Urban Sprawl in Europe

Changes in urban forms and development patterns are crucial to understanding the role of cities as engines of growth. Urban sprawl is usually defined as the spreading of a city and its suburbs over rural land at the fringe of an urban area. Urban planners emphasize the qualitative aspects of sprawl such as the lack of transportation options and pedestrian-friendly neighborhoods. Conservationists tend to focus on the actual amount of land that has been urbanized by sprawl.

Although urban sprawl has been extensively studied in the United States (see, for example, Brueckner 2000, 2001; Glaeser and Kahn 2001, 2004; Nechyba and Walsh 2004), very few empirical studies have been undertaken in Europe. Basic reasons for this lack of research are the conceptual divergences in the U.S. and European geographical definitions of cities and the limitations in the availability of actual data for Europe.

Urban sprawl is one of the most important types of land use changes currently affecting Europe. It increasingly creates major impacts on the environment (via surface sealing, emissions by transport, and ecosystem fragmentation), on the social structure of an area (by segregation, lifestyle changes, and neglect of urban centers), and on the economy (via distributed production, land prices, and issues of scale). It is therefore crucial to understand it better.

During the second half of the twentieth century, urban sprawl has become a mass phenomenon throughout the Western world. Although suburbanization took also place in Europe during the post war period, its dimensions were by far less expansive than in the United States. In the 1950s, numerous European countries were concerned about reshaping their cities. Besides, a lot of countries had been decimated by the war, and many large cities such as Berlin, Vienna, Glasgow, and Birmingham were stagnating, or even lost population (Bruegmann 2005). The postwar period in the United States, on the contrary, was characterized by economic prosperity and a vast population growth (Burchfield and others 2006). Within less than twenty years, the U.S. population increased by 50 million people from 150 million in 1950 to 200 million in 1968 (Couch, Leontidou, and Arnstberg 2007). Some cities were even growing to a faster degree. In the same period, the Los Angeles area more than doubled from under 4 million to more than 8 million people. The Phoenix urbanized area grew almost fourfold, the San Jose area more than fivefold. In Europe, there was generally less growth in urban areas and therefore less pressure to develop the countryside. Besides, urban expansion was usually highly regulated. Planners and other government officials were able to intervene in city development more actively than their U.S. counterparts. In Paris, for example, large parts of suburban settlements consisted of high-density houses that were directly built by governmental bodies or were at least highly subsidized. This procedure was not common in the United States, where the private-market, single-family home was the norm.

As long as the American metaphor has not been replaced by a European one, it will shape the perception of many Europeans. In fact, remarkably little research has been done on the development of a "European" model (at least from economists). As stated above, this is mainly due to the scarce availability of data on indicators of urban performance such as urban amenities, housing, job opportunities, skills, and economic structure that limit the research possibilities in the European Union.

In this paper, we use a recently available dataset, the Urban Audit, which claims that it contains information on more than 300 variables for 258 towns and cities in the European Union's fifteen member states and its twelve eastern European candidate countries, measured at three different points in time: in 1991 (or in the period 1989–93), in 1996 (or in 1994–98), and in 2001 (or in 1999–2003).¹ Unfortunately, the effective information delivered is much more limited. Its coverage is rather poor. For several countries, many indicators are not provided, and missing values do not generally occur for the same cities across variables (even though data coverage does improve over the course of the decade). Nevertheless, collection of...