The literature on the stylised facts of structural change in LDCs has been bedevilled by three problems: (i) drawing time-series inferences from cross-section results; (ii) endogeneity of the variables involved; and (iii) an inability to separate short-run from long-run effects. This paper uses...