Part I
Spatial transformations and new geographies of cities and regions
1 Urban transformations and the future of cities
Edward L. Glaeser
Introduction
For over 100,000 years, our species existed as low-density hunter-gathers. For the past 10,000 years, humans lived as rural farmers and as late as 1900 only 15% of humanity lived in cities. In the shockingly short time of a century, the human experience has transformed itself dramatically, so that we are now an urban species. In the West and in East Asia, urbanization has been associated with rapidly rising incomes, but a surprising number of mega-cities have emerged in countries like the Democratic Republic of the Congo that are still extremely poor.
In the second section of this chapter, I discuss the three trends that have come together to make humankind an urban species. In the wealthy West, many cities have enjoyed a remarkable resurgence as former industrial cities have retrofitted themselves as centers of knowledge creation and business services. In some cases, particularly in Europe, these wealthy cities have also succeeded as centers of consumption as well as production. In East Asia, urbanization has been part of a collective process of rapid industrialization and development that echoes the earlier Western industrial revolution, but on steroids.
Sub-Saharan Africa today and Latin America during the 1960s experienced urbanization without growth (Fay and Opal, 2000; Gollin, Jedwab, & Vollrath, 2016). In the second section, I discuss two competing explanations for this phenomenon. The centralized power hypothesis argues that urbanization in these poorer places reflects political power that has expanded within the capital and that has attracted migrants for rural areas (Ades and Glaeser, 1995). The agricultural trade hypothesis argues that today’s poorer countries have a comparative disadvantage at farming and consequently have specialized in more urban tasks (Glaeser, 2014). The second section suggests that the agricultural trade hypothesis may explain why poorer countries have urbanized, but that the centralized power hypothesis explains why this has produced mega-cities (Davis and Henderson, 2003).
In the third section, I turn to the positive and negative consequences of massive poor world urbanization. The most obvious upsides of urbanization are economic. Agglomeration economies imply that mass urbanization should lead to short-term increases in productivity; more suggestive evidence hints at the possibility of dynamic benefits if cities increase the speed of economic growth. Cities may engender longer-term growth directly, by increasing the flow of knowledge and the production of new ideas, and indirectly, by facilitating investment in physical and human capital.
The downsides of density weigh against these benefits. Urban crowding leads to traffic congestion and the spread of contagious disease. Urban proximity may also abet crime and it is more expensive to build in dense urban areas. These downsides imply a need for effective urban governance to address the negative externalities that occur in dense neighborhoods.
The fourth section turns to the governance problem in cities, which I separate into policy and politics. The technocratic approach to governing mega-cities emphasizes better policies, such as regulations that nudge slum dwellers to adopt clean water and sewerage, or congestion pricing. But these policies cannot happen as long as governments are weak and unwilling to act in the public interest. Consequently, I also discuss the political development of developing world mega-cities.
One question is whether privatization or independent public agencies can meaningfully improve the quality of city life. There are examples of benign privatization and benign parastatals, but also examples of massive corruption and indolence. I follow Djankov, Glaeser, La Porta, Lopez-de-Silanes, and Shleifer (2003) and suggest that developing world cities have a tradeoff between disorder and dictatorship, and that urbanization may lead to either more demand for governmental control or more demand for freedom. Finally, I discuss the “Boston Hypothesis,” raised by Glaeser and Steinberg (2017) which suggests that cities will improve their own governments through democratic revolution and peaceful reform. While urban concentration does make regime change more likely, it is less clear that stable democracies will naturally emerge from even successful uprisings.
The urbanization of humankind
In 1910, when my grandparents were young, less than one-third of Europeans lived in cities with more than 5,000 inhabitants. The only countries on earth that were mostly urban were Belgium, the Netherlands, and the United Kingdom (Bairoch & Goertz, 1986). The United States became a predominantly urban nation only in 1920, when its per capita income level was above $8,000 in 2016 dollars. By 1967, when I was born, Europe and North America were overwhelmingly urban and Latin America was also rapidly urbanizing, but Africa and Asia were still overwhelmingly rural. In that year, fewer than 36% of humans were urban. Today, 54% of humanity is urban and poor countries, as well as rich ones, enjoy the benefits and experience the costs of dense cities.
This change is without precedent. Throughout almost all of history, the experience of humankind has been non-urban. We evolved along the hot banks of the Awash River in the Afar Triangle in modern Ethiopia. We needed abundant land to survive as hunter-gatherers and so we spread globally over hundreds of thousands of years.
About 10,000 years ago, the Neolithic Revolution enabled our ancestors to produce significantly more calories from a fixed amount of land. That meant that humans could become more sedentary and villages began to emerge, such as Catalhoyuk in Southern Anatolia. The first cities appeared in the fourth millennium before the Common Era in Mesopotamia (Smith, Ur, & Feinman, 2014), unsurprisingly near the heartland of the first agriculturalists. Agricultural productivity enabled urbanism then, as it would throughout almost all of human history.
Politics and trade seem to have been the forces that brought humans together into urban areas. Memphis grew great as the capital of ancient Egypt and housed perhaps 60,000 people four millennia ago, making it the world’s largest city. Tyre emerged as a trading powerhouse selling its purple dye throughout the Mediterranean. Commercial and imperial cities have coexisted ever since, but before 1800, the true urban giants were almost always the capitals of great empires.
Rome, Xi’an, Baghdad, Kaifeng, Hangzhou, Istanbul, and Beijing may have all reached one million inhabitants before the industrial revolution.1 While their level of economic development was low by modern standards, these were all the capitals of great empires. Empires typically become great through public capacity in conquest and administration, and consequently these cities were governed by the strongest states of their time.
Julius Caesar might not have had access to electronic road pricing, but he was able to ban wheeled vehicles from Rome during daylight hours. Xi’an had an amazingly sophisticated set of sanitary canals. Kaifeng built great towers for spotting and responding to fires. The downsides of density were mitigated by public competence.
Smaller, commercial cities began to reappear in Europe during the Middle Ages. By 1300, Florence, Genoa, and Venice each had about 100,000 inhabitants. In the north, the trade and weaving center of Bruges had a population of 50,000. As trade moved from the Mediterranean to the Atlantic, northern powerhouses such as Antwerp, Amsterdam, and London emerged. Similar trading hubs, such as Boston, New York, and Philadelphia would appear on the other side of the Atlantic.
The great watershed in urban growth occurred during the industrial revolution, which was preceded by revolutions in agriculture and transportation. The productivity of agricultural land rose dramatically during the 18th century due to crop rotation, the Dutch plough (borrowed from the Chinese during the 17th century), and the spread of the potato (Nunn and Qian, 2011). An explosion of English canal building made it easier to move food over space. Urban expansion during English industrialization was only possible because these earlier advances had made it possible to feed so many urbanites.
Industrialization then acted as a pull on former farmers and their children who were able to find work in urban textile mills and car factories. The earliest factories were typically located near water sources that lay outside of older urban cores, such as the Merrimac River in Massachusetts or the Cromford Sough in Derbyshire. Continuing improvements in steam engine efficiency meant that industrial production could be powered by coal, rather than water, and could therefore be urbanized, albeit at a cost in urban air quality. Factories located in cities like Detroit, Manchester, and Gothenburg, because of access to ports and later rail lines, enabled industrial supplies to be readily shipped in and finished products to be shipped out.
In a sense, heavy industry was always a strange match for urban density. Factories are space intensive, and they produce pollution that makes a city less attractive as a place to live. During the era of Henry Ford, factories became massive self-contained units and it was hard to see why such entities benefitted from proximity to other urbanites. Ford himself moved his operations from downtown Detroit to the enormous, suburban River Rouge complex during and after World War I. After World War II, heavy industry fled big cities far more dramatically and many cities experienced depopulation along with deindustrialization.
The mid-20th century belonged to the automobile and the air conditioner, and America built endless acres of Sunbelt sprawl. As richer urbanites fled to California or at least to Westchester, city populations began to decline during the 1950s. By the 1970s, almost all of America’s older colder cities seemed like they were headed for the trash heap of history. Seattle seemed as doomed as St. Louis. New York’s prospects looked no better than those of Detroit.
But then something unexpected happened. In the 1990s, many city populations stabilized in the wealthy West, and some cities started experiencing robust growth after 2000. Some older cities, including New York, London, Paris, and Frankfurt, became places of extraordinary wealth. High housing prices signaled the demand for urban lifestyles.
The growth was not uniform. Skilled cities were far more successful than unskilled cities, both in terms of population and income growth (Glaeser & Saiz, 2004). The success of Seattle relative to Detroit can be readily explained by the fact that 57.5% of Seattle’s adults have a college degree or more, while only 13.1% of Detroit’s adults are so well educated. Entrepreneurial human capital, at least as measured by average establishment size or the share of employment in new establishments, is also a potent predictor of urban employment growth (Glaeser, Kerr, & Kerr, 2015).
One interpretation of these facts is that globalization and new technologies did initially disperse urban populations, but these forces also increased the returns to human capital and innovation. Literally hundreds of studies have confirmed the rise in returns to skill since the 1970s (e.g., Goldin & Katz, 2009). Cities, like schools, have the capacity to build skills, because cities enable workers to learn from the people around them (Glaeser, 1999). Declining costs of electronic communication can lead to a more interactive society that ultimately favors urban density (Gaspar & Glaeser, 1998). The dependence of so many Western cities on financial services is perhaps unsurprising since the returns to better information are so high in that sector.
Western cities have also thrived as centers of consumption as well as production (Glaeser, Kolko, & Saiz, 2001). As city streets have become safer, prosperous people are willing to pay a premium to enjoy urban amenities, including museums, restaurants, and quirky urban shops. The rise of the consumer city is demonstrated by the correlation between urban amenities and city growth and by the rise of reverse commuters: people who are willing to pay the high cost of urban living despite working in a suburb. The increasing importance of urban amenities helps explain why urban housing prices have often risen more than urban wages.
Indeed, one reasonable view is that in the 21st century, the best local economic development strategy is to attract and train smart people and then get out of their way. Attracting smart people depends on quality of life, which can lure young people to Milan or repel them from older rustbelt cities. The enduring success of London, Stockholm, and San Francisco partially reflects the appeal of these cities as places to live as well as places to work.
Urbanization in East Asia
The growth of East Asian cities seems almost like a replay of the Western industrial revolution run at fast forward. High levels of agricultural productivity meant that Japan was more highly urbanized than France, Germany, or the United States in 1850. The country’s urban population was particularly centralized in the capital city partially because political power was so centralized in the hands of the Tokugawa Shogunate. When Japan opened itself to the West after the Meiji Restoration, industrialization occurred quickly and urbanization grew, and by 1950, three-quarters of the Japanese population were urban.
By contrast, Korea was an overwhelmingly rural nation at the start of the 20th century. During the 1930s, Korea’s Japanese masters promoted the growth of war-related industries and Seoul started expanding. Yet Korea’s urbanization rate only reached 14% at the end of World War II. The bombing of Korea’s cities, especially Seoul, during the Korean War helped ensure that Korea’s urbanization rate remained close to 30% in 1960.
But then over the next 40 years, Korea’s urbanization rate soared from 30 to 90%, an almost unbelievable change. While U.S. and European urban systems were dispersed, partially because 19th-century cities needed to be close to natural resources, Korea’s urbanization was dominated by the expansion of Seoul, which is now larger than New York City. Korea has few natural resources other than the wits of the Koreans themselves, and consequently, manufacturing could be highly centralized. The region around Seoul contains 24 million people and is one of the largest metropolitan economies in the world.
Korea’s urbanization was driven by a dramatic rise in government-supported export-driven industrialization, which moved up the technology ladder swiftly. This movement is typified by Samsung’s moves from sugar refining (a classic pre-modern urban industry) to textile production during the 1950s and then into electronics during 1960s. Today, the Seoul-based behemoth is a global technology leader. South Korea has moved relatively easily from industrial to post-industrial economy because the government invested so heavily in human capital, meaning that Seoul is far more like Seattle than like Detroit.
China’s urbanization has been even larger ...