6.2Some Techniques for Explaining Spatial Productivity Differences

Explaining economic growth is an important task for economists. Traditionally, studies have focused on understanding the process at the level of the aggregate economy. The Solow-Swan model helped to explain output growth in many countries. It has also been used in cross-country analysis, where it left most of the disparities unexplained, contained in the “Solow residual” term. More recently, with the ongoing globalization phenomenon and the fluidity of national boundaries, it has become more relevant to use smaller units of analysis: the region, the state or the city. For instance, productivity levels vary tremendously between these units, and many authors have tried to explain these variations. Moomaw (1983) summarized the earlier work on this subject. Pioneer authors found that the main sources of spatial variations of productivity were: the capital-labor ratio, the age of the capital stock, the rate of technology adoption and diffusion, the quality of the labor force and agglomeration economies. This section presents a survey of the earlier work on spatial growth variation through shift-share analysis, various empirical techniques, and the sources of growth framework. This description is largely inspired by the work of Gerking (1994). Finally, different techniques used to study agglomeration economies (i.e. the use of particular production functions) are presented. Following Abdel-Rahman (1988) and Rivera-Batiz (1988) I will show how product differentiation and monopolistic competition relate to agglomeration economies.