In lieu of an abstract, here is a brief excerpt of the content:

C H A P T E R 2 The Growth of Metropolitan Areas in the United States Gilles Duranton Between 2000 and 2010, the population of metropolitan areas grew on average by 10.8 percent in the United States. This average masks considerable heterogeneity. Of 366 metropolitan areas, 42 lost population while 65 grew by 20 percent or more during this decade. City decline leads to calls for a policy response. In the United States and other developed countries, a wide variety of revitalization policies have been proposed, from the construction of major infrastructure to high-tech clusters to various types of urban beautification policies to attempts to attract “talent.” In the absence of a good understanding of what drives city growth, many of these initiatives are likely to prove futile. This chapter provides a brief summary of what is known regarding the drivers of the growth and decline of cities. Given its breadth, this chapter will focus mostly on U.S. cities. The exposition attempts to remain accessible to the nontechnical reader; a more technical exposition of the issues surrounding the growth of cities can be found in Duranton and Puga (2013). The chapter explores six mains “engines” of metropolitan growth uncovered in the literature: roads, amenities, human capital, entrepreneurship, industry clusters, and land regulations. Given space constraints, the exposition can focus on only a few key papers for each engine of growth. Other potential engines of city growth—such as non-road transportation infrastructure like airports or the quality of local governments—may also mat- The Growth of Metropolitan Areas in the United States 27 ter, but the extant literature does not provide significant empirical support for these alternative drivers of city growth. This does not imply that the six engines of city growth explored here are exclusive: We simply know extremely little outside these six engines of city growth. Roads and Land Use To understand why we expect roads—and particularly roads within cities— to matter, returning to the core model of cities as developed by Alonso (1964), Mills (1967), and Muth (1969) is useful (for a historical perspective on transportation as a driver of city growth, see Edward Glaeser’s chapter in this volume). This model represents cities as a central business district surrounded by residential areas. Potential residents, before moving to a city, consider the wage the city offers and the cost to live there. This cost of living is determined by the cost of housing and of commuting to the central business district where jobs are located. While arguably primitive in its representation of the geography of cities, this model highlights the importance of “accessibility” and the endogenous determination of housing costs. More specifically, higher housing costs close to the center offset shorter commutes, while higher commuting costs offset cheaper housing on the outskirts of a city. Keeping population constant, a transportation improvement should, thus, lead to both lower commuting costs (a direct effect) and lower housing costs (an indirect effect, given that accessibility is now less binding). Both effects serve to make the city more attractive to potential residents since it now offers higher wages net of urban costs. This simple argument predicts that transportation improvements within cities should cause city population and city employment to grow. However, we expect city population and employment to increase only gradually after an increase in roadways since new construction and other adjustments to housing stock do not happen overnight. Together, this suggests looking at city population (or employment) growth over time as a function of initial roads and population. Roads, and more specifically the development of the U.S. interstate system, are arguably the quintessential transportation improvement of the second part of the twentieth century. Duranton and Turner (2012) gather data about the mileage of interstate highways in U.S. metropolitan areas in 1983 (the earliest 18.118.150.80] Project MUSE (2024-04-25 05:06 GMT) 28 Gilles Duranton available year of data) and employment in these metropolitan areas in 1983 and 2003. They first perform a simple regression using employment growth between 1983 and 2003 as a dependent variable and miles of interstate highways in 1983 and employment at the same date as explanatory variables. They measure an (apparent) elasticity of population with respect to interstate highways of about 7 percent. That is, doubling the roadway in a city is associated with an increase in employment of about 7 percent over the subsequent twenty years. Regressing employment growth in cities on a measure...

Share