Urban Growth Across Three Continents: The Blessing of Human Capital and the Curse of Volatility

Defence date: 19 May 2008Jury Members: Professor Frederick van der Ploeg, University of Oxford, (Supervisor) Professor Giancarlo Corsetti, EUI Professor William Strange, University of Toronto, Canada Professor Harry Garretsen, University of Utrecht, The NetherlandsThis dissertation is concerned with the empirical determinants of urban growth in developed countries, and the effects of macroeconomic volatility in less developed countries, on urbanization and long run income growth. Four chapters are presented. Chapters 2 and 3 deal with urban population growth in the USA and employment growth in Germany. In both cases metropolitan areas grow faster if the initial concentration of highly skilled workers is higher. Chapter 2 looks at which city characteristics lead some cities to attract more skilled workers than others, resulting in a diverging pattern of skill concentration among cities. The chapter shows that this is not a selfreinforcing process. One of the more robust findings is that a personal services sector is a positive amenity. The number of skilled workers is expected to increase faster in metropolitan areas where this sector is abundant. The personal services sector is intensive in low-skilled labor which gives a new sound to the popular belief that only a highly-skilled culture sector will attract more skilled workers. The second chapter shows that divergence of the concentration of skills and its effect on metropolitan growth is very similar in Germany. It also shows that there may be positive interaction effects between similar but different skill levels. For example, workers with vocational training have a larger effect on total employment growth if the local concentration of technical professionals is high as well. In Chapter 4 we turn to the developing world. Cities in developing countries have been growing very fast, but surprisingly also in periods when economic growth to create manufacturing jobs was absent. Rural-urban migration is modeled as a response to aggregate volatility. Households in rural areas cannot cope with negative shocks and are forced to move to cities which offer more diverse sources of income, because financial markets are incomplete and households have credit constraints. We furthermore show from country-panel evidence that rural natural resource production (including agriculture) is much more risky than urban manufacturing sectors. Shocks come from natural causes such as rainfall, and from volatile world resource prices. This mechanism may be more important than the traditional view which models urbanization as a transitory process when countries move from an agricultural base to a manufacturing base. Chapter 5, in collaboration with Frederick van der Ploeg, probes deeper into the ii detrimental effects of aggregate volatility. The high volatility of world prices of natural resources causes severe volatility of output per capita growth in countries that depend heavily on them. This has a robust negative effect on long-run growth itself and presents a new explanation to the natural resource curse. Volatility can fortunately be substantially reduced provided that countries have a sound financial system to cope with large and sudden fluctuations in resource income. The curse can even be turned into a blessing because we find evidence for a positive direct effect of natural resource dependence on growth after controlling for volatility. However, the indirect negative effect of resource dependence on growth, via volatility, is much larger than any direct positive effect