Economic geography with spatial econometrics : a ‘third way’ to analyse economic development and ‘equilibrium’, with application to the EU regions

The main item of agreement between the ‘new’ and ‘old’ economic geography is the role of increasing returns in regional economic development. This provides a focal point for the model of this paper, which aims to highlight the existence of a ‘third way’ somewhere between the analysis provided by these two competing modes of explanation. Increasing returns are represented by the Verdoorn Law linking manufacturing output and productivity growth, which is augmented to include endogenous technical progress involving diffusion, spillover effects and putative human capital effects. The model is estimated using data for regions of the EU, thus emphasizing the need to confront theory with data. The approach of the paper thus avoids ‘the lost scientific cause’ of much of contemporary ‘economic geography proper’ and the constraints posed by the theory of ‘new economic geography’. The implications of the model are explored and assumptions are imposed leading to a ‘stochastic steady state’ as an approximation to real world turbulence, and as an alternative to the Markov chain stochastic equilibrium suggested by Quah(1993). The paper shows that the implications of interregional spillovers are faster productivity growth and higher productivity levels, a trend that is accelerated with endogenously determined spillover. Without catch up, regional productivity levels diverge with no stable steady state and one region becomes increasingly dominant, but catch up ensures that cross-regional productivity growth rates tend to equality.