Spatial Spillovers and Labor Productivity Convergence in Canada

Various econometric specifications and levels of data aggregation have yielded different results for labor productivity convergence. However, few studies have modeled convergence while formally accounting for regional spatial interactions, notwithstanding a wide recognition of their importance to convergence. Thus, this paper bridges this gap by studying labor productivity convergence across Canadian provinces by incorporating a spatial lag in a standard empirical specification of a neoclassical growth model. We show that ignoring spatial effects yields a significantly slower speed of labor productivity convergence. Furthermore, the decomposition of the spatial effects reveals that spillover effects dominate convergence, suggesting that the initial labor productivities in neighboring provinces significantly influence a province’s labor productivity growth rate