Three essays on economic geography: Theory and estimation (European Union)

In the last decade, spatial economics gained importance in the agenda of both academic researchers and policy makers. Among academic researchers, the "new economic geography" initiated in the early 1990s took traditional questions from urban and regional economics and brought them to mainstream economics. From a policy point of view, the phenomena of trade liberalization, the multinationality of the firm and the new ways of organizing production have raised questions about the geographical implications of these processes and on how to guide regional and industrial policies. This dissertation consists of three essays that develop and empirically test models to get a better understanding of the determinants of economic activity at three different geographical levels: urban, regional and national.; The first essay considers the effect of trade liberalization on the geographical distribution of the economic activity, which has long been a concern of the European Union (EU). I examine theoretically the ability of different public policies in altering the market outcome of agglomeration and then try to empirically assess the effects of the Structural Funds, the main instrument of the EU regional policy. In the second essay, I study the relation between wages and the distribution of skills in American cities. A recent strand of literature argues that labor productivity, and thus wages, depend not only on the average level of skills but also on their distribution over workers. Using manufacturing and Census data on individuals, I find that that, after controlling for the average level of skills in the city, the distribution of skills also matters. The dispersion of skills has a positive and significant effect on city average wages.; The third essay looks at the choice of Foreign Direct Investment (FDI) mode of entry across countries. Two major modes are considered: mergers and acquisitions (M&A) and greenfield investment. I empirically test the model on data of FDI entries into U.S. manufacturing. As predicted by my model, foreign investors seem to use M&A as a way to acquire specific assets, although this choice differs significantly across countries. The effect of specific assets on M&A is larger for Anglo-Saxon countries that are presumably closer to the U.S. in terms of corporate and organizational culture.