they disproportionately prevent poor and working-class people from taking advantage of high-wage job markets. Housing costs eat up a larger percentage of a poor person’s paycheck than that of a wealthy person. Thus, even in a city that can provide marginally higher wages, low-income persons simply may not be able to afford the cost of living in rich, land-use-restricted areas. While nominal incomes for janitors in New York are much higher than in poor states in the Deep South, real incomes, factoring in housing costs, are lower. As a result, restrictive land-use rules have meant that poor and middle-class people have little incentive to move to places where higher incomes are available.157 Therefore, these restrictions reduce labor income at the bottom of the income distribution. As Jason Furman, former Chair of the Council of Economic Advisers, notes, land-use restrictions “can increase inequality by reducing one of the channels through which workers get a raise, specifically moving from job to job.”158
This is no small effect. Thomas Piketty famously argued that increasing returns to capital relative to economic growth are a major driver of economic inequality.159 But Matthew Rognile and others have found that nearly all of the increased returns to capital in Piketty’s work came from housing capital: “[T]he long-term increase in capital’s net share of income in large developed countries has consisted entirely of housing.”160 This is a stark and important finding: Piketty’s result about capital is almost exclusively about real estate. Why? The most important reason is land-use restrictions. Today’srentiers have something in common with the classic rentiersof old—they are landowners, but instead of deriving income from owning arable soil that can produce income when farmed, they own access to valuable labor markets, which increases either their own income or produces rental income. Today’s rentiers then use zoning to ensure, and increase through monopolization, the value of their rents.161
These restrictions create income disparities not only between individuals, but also between states and regions. As discussed above, Ganong and Shoag have shown that from 1870 to 1970, states’ per-capita GDPs were converging. People would leave poor states like Mississippi to move to richer ones like Connecticut.162 But starting in the 1970s, the combined effect of land-use restrictions in many of the richest metropolitan areas stopped that convergence in its tracks. Though doctors, lawyers, and other high-skilled individuals who could afford the higher housing costs could still move to capture higher wages in richer, restrictive states, less-skilled workers could not.163 Notably, although the GDP gap between richer and poorer states is now static, convergence continues to this day within the set of states with less-restrictive land-use regulations.164
Limiting housing construction is not the only way in which states and localities prevent people from moving in. Labor regulations, particularly occupational licensing, also play an important role. While professions like law and medicine have long required licenses, over the past few decades states have dramatically increased the number of jobs that require state approval before practicing.165 Decades of work by Morris Kleiner and a recent outpouring of scholarly interest have documented a massive increase in state licensing requirements.166 Occupational licensing affects a greater percentage of the workforce than much more frequently discussed facets of labor law, like the minimum wage or private-sector unionization. According to a recent report released by the White House, “the percentage of the workforce covered by state licensing laws grew from less than five percent in the early 1950s to twenty-five percent by 2008, meaning that the State licensing rate grew roughly five-fold during this period.”167 These rules proliferated exactly as mobility declined.
About two-thirds of this increase is attributable to changes in state law, which added licensing requirements for a variety of professions, and the rest is due to increased participation in regulated industries.168 Occupations that require licenses in some states now include animal breeder, auctioneer, bartender, florist, interior designer, turtle farmer, hair braider, and scrap metal recycler, among many others.169
Proponents of licensing requirements argue that such laws increase the quality of goods and services, protect public safety, and increase consumer confidence, particularly when there are substantial information asymmetries between consumers and sellers.170 Others suggest that licenses increase professionalization in a given industry, incentivizing investments in training and boosting prestige.171 But research suggests that such positive effects of licensing requirements remain uncertain at best and are often minimal to nonexistent.172 Of course, there are anecdotal examples of such regulations serving a useful function. But there is not much evidence that, on average, occupational licensing requirements provide the benefits that many ascribe to them.
However, such requirements do work effectively as supply constraints, driving down employment in licensed industries and increasing wages and prices. Prices in such industries have been estimated to be three percent to sixteen percent higher than those in unlicensed industries.173 On average, employment growth in licensed industries is lower and wages are higher, although estimates of the size of the licensing effect vary substantially.174
Restricting competition and increasing wages and prices is in fact the goal of most licensing regimes. The politics of occupational licensing follow a classic Olsonian script.175 Workers in a particular field share a strong interest in establishing licensing requirements.176 The workers’ concentrated interests defeat the diffuse and disorganized interests of consumers and future entrants.177 What’s more, the administrative bodies that define the “scope of practice” in many licensing regimes are frequently staffed by representatives of the regulated industry, further entrenching the concentrated interest’s power.178
What matters most for the purposes of this Article is that licensing requirements limit interstate mobility. Licensing requirements differ across states. Michigan, for example, requires that licensed security guards undergo three years of training, while most other states require eleven days or less.179 This poses a significant barrier to entry for security guards who might otherwise consider moving to Michigan. In general, starting one’s old profession in a new state requires costly courses, tests, and training.180 Even when the tests are the same, states often require different scores to pass, making it difficult to transfer licenses.181
While multistate licenses and reciprocal agreements are available for a few industries (like nursing or lawyering), most occupational licensing regimes are state-specific and thus create substantial barriers to entry for many classes of workers.182 Indeed, states vary a great deal even as to which professions require licenses. More than 1,100 occupations are regulated in at least one state, but fewer than sixty jobs are regulated in all fifty states.183
Numerous studies support the idea that licensing laws limit interstate mobility. Kleiner and Janna Johnson found that licensed professions tend to have similar within-state mobility rates as compared to non-licensed professions, but far lower rates of interstate mobility.184 For example, they observed that “barbers and hairdressers are 27% less likely to move between states but only 7% less likely to move within state than their peers in other [non-licensed] occupations.”185 President Obama’s White House performed a similar analysis across a broader range of industries and found that “interstate migration rates for workers in the most licensed occupations are lower by an amount equal to nearly 15 percent of the average migration rate compared to those in the least licensed occupations. But the difference between these workers in within-state migration is much smaller, only about 3 percent of the average rate.”186 This affects labor mobility by reducing the degree to which workers in regulated industries can move to metropolitan areas in other states with different licensing regimes.187
There has been some positive movement toward reform in this area, and the push to remove licensing requirements enjoys bipartisan support.188 Following a proposal by President Obama, Congress appropriated a small amount of money to reduce technical difficulties for states that wanted to make occupational licenses work across borders.189 President Obama’s Fiscal Year 2016 Budget included $15 million in new discretionary funding for the Department of Labor to identify, explore, and address areas where licensing requirements create barriers to labor-market entry or mobility.190 The Obama White House even publicized a list of best practices for state governments, which has spurred some reform.191 President Trump’s Labor Secretary Alexander Acosta has continued this push, calling for the elimination of many licensing rules, and several Republican members of Congress have proposed legislation that would condition protection from antitrust review for licensing regimes on a reduction in their scope.192 Further, interest groups have become concerned with the issue. The Institute for Justice and the Pacific Legal Foundation have challenged the legality of various occupational licensing regimes with some success.193 The Supreme Court made a minor move against occupational licenses when it held that, when taken to the extreme, delegation of licensing to professional bodies could result in the forfeiture of state-action immunity under antitrust laws.194 But these small steps in the right direction have done little to change the underlying problem. A huge part of the labor force remains in industries subject to occupational licensing, with licensing requirements varying dramatically from state to state. These occupational licensing rules thus impede the internal market for labor.
While limits on entry make it harder for people to move to opportunity, limits on exit make it harder for people to flee depressed economic regions in the first place. These two factors work in tandem to keep people in stagnant labor markets, drastically limiting output and increasing inequality.
When a negative economic shock hits a city or region, driving unemployment up and wages down, people may nevertheless elect to stay for various reasons. Many of those reasons are nonmonetary. We make friends, build social networks, and raise our families where we live. So even when moving could lead to employment or higher wages, some people still may choose not to go.
Further, local government by its very nature limits individuals from moving to the best economic location.195 State and local governments provide a host of public services and legal regimes—from schools to police to gun laws—that are tied to geographic location. When people move from Detroit to Houston, they can keep their cell phone plans and Amazon Prime accounts. But they must take on a whole new set of public goods and legal regimes—changing schools, gun laws, and so forth. As I have argued elsewhere, this is a cost of having public goods provided locally, as opposed to federally or by private contract untethered from geography.196 Of course, local provision of services has many benefits, particularly (as Charles Tiebout famously argued) by improving fit between preferences and policies.197 But this fit comes at a cost: it undermines agglomerative efficiency because it forces people to choose their locations based on packages of government services, rather than solely based on an economic calculus.198 If parents like their child’s school, they will be less willing to take a new job in a different city—especially when there is no guarantee that a school of similar quality will be available at a similar tax rate.
It is not merely the fact that moving requires breaking social ties, or receiving different packages of local services, that limits mobility. There are many specific policies that inhibit exit among certain populations. This Section delineates a few of them.
Almost thirteen percent of Americans work for state and local governments, and ninety-two percent of them have defined-benefit public pensions.199 State and local government employees often must rely exclusively on these pensions, as many do not receive Social Security benefits.200
“Defined-benefit” plans guarantee benefits for workers who stay in a system for a set period of time. Such plans often have long vesting periods, requiring people to stay in a single system for many years. As of 2012, eighteen states require a teacher to stay in the system for eight or more years before she can access her benefits.201 In many states, a teacher who leaves before her defined-benefit plan vests can only withdraw her own contributions, not the contributions made by her employer; in some states, a teacher who leaves the system loses most or all of her own contributions as well.202 Thus, in contrast to “defined-contribution” plans, which an employee can take with her when she leaves her job, defined-benefit plans are often non-portable and thus make moving difficult. The disincentive to move is huge: Robert Costrell and Michael Podgursky estimate that “teachers who split a thirty-year career between two pension plans often lose over half their net pension wealth compared to teachers who complete a career in a single system.”203
Though researchers have found it difficult to calculate the precise magnitude of the effect on mobility, it is clear that the prevalence of defined-benefit pensions with long vesting periods hinders workers’ ability to move at will.204
Citizens who receive public benefits risk losing access to those benefits when they move. Many federal programs allow states to determine eligibility standards. This flexibility allows states to tailor their programs to their citizens’ needs and preferences, but it also leads to wildly fluctuating standards across states.205 When poorer citizens move from high-benefit to low-benefit states, they can lose access to health care and other poverty relief programs.206 This phenomenon has become even more dramatic since the passage of the Affordable Care Act, as the differences in eligibility standards have expanded.207 Under the block-granted TANF program, the standard for eligibility varies by huge amounts from state to state. A single mother with two children qualifies for benefits even if earning as much as $1,605 per month in Arkansas, but the same mother would need to earn less than $859 per month to qualify in New York.208 Many other nationalized programs have different standards based on state policies. For example, the Supplemental Nutrition Assistance Program allows states to set an asset limit for individuals who seek to qualify for benefits.209
Notably, differences in eligibility standards have increased just as the poor have become less mobile.210 Since the 1980s, states repeatedly have been granted greater operational control over the federal welfare state.211 In 1981, Congress terminated a huge number of federal programs and replaced them with block grants to states.212 In 1996, Congress replaced the Aid to Families with Dependent Children with the TANF program, which uses block grants to administer traditional welfare funding.213 Successive Presidents have issued states waivers to experiment with the use of Medicaid and other federal programs.214 The Supreme Court’s decision in NFIB v. Sebelius gave the states greater capacity to turn down the expansion of Medicaid in the Affordable Care Act,215 and nearly twenty have done so.216
There are good reasons to give states greater ability to determine benefits levels, but these differing standards hinder or bias interstate mobility in a number of ways. First, at the simplest level, people who receive benefits are less likely to move to states where they will no longer be eligible. This will, for instance, impede moves to Texas, due to its generally sparser benefits—and in spite of its hot job market.217 Second, in theory, more generous states may have less out-migration (and more incoming migrants), meaning there will be a greater mismatch between the “natural” needs of the labor market and the supply of labor.218 Finally, and perhaps most importantly, even where eligibility standards are similar, transfer can be difficult due to administrative requirements. These requirements are “dauntingly complex and often adversarial,” with a “degree of nonuniformity, ambiguity, and opaqueness [that] is astonishing.”219 As a result, even formally transferable benefits can create limits on exit.220 This limits interstate mobility, and thus limits the capacity of residents to move to out-of-state metropolitan areas.221
Homeownership, another barrier to exit, is encouraged and subsidized by a number of federal policies. These include: the mortgage interest deduction, preferred capital gains tax treatment for housing, mortgage insurance through the Federal Housing Administration and other agencies, the failure to tax imputed rent on owner-occupied housing, secondary market support for mortgages through Fannie Mae and Freddie Mac, direct federal spending, and so on.222 The magnitude of these subsidies cannot be overstated. The tax benefits alone constitute $121 billion in lost revenue annually.223
Some might argue that the supposed public benefits of homeownership justify tax subsidies this large. In particular, homeownership may create better citizens, encourage involvement in local affairs, and lead to contributions to other local public goods or social capital.224 But as mentioned earlier, most of these benefits are felt locally, meaning that state and local governments would likely subsidize them even if the federal government did not.225
And furthermore, these local benefits of homeownership come at the price of national economic harms. David Blanchflower and Andrew Oswald have shown that homeownership rates correlate with substantially higher unemployment: “[A] doubling of home-ownership in a state would be associated in the steady state with more than a doubling of the unemployment rate.”226 They also find that homeownership rates result in substantially lower labor mobility.227
The causal explanation for why homeownership rates increase unemployment is less clear. The most intuitive explanation is that homeownership directly limits exit. Because owners must sell to move, and selling is costly, one might think homeowners are less likely to move to new jobs. But this causal connection is questionable, as homeowners do not appear to have higher unemployment rates than renters.228
That said, during economic downturns, homeownership might have a particularly negative effect on exit. Fernando Ferriera, Joseph Gyourko, and Joseph Tracy find that owners with negative equity face “lock-in” effects.229 They cannot move because the revenue from the sale of the home does not cover the loan balance. Further, these homeowners might be loss-averse. As housing prices decline, people might be unwilling to take substantial losses on the value of their home—even if taking a loss would be rational. “Owners suffering from negative equity are one-third less mobile, and every added $1000 in real annual mortgage costs lowers mobility by about 12%.”230
Blanchflower and Oswald suggest another possibility: homeownership incentivizes people to enact restrictive land-use laws, so new firms cannot enter markets with high homeownership rates.231 Zoning restrictions may directly stop employers who cause local externalities, like factories or tall buildings. Many scholars, particularly William Fischel, have pointed to homeownership rates as a causal force behind greater land-use restriction.232 Homeowners care about preserving the value of their largest asset and use zoning to insure it against the risk associated with new development. A homeowner would vote to block construction of a waste disposal plant right next to her house, even if it promised to bring lots of jobs to town. As a result, homeownership may limit not just exit, but also entry—particularly entry by employers—thus increasing unemployment.
Yet another possibility is that high homeownership rates create externalities that apply to both owners and renters.233 Blanchflower and Oswald note that this could work through a channel of information about the local work force—low levels of mobility may tell employers something about the qualities of the local work force—although they note there has not been empirical work testing this hypothesis.234 But it is not hard to tell a story consistent with this idea. For instance, employers may make less effort to hire from regions with high homeownership rates, as they are worried their offers will not result in acceptances and moves (because they might believe that homeowners are less likely to move than renters). Homeownership would thus drive down the mobility of renters, because it would scare employers from even trying to recruit in places with high homeownership rates.
Whatever the cause, the empirics show that homeownership rates correlate with unemployment and lower interstate labor mobility. Because federal housing policy aims to increase homeownership rates, it reduces interstate churn and harms the efficiency of the labor market—both substantial costs that policymakers should take into account.
Like federal tax policy, state and local real property laws hinder labor mobility. Almost all property in the United States is held in “fee simple absolute,” the most extensive estate in property, providing ownership of endless duration, unencumbered by future interests, and with a full set of rights to use, alienate, or bequeath the property. Fee-simple ownership emerged as a norm when most of the value of property came from the use of property itself.235 The slow-to-change common law of property enshrines this default by limiting the development of new forms of ownership through the numerus clausus principle.236
But the fee simple is just one way, and perhaps no longer the best way, to structure property ownership. In an urban age, most of a property’s value comes from its proximity to other properties and from the ability to combine property uses in advantageous ways.237Lee Anne Fennell argues that the dominance of the fee-simple form of ownership makes it harder to assemble properties because of holdout problems.238 She also notes that the fee-simple form requires people to own property in a specific place.239
The fee simple encourages stasis, Fennell argues. When faced with a cross-country move, for example, a homeowner must decide whether to sell her house, and whether to buy a new one (making the move more permanent) or to rent (forgoing the savings and tax advantages of ownership). The realities of the home-loan market also make ownership less portable. If interest rates rise after someone has taken out a mortgage, homeowners who want to sell will not be able to keep their lower rate when they seek a new loan to buy.240
To address these concerns, Fennell proposes a different form of ownership: a “floating fee.”241 This new form of property ownership would give owners shares and use rights in a set of equivalent properties around the country.242 Owners of floating fees would be capable of building equity and receiving tax benefits, but they would remain mobile, rather than forced to put down roots.243 A “floating fee” would allow citizens to move to jobs without forgoing the benefits of ownership.
Whether the “floating fee” is a good idea is not important for the purposes of this Article. What is important is Fennell’s point that deep common-law rules, and their interaction with institutions like the mortgage market, do not facilitate a flowing national market for labor and capital. Further, states are unlikely to adopt the floating fee for the very reasons suggested in this Article: instead of encouraging homeowning, tax-paying citizens to remain within the state in times of economic downturn, a floating fee rule would allow them to exit the state just when the state needs them most. States have no reason to care about mobility and may in fact be inclined to obstruct it.
Common-law property rules are not the only, or even the most important, real-property institution that limits exit. Some states limit annual increases in property valuation for tax purposes, but require full revaluation upon sale. This practice became common after California passed Proposition 13 in 1978.244 Owners therefore have a disincentive to sell, since they cannot transfer the tax benefit to the next owner. As a result, people stay in their homes longer.245 Real estate transfer taxes, which vary widely across the United States, are direct limitations on mobility.246 A family that wants to sell its house and move out of Washington State, for instance, will lose a full two percent of the value of its home in estate transfer taxes.
Rent-control regulations work similarly. A tenant can usually keep her below-market rent for life, but cannot sell access to below-market rent to the next tenant, nor keep the protection in a new apartment. As a result, moving means losing access to the stream of value that is having an apartment at below-market rates. If a person moves, she thus loses the benefit she gets from the policy and there is no guarantee that her next apartment will have below-market rent.247
American property law is not unique in this regard, and it is not the worst offender. My colleague Bob Ellickson has shown clearly that American property regulation promotes mobility more than French law does.248 But American property regimes do restrict exit, both through tax policy and the basic structure of the common law.
As the previous two Sections have shown, laws limit the capacity of people to move to areas with economic opportunity, and to move away from areas that are economically stagnant. The laws discussed so far operate directly on people, constraining their individual mobility. This third Section concerns a different set of legal limits: laws that operate not on people, but on municipalities. These laws prevent cities in economic decline from shrinking appropriately. The failure of cities to shrink inhibits and distorts interstate population movement as well.
As the work of David Autor and others has shown, cities and regions suffering extreme economic dislocations do not necessarily see commensurate degrees of population exit.249 This is in part due to a specific set of exit limits that reduce and bias mobility from these types of cities. Buildings, roads, and other physical stock of cities tend to be inertial; they remain in place long after they cease to be of use. Nor do local governments usually reduce in size at the same rate as local economies. In fact, these governments often expand as economic opportunity fades.
When cities are unable to shrink—both physically and governmentally—to meet changed circumstances, capital and residents remain in declining places. Laws that make it hard for cities to shrink act as a form of selective exit restriction, creating perverse incentives for unemployed and poorer residents to stay in stagnant job markets. At the same time, as cities increase taxes to fund services for the growing number of poor residents, well-off and employed residents will move away. The result is a bias in mobility—encouraging the employed to leave while encouraging the unemployed to stay.
This phenomenon is best illustrated by a comparison between urban decline and corporate decline. One goal—according to some, the only goal—of corporate bankruptcy law is to reduce the cost of capital.250 When a firm becomes excessively indebted, a collective action problem among its creditors ensues; each creditor has incentives to act in ways that reduce the capacity of creditors as a whole to recover as much as possible.251 Bankruptcy law offers an orderly solution to that problem. Whether a bankruptcy involves writing down or delaying some debt while allowing the firm to continue operating, as in a Chapter 11 reorganization, or whether it involves selling off a company’s assets, as in a Chapter 7 liquidation, the bankruptcy process avoids piecemeal litigation and allows capital inside an insolvent company to be put to better use. In other words, bankruptcy takes capital stuck inside an excessively indebted company and unsticks it. And by offering a way to move capital ex post, bankruptcy reduces risks ex ante, encouraging investment.
By contrast, the capital trapped inside cities is not easily unstuck following a negative economic shock. Much of a city’s physical stock is by its very nature immobile and long-lasting—a reality often exacerbated by regulations. The suite of services municipal governments provide are enduring as well, often expanding in the face of negative economic news, sometimes well beyond the capacity of a city’s tax base. After a negative economic shock, welfare could potentially rise if a city’s building stock and government shrank to fit its new conditions. But such an adjustment is tremendously challenging to achieve. And, as will be discussed below, neither our current municipal bankruptcy regime nor current policies for knocking down buildings offer effective solutions for this adjustment problem.
The problem of urban bloat has plagued cities throughout history. Consider Guy Michaels and Ferdinand Rauch’s recent work on differential urban growth in France and Britain before 1700. Due to differences in the way troops retreated following the first sack of Rome, Roman garrisons in Britain did not survive as cities, but many French garrisons did.252 Hundreds of years later, French cities remained in these former garrison locations, arrayed along Roman road networks. The British population, meanwhile, began concentrating in coastal locations to take advantage of new shipping technologies. These coastal towns experienced faster economic growth, while French cities languished in their original, land-locked locations. By the Middle Ages, Britons were two and a half times more likely to have access to coastal waters than they were during Roman times, while the French were no more likely to have such access than they were more than a thousand years earlier.253 Residents of French cities stayed in place even when the relevant economic technology made obvious the benefits of coastal proximity.254
In part, of course, people did not migrate from inland French cities to the coast because of simple path dependence. But they also stayed because French cities developed governmental structures that were hard to relocate—for example, local bishops tended to be the providers of public goods.255 Because local public goods were only available in cities built on old garrisons, citizens could not move to locations more suitable for economic growth. Public policy limited the capacity of the populace to adapt to economic and technological change by failing to allow for cities to shrink.
In recent decades, the United States has felt acutely the problem of cities that are too large. Economic shocks like mechanization, declining transport costs, and greater trade competition have decimated a number of American cities built around manufacturing, leaving their building stock and governments too large for their current economic needs.256
On top of these broader economic developments, idiosyncratic shocks have hit particular U.S. cities. Atlantic City has been pushed into crisis because the advent of cheap flights made it easier for east coast vacationers to visit Floridian beaches, and the rise of casinos on Native reservations deprived the city of its monopoly over legal table games in the region.257 The result has been economic collapse—rusting husks of buildings with unclear uses, huge amounts of unused or barely used housing, and a heavily indebted local government attempting to provide services to a needy and poor population.258
Although the causes of these various municipal declines have varied, their outcomes have been similar: fiscal crisis and an inability to adapt. Increasingly poor populations are unable to pay enough taxes to maintain local governmental services. Why don’t places like Detroit or Atlantic City simply shrink as they might have grown, with people leaving, fields of wheat replacing houses, and government programs diminishing to fit a smaller tax base? The answer lies in the fact that, for municipalities, shrinking is much harder than growing. Public policy should seek to free up capital stuck in bloated cities.
Buildings, both residential and commercial, are usually built to outlast their first users. The private market for housing partially drives the construction of these enduring edifices. Homebuyers want to be able to sell in the future.
But it is also the case that local and state regulations encourage builders to construct houses that will not be easily destroyed. Building codes require housing to be built in ways that make it less likely to fall down.259 Zoning laws that limit heights and densities incentivize developers to build long-lasting houses, because new buyers know they will not be able to tear down and build higher in the future if there is demand. Local governments also use zoning and other rules to bar or limit mobile homes.260 Such regulations encourage capital to be spent on more durable housing, raising the costs of abandonment.
The widespread physical durability of homes means that when cities face negative economic shocks, they are generally left with far more housing than their populations require. Detroit’s population has fallen from 1.85 million at its peak in 1950 to slightly less than 700,000 today.261 Unsurprisingly, much of the contemporary housing stock dates back to the city’s golden era. The median house in Detroit was built in the 1940s, and thirty-three percent of the city’s housing stock was built before 1939.262 The housing stock of Detroit is old and massive relative to its current population.
An excessive supply of housing reduces the incentives for newly-unemployed workers to leave dying cities. Moving to another place would bring with it a higher income. Huge housing stocks in declining cities, however, mean that rents are lower than residents could find in places with more jobs, off-setting the income gains.263 Housing prices in Detroit are far lower than construction costs, while houses in ordinary markets are roughly equal to construction costs (and in restrictive markets, houses cost much more).264
This effect on mobility is even more dramatic if the places with more opportunities for employment also have artificially high rents due to local zoning restrictions. For example, to experience an increase in real wages when moving from Detroit to San Francisco, a worker would need to enjoy a huge increase in nominal wages because housing in San Francisco is very expensive due to local building restrictions. This stickiness should be particularly strong among lower-wage workers because, as discussed in Section II.A, housing takes up a larger proportion of their incomes.265
Further, for declining cities, a too-large housing stock produces all sorts of social ills. Empty houses can be criminogenic in urban areas and create nuisances that drive down property prices in nearby occupied houses.266 Neighborhoods consisting of many empty houses still draw on local resources—police have to guard them and street lights still need to be lit—but do not contribute much in the way of property tax revenue.267
In theory, shrinking a city’s housing stock after a negative economic shock should create general benefits—by reducing crime, eliminating the need for costly government services in depopulated areas, and increasing property values for remaining houses, to name just a few. But cities have found it very difficult to shrink their housing stock in practice. For instance, Detroit considered using eminent domain to take properties in certain zones, which would have allowed the city to allocate its scarce resources to more populated neighborhoods.268 But in the face of contentious debates over which neighborhoods should be closed and distaste for the use of eminent domain, the uproar that followed the proposal ultimately doomed it.269
Cities have come up with creative responses to the problem of excessive housing stocks, but these piecemeal, controversial solutions have been inadequate to combat a problem of such magnitude. For instance, many cities have developed “land banks” to acquire vacant and tax-foreclosed properties, clear their title and tax delinquencies, and then lease or sell them in ways that fit local purposes.270 This can, and frequently does, involve taking foreclosed residential properties and converting them into low-intensity uses like tree farms, or giving them away to neighbors to use as larger backyards.271 Many cities have started demolishing vacant houses—10,000 in Detroit, 3,000 in Buffalo, 2,500 in Cleveland.272 But the cost of knocking down vacant and underused housing is high, due both to physical constraints and federal regulations on displacement and environmental harm.273 Some state laws impose additional regulatory hurdles, resulting in huge disparities in how much it costs to knock down a house—from around $7,500 in Cleveland to $19,000 in Buffalo.274 And the act of government seizure and potential destruction of property is necessarily controversial, especially because of the political difficulties inherent in deciding which vacant houses and neighborhoods to demolish.
The basic problem therefore remains. Buildings stay up for long periods of time, for reasons both “natural” and policy-driven. As a result, cities find themselves with excessively large housing stocks after economic crashes, and the populations of these cities remain correspondingly too big. If housing was less durable, it would also presumably be less costly to produce ex ante. And if local governments permitted more mobile homes or just cheaper housing, the capital invested in housing could move to other sectors. Furthermore, if ex post policies existed to easily knock houses down, housing stocks could be reduced more easily. Defeating the problem of bloated housing stocks in declining cities will not be easy, but it is crucial for promoting a mobile labor force.
A related problem emerges with respect to city budgets after economic crises. When a city faces a negative economic shock, its tax revenue falls. However, the local government does not automatically shrink in response—contracts signed, debt issued, and programs established remain in place. Further, negative shocks result in increased demand for local services. Local redistributive services are needed more in times of economic dislocation. And when the private sector workforce diminishes, public aid recipients and public workers—all of whom have an interest in expanding government services—gain political influence.275
In theory, cities should save for downturns. However, as Brian Galle and Kirk Stark have demonstrated, municipalities wildly undersupport rainy day funds during boom periods.276 This is in part a problem of moral hazard: cities may expect federal or state aid should an emergency arise. Voters are also biased toward present consumption because if the debt gets too big, they can always leave to avoid paying the necessary taxes (unless the taxes are on real property that they own).277
Shrinking local government in response to a crashing local economy is a difficult task. As will be discussed below, though the federal law of municipal bankruptcy provides some limited tools for addressing debt burdens, it has significant limitations. State-level reforms are possible and frequently employed, but often fail to cure the long-run mismatch between local ability to pay and the cost of services in declining cities. As a result, declining cities end up with governments that are too large for them to fund on an ongoing basis.
But unlike excessively large housing stocks, the direct effect of this problem on mobility is mixed. When government services do not decline in proportion to a city’s diminishing fiscal capacity to provide them, it might artificially limit mobility, particularly among the poor. Under this line of reasoning, unemployed people will remain in dying American cities in part because these cities continue to provide public goods and services. But the tax increases needed to fund these services can in turn lead to an exodus of employed workers from the dying city.278 Such an exodus would technically increase labor mobility, but not the socially useful form of mobility this Article has praised. Driving employed workers out of dying cities does not balance out unemployment rates around the country, does not promote agglomeration economies, and makes funding redistributive services harder, not easier.279
In declining cities, the difficulties of reducing government services worsen over time. As private-sector workers leave, particular populations—net recipients of public services, public sector workers, pensioners—become more powerful in local politics, giving them power to prevent reductions in benefits. Consider Detroit, where the political power of pensioners increased during an economic decline. Despite a spiraling economy, active and retired public workers demanded and received “13th checks,” or non-contractually-owed, undisclosed bonuses.280 These checks cost the city nearly $2 billion from the 1980s through the 2000s.281 The city provided for pensioners in need, even though doing so was not prudent fiscal management.
State and federal responses to cities in crisis are detrimental to mobility as well, in large part because they confer benefits on geographic locations rather than on individuals. When cities face fiscal crises, states or the federal government may respond by providing substantial bailouts. Bailouts attempt to address the immediate fiscal challenges of cities and reinvigorate local economies. For instance, the federal government provided huge financial benefits to New Orleans after Hurricane Katrina. While providing support to the city of New Orleans was a standard form of government aid, it is important to understand that the federal government declined to exercise an alternative course of action: simply cutting each New Orleans resident a check.282
Bailouts have obvious short-term benefits and provide needed services to people living in poor places.283 Some people will never move and need services and jobs. But targeting places rather than people has substantial downsides as well.284 First, paying for these subsidies requires taxing people in places that are economically vibrant.285 Second, the benefits may accrue to landowners (who are less likely to be truly needy) rather than residents, as owners can increase rents following the announcement of a bailout policy. Finally, taking a longer-term perspective, unless place-based subsidies fundamentally alter the structure of local economies, they ultimately encourage people to stay in declining places.286 It is not clear why the country as a whole or a state in particular should want residents to remain in, say, Atlantic City rather than move to the New York City suburbs, which would give them access to a better labor market.287
From a cynical perspective, subsidies to declining regions sometimes can appear to be policies designed to serve the interests of rich areas—as efforts to keep the riffraff out—rather than genuine efforts to reduce poverty. Consider Michelle Wilde Anderson’s argument in favor of continuing subsidy programs for rural Oregon.288 For many years, local governments in rural Oregon benefited from a share of revenues from logging on federal land in their jurisdiction. Logging revenues, however, declined sharply in the 1980s and 1990s. The federal government agreed to keep sending funds to local governments in place of the defunct logging revenues.289 But there is pressure in Congress to take these communities “off the federal dole,” and the amounts paid have started to decline. If this funding dries up, local services will collapse. Even so, local voters have refused to approve tax increases to pay for needed services, keeping rates lower than those paid elsewhere in Oregon. Anderson argues that state taxpayers should subsidize the maintenance of minimum service levels in rural Oregon even if rural voters refuse to raise their own taxes to fund them as best they can, and even if few people want to move to these communities.290 Why? “[T]he restoration and continued population of our historic places, whether urban or rural, is a policy imperative for both environmental and humanitarian reasons.”291 She continues:
In my view, historic places and modes of living have existence value, even when they have trouble attracting residents and businesses in a competitive system . . . . [P]erhaps there is existence value to rural living, just as there is existence value to the forest ecosystems themselves—humankind made spiritually and morally more whole through the existence of households and environments beyond the hustle bustle of urban materialism. Perhaps we are made more whole not only by preserving ecological diversity but also by preserving knowledge—everything from animal husbandry to the DIY of home goods to survivalism.292
Anderson seems to acknowledge that aid to poor places is often proposed for purposes entirely unrelated to alleviating poverty.293 In fact, her argument that permanent subsidies will allow residents of rural Oregon to stay in place and remain adorable repositories of homespun knowledge is premised on the idea that they are unlikely to prosper economically. (Were they to prosper, it is hard to see why residents would need to make their own home goods!) In effect, she argues that Oregon taxpayers should incentivize families to remain in low-wage areas that are not supportive of human capital development (for adults or, more importantly, for children) because their maintenance of “historic modes of living” is something the rest of us value.
In many cases, the need for bailouts to economically declining places is both intense and compelling. Permanent subsidies, however, without any effort to encourage mobility away from dying places—or to credibly generate sustainable economic growth in the targeted area—are counterproductive. Though often motivated by noble instincts, they most likely reduce rather than increase opportunity among the poor, while harming the broader economy.
A city facing an economic crisis that fails to reduce the size of its government will frequently face a debt crisis—a shrinking tax base and growing expenditures are only sustainable for so long. Federal law provides tools for addressing excessive local indebtedness, most notably Chapter 9 municipal bankruptcy.294 But Chapter 9 municipal bankruptcy is an imperfect tool and could be improved to better promote interstate mobility.
When authorized by the state and after making good-faith efforts to negotiate with creditors, insolvent local governments can file for bankruptcy under Chapter 9. If a court finds that a municipality is eligible for Chapter 9 protection, it is granted a stay that bars efforts by creditors to recover debts and is allowed to develop a sustainable plan to adjust those debts. These efforts are backed by the court’s power to approve the writing down of some debt, without creditor approval if necessary.295 Further, Chapter 9 allows local governments to renege on collectively bargained contracts and encompasses all forms of indebtedness, including pension obligations that are protected specifically in state constitutions.296
Municipalities have not traditionally turned to Chapter 9 for relief. Between 1976 and 2009, there were only about forty filings by general-purpose municipalities.297 And most of those cases involved sudden financial shocks, like a major tort claim against a small municipality.298 Instead, most cities facing real crises have turned to state bailouts or other types of state intervention.299 Only in recent years have we seen bigger local governments with ongoing problems resort to Chapter 9, with the filings of Detroit, Alabama’s Jefferson County, and several California municipalities.300
The reluctance to use Chapter 9 to address systemic problems partially results from the law’s structure.301 Much of the content of Chapter 9 is borrowed from Chapter 11 corporate reorganizations.302 But, in order to avoid constitutional challenges,303 Chapter 9 places substantial limits on the power of bankruptcy courts to order policy changes. Section 903 of the Bankruptcy Code states that “this chapter does not limit or impair the power of a State to control, by legislation or otherwise, a municipality of or in such State in the exercise of the political or governmental powers of such municipality, including expenditures for such exercise.”304 Section 904 states:
“Notwithstanding any power of the court, unless the debtor consents or the plan so provides, the court may not, by any stay, order, or decree, in the case or otherwise, interfere with (1) any of the political or governmental powers of the debtor; (2) any of the property or revenues of the debtor; or (3) the debtor’s use or enjoyment of any income-producing property.”305
Chapter 9 is reasonably effective for restructuring and resurrecting municipalities in the types of cases for which it was traditionally used—cities facing huge debts due to certain one-time problems, such as a huge tort judgment or a theft by the city treasurer. That is, where Chapter 9 is used to achieve the municipal equivalent of a corporate reorganization, it works well enough according to its terms. But Sections 903 and 904 prevent courts from using Chapter 9 to alleviate a city’s larger structural problems, particularly where local politics make it difficult for government to shrink. The ordinary reading of the statute does not allow courts to order localities to raise revenues, sell property, reduce spending, or change their operations in any way, limiting courts’ capacity to radically reform local governments to bring them into line with changed local conditions.
Scholars and judges have gotten quite creative in attempting to fix Chapter 9. Randall Picker and Michael McConnell argue that Chapter 9 gives courts discretion to force cities to raise taxes and cut spending, notwithstanding the limitations of Sections 903 and 904.306 Specifically, they argue that the standard for insolvency for access to bankruptcy could be interpreted to force cities to make an effort to pay debts through spending cuts and tax increases. Courts could also find that city-proposed debt adjustment plans that do not institute reforms are not in the “best interests of the creditors” or are not “fair,” “reasonable,” and “equitable.”307 Clayton Gillette and David Skeel argue that judges should use these same tools to require local governments to change their governing structures, pushing reforms like at-large districts for city councils or reducing the number of city departments.308 However, as Gillette and Skeel admit, the constitutionality of such aggressive actions by courts under Chapter 9 is unclear.309 The use of judicial discretion to push cities to cut spending or reform their governing structure remains somewhat at odds with the spirit of Sections 903 and 904. But it may be permissible under the letter of the statute and normatively attractive, insofar as it would give courts tools to push cities to shrink and adapt to their current conditions.
Further, to make Chapter 9 effective in the context of shrinking cities, courts have already had to rethink the concept of insolvency. Under the statute, a city is only insolvent if it meets one of two criteria: either it is “generally not paying its debts as they become due” or it will be “unable to pay its debts as they become due” in the future.310 In a strict sense, most debt-ridden cities do not satisfy this definition—they could pay their bills today if they sold off all their assets, including land and buildings. Such a fire sale, however, would be disastrous for residents and harmful in the long run. In response to this dilemma, the courts in the Detroit and Stockton bankruptcies created a new concept, which they termed “service delivery insolvency.” The courts held that a situation in which services are so bad that they simply cannot be further cut to pay bills as they come due meets the statutory definition of insolvency.311 This implicitly lowers the standard for insolvency, as it suggests a limit on what cities need to cut before the court determines that they are unable to pay their debts.312
The legacy costs of both debt and government size distort migration patterns. Courts applying Chapter 9 should consider the effect of their decisions not only on a particular city but also on the national economy.313 Judicial innovations that make access to Chapter 9 easier for municipalities or push governmental reform would help alleviate the distorting effect local debt crises have on migration patterns. Judges should consider the effects of such decisions on the broader economy, and ask for advice from macroeconomic policymakers.
More broadly, the question facing a bankruptcy court in the case of a declining city is very different from one facing unexpected one-time debt. A place like Atlantic City does not need the equivalent of corporate reorganization. It needs liquidation, both of its physical assets and its governmental structure. Without a monopoly on gambling or substantial numbers of beachgoers, Atlantic City should be much, much smaller. The question is how to get it there.
Bankruptcy courts have started to notice the newfound nature of the problem they face with regard to declining municipalities. For instance, the Detroit plan included $1.7 billion for revitalization, including money for reducing the size of the housing stock through demolition.314 That is, the court allowed the city to pay back less debt in order to ensure that the city became a more sensible size going forward. Whether this is appropriate or necessary should be informed by its effect on macroeconomic conditions, not just city-level factors.
The primary goal of this Article has been to identify a problem of national import: sluggish interstate mobility, particularly away from poor metropolitan areas and toward rich ones, harms the broader economy. While it does not attempt to provide comprehensive answers, this Conclusion seeks to provide a framework for thinking through future solutions.
A too-easy answer is that states and localities should simply fix the problem by changing their policies. These governments created the problem, the argument goes, so they should fix it. But it is not so easy. As Adrian Vermeule and Eric Posner have noted, articles like this one that point to structural problems in public decisionmaking should not suggest solutions that ignore the very structural problem they diagnose.315
In this Article, I have argued that state and local governments harm national labor markets and macroeconomic policymaking largely due to countervailing state and local interests. To simply tell states and localities to solve the problems discussed would ignore the analysis above. That said, a reform-minded state or local politician, elected by chance, might establish procedural changes that would alter the policies that local politics produce. In other work, for example, I have discussed how such a politician might propose procedural changes in land use that would create a stable political equilibrium resulting in equally democratic but less restrictive policies than we see today.316 For instance, a city might require the local legislature to set a “zoning budget” every two years, or a goal for housing construction, that would bind itself from approving down-zonings until that goal is met (and require one-for-one up- and down-zoning tradeoffs after the target is hit). Doing so would solve collective action problems inside the legislature, much as fiscal budgeting is designed to limit sequential pork-laden appropriations bills. Further, the mayor—a citywide official concerned with citywide needs, not local neighborhood ones—could be given the power to propose a budget, and presumably would do so at the maximum level acceptable to the city council. Changing the order of voting on issues, and the decisionmakers who get to set the agenda, can influence the result.
Still, the central problem is that state and local policymakers do not have incentives to protect the broader national labor market. Only federal policymakers can be expected to act on behalf of such an interest. As a result, while state and local policies are primary contributors to the mobility problem, the focus of reform must be at the federal level.
Rather than lay out a full program to address the multitude of individual policy questions raised in this Article, it seems more useful to discuss how a reform-minded national politician should think about the problem of stagnant internal labor markets. The logic of the Article suggests that she should ask herself a few questions:
First, have the costs of local control become too high? There are plenty of very good reasons to allow states and localities to regulate land use, occupational licensing, property law, public benefits, and so forth. Even if local control makes the labor market more sclerotic, these might remain sensible institutional-design decisions for which the benefits outweigh the costs.
However, it may also be the case that state and local control no longer make sense, even if they once did. Changes in the economy and the political landscape might have diminished the benefits of local control such that the costs outweigh them. With respect to occupational licensing, for example, the nationalization of medical or legal markets might make state control less attractive. A national bar exam or national rules on the scope of practice for doctors and nurse practitioners may make sense. More generally, changes in the economy may have made it such that the costs of local control have become too severe. If so, the federal government should preempt state or local regulations and establish its own.
Second, is reform that encourages interstate mobility possible without displacing the useful aspects of local control? For instance, the Obama Administration proposed creating interstate compacts for harmonized occupational licensing rules across states.317 It also proposed best practices for limiting the negative effect of such rules. Perhaps the Commerce Department could issue a new model Standard State Zoning Enabling Act that includes limits on local capacity to exclude.318 The federal government could also broker deals between states that allow public workers to move without losing access to their pension benefits. Such federal policies could target national-level pathologies while still largely leaving regulatory authority in the hands of state and local governments.
More coercive efforts might also make sense. Aaron Edlin and Rebecca Haw Allensworth suggest that the federal government could go after occupational licensing by limiting the state-action immunity doctrine in antitrust.319 Similarly, Fair Housing Act enforcement and strong enforcement of the recent Affirmative Furthering Fair Housing regulations could limit certain exclusionary zoning practices when they disproportionately harm racial minorities.320
The federal government could also create tax incentives for states and localities that decrease entry limits. On the one hand, it could hand out carrots for good behavior. For instance, Josh Barro has suggested a Race-to-the-Top-style program for localities that increase housing construction.321 And if the carrot fails, policymakers could try a stick. Noting that homeowners tend to push for restrictive housing policies and also tend to reap the benefits, Edward Glaeser and Joseph Gyourko propose that the federal government suspend the mortgage-interest deduction for localities that fail to allow sufficient housing construction.322 The federal tax deduction for state and local taxes could be modified to not apply to any property-tax payer with property assessed at lower than some percentage of the property’s real value.
These proposals do not force the federal government to become directly involved in areas of local policymaking. Instead, they vindicate the federal interest by offering financial benefits or imposing costs—then leaving the states and localities to figure out how to comply.
Third, can the federal government bypass some of these problems, and the prodigious political difficulties facing reform in these areas, and address the problem of interstate mobility in a second-best manner? As one example, the federal government could simply provide incentives to relocate. During the 1960s and 1970s, the federal government did just this, although the incentives were narrowly targeted, and it proved difficult to measure the success of the program.323 Congress has also provided trade adjustment assistance to workers who move long distances after losing their jobs because of trade deals or import competition.324
These mobility incentive programs largely have been targeted at specific problems, such as layoffs following new international trade agreements. But it is not hard to imagine more general policies. For instance, Congress could increase the Earned Income Tax Credit for otherwise-qualifying taxpayers who relocate to a new state (or move a certain number of miles). Such a program could create an incentive to move that might overcome the drawbacks that federal, state, and local laws create. Alternatively, the federal government could create relocation-subsidy programs that specifically target economically depressed areas. The United States could follow Canada’s lead: The provincial government of Newfoundland and Labrador has begun a program that offers relocation subsidies to residents of depressed coastal villages.325
Another possibility is subsidizing information about mobility.326 College graduates are more mobile than people who didn’t go to college, and while much of this effect is due to increased economic opportunity, it is at least possible that some of this is due to the psychological and informational benefits going to college has on learning about other places. Other historical policies—like mass conscription in the Army—created similar opportunities for learning about other places.327 Perhaps subsidies for higher education could be understood as encouraging people to move. Other programs, like Teach for America, serve to provide small groups of people with information about working in other areas of the country.
Fourth, are reforms possible in the areas the federal government already regulates? Congress could order entities that are tasked with macroeconomic policymaking—the Federal Reserve and the Treasury Department—to conduct studies about the effect of policies on labor mobility. All policymakers, from regulators at the Centers for Medicare and Medicaid Services to enforcers of the Fair Housing Act could use this information to better determine the macroeconomic effects of their policies. Congress and courts could reform the Chapter 9 process, as discussed above.
Whether or not it adopts these particular policy proposals, it is imperative that the federal government take action to combat stagnation in our labor market. Many of the problems described here—zoning, occupational licensing, fiscal crises in declining cities—are getting worse over time. As these problems fester, labor mobility will continue to decline, monetary policy will be less effective, growth will be slower, and the tax burden of funding the safety net will be higher.
These economic costs come along with perhaps more severe political ones. The 2016 presidential election has been interpreted by some as the revenge of residents of rural and exurban areas left out of the economic boom we have seen in richer metropolitan areas.328 To the extent that this interpretation of the 2016 election is correct,329 a policy agenda aimed at increasing mobility is a tool for addressing the concerns of those rural and exurban voters. Further, such an agenda would capture the spirit of this Article’s argument that residents of rich urbanized areas are excluding residents of poorer exurban ones from opportunity. Politicians should consider pushing for reforms that will break down geographic barriers to opportunity. Doing so will not only make the country richer, but will further the political ideal of forging a unified economy and people from our many regions and groups.