The bell-shaped curve and regional growth
The core-periphery model has triggered a huge flow of extensions, which have contributed to make new economic geography one of the most lively research topics of the 1990s. Fujita et al.
(1999) is the first place to go, whereas Neary (2001) remains the best critical review of the canonical model. One of the main criticisms is that the model ignores the congestion costs generated by the gathering of people and firms within the same territory. In particular, armchair evidence shows that a human settlement of a sizable scale almost inevitably takes on the form of a city. As a consequence, a growing concentration of people intensifies competition for land and, therefore, leads to higher housing costs and longer commuting. In other words, even when nominal wages increase with employment density, housing and commuting costs, as well as pollution and crimes, make such large agglomerations less attractive.As transport costs steadily decrease, the spatial economy now moves through three phases instead of two: dispersion, agglomeration, and re-dispersion of the mobile sector (Helpman 1998). Agglomeration arises in the second phase for the reasons highlighted in the core-periphery model. The dispersion in the first and third phases emerges for very different reasons. In the first phase, firms are dispersed because shipping their output is expensive whereas, in the third phase, dispersion occurs because housing and commuting costs are too high for the agglomeration to be sustainable. Put differently, beyond some threshold congestion prompts firms and workers to re-disperse in order to alleviate the corresponding costs. At the limit, high commuting costs are sufficient to prevent the formation of a large city and guarantee the continuation of industrial activities within several small cities, a situation fairly characteristic of pre-industrial economies.
Another major shortcoming of the core-periphery model is that it overlooks the importance of intermediate goods. Yet, the demand for consumer goods does not account for a very large fraction of firms’ sales, often being overshadowed by the demand for intermediate goods. Therefore, in making their location choices, it makes sense for intermediate-goods producers to care about the places where final goods are produced; similarly, final-goods producers are likely to pay close attention to where intermediategoods suppliers are located. Giving intermediate goods a prominent role is a clear departure from the core-periphery model, which allows one to focus on other forces that are at work in modern economies. To this end, note that, once workers are immobile, a higher concentration of firms within a region translates to an increase in wages for this region. This gives rise to two opposite forces. On the one hand, final demand in the core region increases because consumers enjoy higher incomes. As in Krugman, final demand is an agglomeration force; however, it is no longer sparked by an increase in population size, but by an increase in income. On the other hand, an increase in the wage level generates a new dispersion force, which lies at the heart of many debates regarding the deindustrialization of developed countries, that is, their high labor costs. In such a context, firms are induced to relocate their activities to the periphery when lower wages more than offset lower demand (Krugman and Venables, 1995).
Finally, the vast majority of economic geography models rely on a fairly naive assumption regarding migration behavior: individuals care only for real wages. Leaving aside migratory movements triggered by wars, people are heterogeneous in their perception of the non-economic attributes of the different regions, and this heterogeneity affects the nature and intensity of migration flows, very much as it shapes consumers’ shopping behavior. Because labor mobility is also driven by non-economic variables, workers do not react to economic inequalities in the same way.
In such a context, the bell-shaped curve emerges again: workers move to the core when spatial inequalities are large but stay put when they are small (Tabuchi and Thisse 2002). This is because workers bestow increasing relative weight on non-economic factors affecting the quality of their life once they have achieved a sufficiently high material welfare. If this premise is correct, both economic growth and the development of the welfare state combine to slow down individuals’ mobility by allowing them to satisfy their needs for socializing and/or their attachment to a certain environment.The benefits of using the Dixit-Stiglitz model of monopolistic competition are reaped when we turn to regional growth because new economic geography and endogenous growth theories have been built on the same setting, thus making it easier to combine these two bodies of research within a unified framework. The contributions reviewed by Baldwin and Martin (2004) emphasize the possible geographical concentration of the innovation sector. Innovation being one of the main sources of the long-term growth of the economy, this innovation-driven concentration supplements the core-periphery effect to generate long-run patterns characterized by persistent and sizable income differences. That is, the predominant centers would retain the high value-added activities, whereas the routine activities would be relocated into the periphery. This challenges the unfolding of the bell-shaped curve and keeps open the debate on the spatial diffusion of economic development. However, thanks to new economic geography, we understand much better the various forces at work.