It would be nice if we had a better idea just what causes economies to grow. Adding more people or capital or resources into the mix will do it, of course, but to really make people richer one has to raise GDP per person, and that’s the bit economists are less clear on. Generally speaking, such rises come from improvements in “technology”, but technology is a catch-all for an entire set of productivity-boosting things we don’t understand, nor do we have a good sense of how best to increase the rate at which new technology arrives.
But if we aren’t exactly clear on how to get from point A to point B, we aren’t completely lost either. We can sift through modern economic history and observe where fast growth has tended to occur — or still better, where it is occurring now. In the absence of a clear and reliable recipe for creating growth ex nihilo, the most sensible approach to getting more may simply be to do more of what’s already working, even if we aren’t precisely clear on why it’s working. The best strategy, in other words, is reallocation.
Productive firms should get bigger, and absorb market share, capital, and workers from competitors that aren’t up to snuff. But for this reallocation process to work effectively, workers must be able to move from slow-growing places to fast ones. Perhaps the biggest constraint on growth over the last generation has been the inability of the rich world’s most productive cities to accommodate all the workers who would like to live in them.
Urbanization has always been a critical part of modern economic growth. Early industrialization occurred alongside a mass migration from the countryside and small villages to booming industrial cities. London was already one of the world’s largest cities in 1750, when its population reached about 750,000. Yet over the next two centuries the number of people living in London exploded, to over 8m. Cities around the industrial world — Manchester and Liverpool, New York and Chicago — experienced similar booms.
Big manufacturers needed to be as near as possible to suppliers and customers, because of the exorbitant cost to moving things over land. In practice, that meant locating near coal fields and port cities (into which raw materials could be shipped in and processed goods shipped out). Labor poured into cities in search of work, and the pool of available labor became an attractive force in itself. Urban growth was self-reinforcing. Concentrations of industry encouraged innovation in manufacturing, as firms tinkered with ways to improve productivity. Booming cities seemed to provide a double boost to growth: as people moved from less productive towns and farms to more productive manufacturing cities, and as economies of scale within the cities themselves raised productivity. The industrial revolution would have looked very different had urban growth been held in check, delivering less growth in output and incomes and less societal transformation.
That’s an important lesson for economies today. Technological change is again transforming economic activity. From the early 1980s advances in computing and information technology squeezed employment in middle-skill occupations, via automation as well as the expansion in global trade that new communications technologies allowed. In the many cities that had built their economies on mass employment of modestly skilled workers in industry or routine business services, this development was a killer blow.
But these technological advances were very good for different workers in different cities. Economists Edward Glaeser and Giacomo Ponzetto argue that the “death of distance” was very hard on industrial cities like Detroit, but a reinvigorating force in “idea-producing” cities like New York.1 By expanding the markets in which good ideas can be deployed, information technology magnified the return to knowledge and to the skilled workers who contribute to it. A globalized financial market means that successful trading strategies generate enormous returns, and the concentrations of financial activity that generate those strategies have benefited correspondingly. Workers in technology clusters like the San Francisco Bay area have done very well as a result of their ability to provide their products to users around the world: from e-commerce to online gaming.
Successful cities have also shown the potential to drive employment growth. Economic historians Thor Berger and Carl Benedikt Frey tracked the creation of new job categories in American cities over the past few decades.2 They discovered that newly created work changed in nature between the 1970s and 1980s, from fairly routine tasks to brainier work that relied on more skilled or creative labor. They connect this shift with a geographical change: from the 1980s new employment categories overwhelmingly pop up in highly educated cities like Boston and San Francisco. It isn’t just the brainy that find work in such places, however. Enrico Moretti, an economist at the University of California, Berkeley, has calculated the “local employment multiplier” for various sorts of jobs. A new manufacturing job tends to correspond to the creation of 1 to 2 additional jobs in the local economy. For more skilled work this multiplier is higher, rising to 4 to 5 jobs for each new tech-industry position.3
Past experience and theory suggest that cities whose economic purpose has been destroyed by technology ought to be losing people, the young and ambitious especially, to cities to which technology has provided new life. The population ought to be reallocating itself, in other words, toward places with high levels of productivity and high wages, the better to boost economic growth and lift household incomes.
But that is not at all what has happened. In 2000, there were slightly more people employed around the San Francisco Bay than in Dallas or Houston. Since that time, personal incomes in the Bay Area have grown about as much as those in the big Texas cities, even though Bay Area incomes were far higher at the beginning of the period. The real output of the Bay Area has also risen by about as much as that of Dallas and Houston.
The employment picture, however, could not be more different. Since 2000, employment in the Bay Area has grown by roughly 20,000 jobs. In Dallas, by contrast, employment has risen by more than 400,000 jobs and in Houston by more than 600,000. This remarkable divergence is part of a broader phenomenon in which Americans migrate away from the most productive cities. Between 2000 and 2009 the metropolitan areas of Boston, New York, San Francisco, San Jose, and Washington lost nearly 3m people, on net, to other American cities. The ten biggest recipients of net internal migration, on the other hand, absorbed over 3m Americans during that time, despite the fact that the average wage in the gaining cities was 25 percent below that in the losing cities.4
People are moving toward lower productivity places, and the reason for the phenomenon is clear: housing. The median value of a home in Harris County, Texas, the center of the Houston metro area, is just under $130,000. In Santa Clara County, in the middle of Silicon Valley, the figure is nearly $660,000. Wages in highly productive places are high, but often not high enough to offset the exorbitant cost of housing. And so people migrate, because the pay cut that results nonetheless entails a net rise in disposable income.5
Housing is more costly in the most expensive cities because so little of it is built. In the 2000s, Houston’s housing stock grew by more than 25 percent while that in the Bay Area grew just over 5 percent. In 2013 Houston approved 51,000 new homes while San Jose okayed fewer than 8,000, despite the booming Silicon Valley economy. Glaeser and Kristina Tobio find that since the 1980s, the extraordinarily rapid growth in the population of Sunbelt cities is due primarily to the receptiveness of those cities to new construction.6 A strengthening economy in places like Texas and Georgia leads to a construction boom and rapid population growth, while economic booms in coastal cities lead to very little population growth but soaring housing costs.
The net effect on the economy is difficult to estimate precisely but probably huge. Chang-Tai Hsieh and Moretti reckon that America’s inability to move people to the best opportunities means that over the last half century its GDP has fallen 13 percent below the level it might otherwise be: a tremendous loss in a $16 trillion dollar economy.7 Were builders able to respond to demand, employment in the New York metropolitan area might be eight times higher than it currently is while employment in the Bay Area would be two to three times higher.
Those figures provide a big clue to what is constraining housing supply in productive places. Were the New York metro area to accommodate all the people who probably should be working there, its population would grow about as rapidly as it did in the 19th century. But that earlier population explosion occurred on a much smaller base and led to overcrowded slums, disease, and other urban ills. Avoiding those problems while growing at the necessary clip would force expanding places to spend massively on new infrastructure and to tolerate the almost complete reconstruction of their cities. The residents of places like Boston, New York, and the Bay Area don’t want the trouble, don’t want the crowding, and don’t want the expense. So they have organized methods to block new construction: through strict zoning codes, historical and environmental protection laws, and through overt political pressure.
If one had a magic wand to wave and wanted to boost growth, magically neutralizing opposition to new development in the most productive cities would be one’s best bet. In the absence of a magic wand, solving the problem probably requires a two-pronged approach. On the one hand, it must be made easier for big cities to invest in big infrastructure projects, like the ones that allowed them to get so large in the first place. That means simplifying the regulations that constrain such investments and raise their costs. It means designing project bidding in ways that encourage competition and create the incentives for efficient, on-time construction. It means reforming the federal government rules that channel infrastructure money toward places that don’t need it, and, yes, it means using the federal government’s ability to borrow at remarkably low interest rates to make an economically justified investment in America’s future.
But infrastructure alone will not solve the problem. Instead, metropolitan areas may need institutional reforms that better balance the economic interests of the metropolitan area (and the country as a whole) with the interests and preferences of those living in neighborhoods that are likely to be affected by new development. When land-use decisions are made at a hyper-local level — giving local councilmembers or commissions extensive influence over which projects are approved, or focusing negotiation between residents and developers at the street level rather than the metropolitan level — the result will typically be far too little development. Those living immediately around a project enjoy some of its benefits but bear nearly all of its costs, in terms of disruption and congestion; they are therefore highly motivated to block projects and can succeed when local institutions enable them.
Political reforms could improve the situation. Roderick Hills and David Schleicher propose that metropolitan areas change the way they use city plans to make them “stickier”, so that they function more like development budgets that lay out how much new construction the city could use and where it should occur.8 To keep to the overall budget, they reckon, more political bargaining over how to share the costs and benefits of growth would occur at the city level, rather than within individual neighborhoods. And because more of the benefits of new development are captured within the city as a whole, this negotiation framework ought to mean that more of the allowable construction in given city plan is realized than occurs under the typical planning arrangement.
Fixing planning regimes won’t be easy. But the return to getting it right is likely to be handsome: more people working in more productive cities, generating more economic growth.
4 BEA, BLS; Ryan Avent, The Gated City (2011).
6 http://scholar.harvard.edu/files/glaeser/files/the_rise_of_the_sunbelt.pdf7 See discussion of the paper at http://www.vox.com/2014/7/15/5901041/nimbys-are-costing-the-us-economy-billions8 http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2477125
The opinions expressed here are solely those of the author and do not necessarily reflect the views of the Cato Institute. This essay was prepared as part of a special Cato online forum on reviving economic growth.