What Can Growth Rates Tell Us? A Short-Run Decomposition Method

Consider time series output data for two sectors, industry and agriculture. By examining just the output data themselves, what can we say about the relative impact of institutional/policy factors, intrasectoral competition for resources, and intersectoral linkages on each sector’s growth? Currently the answer might be very little. Our aim is to fill this gap: First, we explain how institutional/policy and other factors can be formally derived from a growth rate term. Second, we offer an empirical illustration of the derivation, such that just the time series output data of the two sectors by themselves contain enough information to make inferences regarding the relative impacts of the institutional/policy and other factors. Thus we provide the formal decomposition of a growth rate term, allowing the relative impacts of key explanatory variables to be estimated from a highly parsimonious data set. For countries that publish limited data sets, our method extends the ability of researchers to make inferences about the impact of institutions and so on, even when data on institutions are unavailable.