Lucas Paradox in the Short-Run

Abstract

This paper is concerned with whether the persistence of the Lucas paradox (that unlike what the classical economic theory would predict, capital flows to richer economies rather than poorer ones where marginal returns to capital are expected to be higher) within developing countries is because of the unobservable county-specific effects. Perhaps capital has been flowing to where it has already flowed and not necessarily where it had already been. Using five-year (rolling-averaged) panel data for up to 47 developing countries over the period 1980-2006, it examines if including the institutional quality index removes the Lucas paradox intertemporally (i.e. in the short-run). The ‘short-run’ relationships are captured by employing linear static (principally within-group fixed effects) and dynamic (system GMM) panel data methods. I demonstrate that the persistence in the Lucas paradox within developing countries is so entrenched that allowing for unobserved country-specific effects, within-group (time series) variation and autoregressive dynamics do not resolve the paradox.