Regional Indexes of Activity: Combining the Old with the New

This paper proposes a framework to construct indexes of activity which links two strands of the index literature – the traditional business cycle analysis and the latent variable approach. To illustrate the method, we apply the framework to Australian regional data, namely to two resource-rich and two service-based states. The results reveal differences in the evolution and drivers of economic activity across the four states. We also demonstrate the value of the Index in a broader context by using a structural vector autoregression (SVAR) approach to analyse the effects of shocks from the US and from China. This Index-SVAR approach facilitates a richer analysis because the unique feature of the index method proposed here allows impulse responses to be traced back to the components.