Global Urbanization

Chapter 4. Urban Growth and Development at Six Scales: An Economist’s View

C h a p t e r 4 Urban Growth and Development at Six Scales: An Economist’s View Stephen Malpezzi The world’s population, roughly 6.7 billion people, spreads over about 13 billion hectares of land. (Of course, much of this land is arid or otherwise inhospitable to settlement, but more on that later.) Variously considered, people live in nations, regions, cities, and neighborhoods. The study of their urbanization is, more or less, the study of density as it occurs in these venues. This chapter classifies urbanization along six scales, ranging from the largest (conceptual) to the smallest (neighborhood ) and outlines the specific issues of each. It employs urban economics to explore these scales and to suggest questions for prospective urban research. Six Scales of Urbanization The first scale, conceptual, crosses geographies and applies to many, if not all, the units of observation. The second scale, global, is selfexplanatory and considers several worldwide patterns over time. The third scale, national, focuses on key patterns and drivers of urbanization in a country. The fourth scale, regional, looks at systems of cities. It is analytically problematic because different definitions of ‘‘region’’ make data collection difficult. Some definitions of region (e.g., regional labor markets) are actually closer to the fifth scale, city. For instance, ‘‘Chicago ’’ can refer to both a metropolitan region and an area defined by political boundaries. The sixth scale, neighborhood, describes patterns of density and location within cities. Urban Growth and Development at Six Scales 49 Scale One: Conceptual: Why and How Cities Exist While a general definition of a ‘‘city’’ holds that it is a place where density (of people, jobs, firms, capital) is above average for a given country, to an economist a city is a place with higher value output than it would have if it were not dense. To understand this, consider two thought experiments . First, imagine that the roughly $24 trillion in capital stock of the U.S. and the country’s 150 million workers randomly moved so that each of the nation’s 10,000 zip codes had an equal share. GDP would drop substantially due to the loss of the benefits of economies of scale and agglomeration (discussed in some detail below); exactly how much is hard to pin down, but perhaps output would fall to a tenth of its current value. The fact that U.S. GDP is surely trillions of dollars more than it would be if there were no cities—and the same is broadly true for world GDP—is a measure of cities’ value. Next, imagine that the entire capital stock of the U.S., or the world, moved to New York. GDP would again be reduced to a fraction of its previous value due to its concentration in one place. This is why there are systems of cities, or regions. From an economic standpoint, cities exist because they offer opportunities for comparative advantage (increased trade), economies of scale (internal firm savings) and economies of agglomeration (external locational advantages). Several elements underpin these concepts. They include transaction costs (mixed but, on balance, lower), creativity and entrepreneurship (existence of and openness to innovation), variety and specialization (improved labor pooling; sharing of inputs; matching among labor, jobs and capital), and information (production, diffusion, and spillover effects). Agglomeration economies, resulting from both public and private activities, are of two broad types: localization economies (external to the firm but internal to the industry) and urbanization economies (external to the firm and external to the industry). In the former, the kinds of firms (or other entities) and their locations matter; in the latter, the size of the urban area is important (Kahnert 1987; Segal 1976). Plentiful research debates the relative importance of each type of agglomeration; both forms seem to matter and the existence of one does not preclude the other (Carlino 1987; Henderson 1988; Rosenthal and Strange 2001). Early examples of industrial agglomeration are legion and include Lancashire’s textile industry (1770s), Glasgow’s development of shipbuilding (1820s), Berlin’s leadership in electrical engineering (1870s), and Detroit’s innovation in automobile design and manufacturing (1900s) (Hall 1998). More recent examples of agglomeration identify such well-known high-tech locations as Silicon Valley, ‘‘Silicon Fen’’ (Cambridge), ‘‘Silicon Glen’’ (Glasgow), ‘‘Silicon Hills’’ (Austin), and 50 Twenty-First-Century Population Prospect ‘‘Silicon Plateau’’ (Bangalore) (Saxenian 1996). Although Rosenthal and Strange’s 2004 study of the software industry around Silicon Valley is...