Killer Cities: Past and Present

摘要

In the nineteenth century, industrial cities were incredibly unhealthy places to live due to a combination of infectious diseases and pollution (Cain and Hong 2009; Kesztenbaum and Rosenthal 2011). Today, industrial cities in developing countries face similar challenges, particularly the threat of high pollution levels. For example, a 2012 World Health Organization (WHO) report attributed 3.7 million premature deaths to ambient air pollution (WHO 2014). Of these, about 88 percent occur in low-and middle-income countries, chiefly in East and South Asia. In China, the air of booming mega-cities is sometimes so thick with soot that news reports have dubbed it an "airpocalypse" and the contribution of air pollution is estimated to have led to 1.2 million excess deaths in 2010. These observations raise a series of questions about the progress of health in cities. How much progress has been made in improving the health of cities over the past century and a half? Are the industrial cities of today's emerging economies less healthy than more rural areas, as was true in the past? What role does pollution play in the health of cities, and how has this changed over time? This paper takes a first step toward answering these questions. To do so, we study urban mortality patterns in nineteenth century England and in China in 2000. In both time periods, these countries represent leading industrial producers, and in both settings we find reports of high mortality levels and polluted urban environments. Our approach to this issue involves constructing measures of urban pollution, mortality, city size, and city density, as well as control variables, that are as similar across the two settings as possible. We then estimate the relationship between mortality and city size, city density, and city pollution in both the historical and modern setting, using simple cross-sectional regressions. Comparing how these relationships have shifted over time can give us an idea of how much progress has been made, and how much factors such as pollution continue to affect urban health. One of the main challenges in this study is a shortage of direct pollution data, which is often scarce in both developing economies and historical settings. We bypass the need for direct measures of pollution by using the composition of industries in a location, together with information about the pollution intensity of different industries, to generate proxy measures for the level of pollution in a location. One advantage of this approach is that it can be applied in settings where direct pollution measures are often unavailable. A second advantage is that our measure will reflect a multidimensional version of pollution, rather than a single pollution measure, such as total suspended particulates (TSP), which will capture only one element of urban pollution. However, this also means that we cannot differentiate the impact of specific types of pollution. This approach is similar to that used by Glaeser and Kahn (2008) for the United States and Zheng et al. (2011) for China, except that we focus on industrial sources of pollution rather than pollution produced by households. One motivation for our focus on industrial pollution is that industry is a major pollution producer. For example, Zheng and Kahn (2013) report that industry consumes 89.1 percent of total energy in China. A second reason to focus on industry is that it tends to be geographically concentrated, leading some areas to have much higher pollution levels than others.