Measuring Economic Convergence

Abstract The sustainability of the ERM and the feasibility of EMU is dependent upon the economies of the EC achieving the appropriate degree of economic convergence. One of the difficulties in determining this is the absence of a broadly acceptable measure of economic convergence. In this paper we consider possible definitions of convergence for time series. To be useful the definition must allow for the stochastic nature of the data under consideration, and for the development of formal tests of a convergence hypothesis. These two considerations lead us to a preferred definition of convergence and a proposed framework for testing, based on a time varying parameter formulation and estimation by Kalman filter techniques. We consider the relationship between our proposed framework and other methods for testing convergence. It is argued that our method encompasses most of the tests that exist in the literature. Tests for convergence are then applied to exchange rates, interest rates and inflation rates for a sample of countries including members of the ERM. We conclude that the ERM may have helped to promote convergence in the nominal exchange rates. Although interest rate differentials have declined, convergence does not seem to be occurring. For inflation rates convergence can be rejected. Taken together these results suggest reasons for the recent disruption to the ERM and bring into question its sustainability. © 1997 John Wiley & Sons, Ltd.