Identifying Aggregate Supply and Demand Shocks in South Africa

CID Faculty Working Paper No. 164

Stan du Plessis, Ben Smit, and Federico Sturzenegger
May 2008

A publication of the CID South Africa Growth Initiative

Abstract

This paper uses a structural VAR methodology to identify aggregate demand and supply shocks to real output for the South African economy. Demand shocks, in turn, are separated into fiscal and monetary shocks. The model is estimated with quarterly data over two overlapping samples: 1960Q2-2006Q4 and 1983Q4-2006Q4. The identified (structural) shocks were used in a historical decomposition to split output into a measure of potential output (resulting from the evolution of supply shocks) and a measure of the business cycle (the gap between actual and potential output). This measure of potential output suggests a significant decline relative to trend in the years prior to the political transition of 1994 and a swift reversal thereafter. The paper presents evidence from three sources to support its identification of aggregate supply and demand shocks. These sources are the following: theory consistent impulse response functions; a close match between the implied measure of the business cycle and independent information about the South African business cycle; and a demonstration of the close match between the identified series of aggregate supply shocks and important historical events in the decades prior to and following 1994 that have been identified by economic historians as important shocks to the South African economy.

Keywords: South Africa, aggregate supply, aggregate demand, monetary policy, fiscal policy, potential output, long-run restrictions

JEL subject codes: C25, C41, E32