The dynamics of agglomeration economies and channels of economic change

Since the 1980s spatial inequality within countries has been increasing. This thesis focuses in three parts on how certain agglomeration advantages of cities, known as agglomeration economies, have changed and why: 1) According to Marshall (1890) agglomeration economies consist of local advantages in exchanging, respectively, products, employees, and ideas. In this chapter, the relevance of each is measured over time by gathering historical U.S. data on the coagglomeration of industries and the extent to which these buy/sell products from each other, employ similar workers, and patent similar technologies. The results show that since 1970 the local exchange of ideas has become more relevant whereas exchanging products and employees has become less relevant. These changes are influenced by the rise of computer technology, as measured by the decreasing share of routine task intensive jobs, and trade competition from low-wage countries that make industries more knowledge-intensive. Surprisingly, the decreasing transportation costs of goods is not significantly associated with the decreasing importance of input-output linkages nor with the other two agglomeration determinants. 2) The division of labour as agglomeration advantage is often attributed to Adam Smith (1776), who argued that physical proximity facilitates the specialisation of individuals as the fruits of their labour are more easily exchanged. In this chapter is argued that this proximity is more relevant for complex activities where individuals are specialised in more narrow fields and need face-to-face contact to collaborate in larger networks. Results confirm that more complex jobs, industries, patents, and publications are more strongly concentrated in larger cities. U.S. patent data shows that since 1830 the concentration of complex activities in large cities has increased over time, in particular during technological revolutions, which is in line with the observation that the expanding frontier of human knowledge requires an ever-larger division of labour and a stronger spatial concentration of human activities. 3) Cities also have advantages in developing future activities by building on current regional capabilities, so-called dynamic agglomeration economies. This aspect is particularly relevant for regional resilience, as the ability to overcome crises lies partly in the possibility of regions to develop new economic activities when others fade. Using the same patent data, it is shown that cities with a larger variety in technological activities are better able to develop new activities during the great crises in the history of the U.S.A.: the Long Depression (1873-1879), the Great Depression (1929-1934), and the Oil Crisis (1973-1975). The former and latter crisis co-occur with technological revolutions when general purpose technologies come up that revolutionise work and production processes. During these time periods advantages in reinventing regional economic structures become particularly relevant to adapt to new economic realities. All in all, this thesis offers insights why larger more innovative cities grow faster than other regions: 1, they have more knowledge-oriented agglomeration advantages; 2, they accommodate the growing division of labour of complex activities; and 3, they reinvent themselves more easily. The technological computer revolution and globalisation are instrumental to understand the recent growth in spatial inequality