Multinational Inflation under Fixed Exchange Rates: Some Empirical Evidence from Latent Variable Models

Abstract

T HE purpose of this paper is to study multinational inflation under fixed exchange rates in the context of a multivariate time series model of national inflation rates. In its most general form, the proposed model describes each national inflation rate as a distributed lag function of several common, serially uncorrelated plus a disturbance specific to that country. The common embody the innovations in the global monetary, fiscal, and wage-push variables thought to influence all inflation rates, while the country-specific disturbance captures events which impinge only on the home country's inflation rate. This representation has several features that recommend it as a framework for analyzing multinational inflation. First, special cases of the general model range from models, in which all national inflation rates are a function of one or more common world disturbances and the country-specific are identically zero, to models in which a specific disturbance has a substantial effect on the domestic inflation rate independently of global disturbances. Second, Johnson (1972), Kaldor (1976), and Gordon (1977), among others, have recently given much attention to the question of whether global monetary growth, fiscal demand pressures, and labor union militancy were all autonomous of inflation. Their hypotheses about the number of basic common causes of multinational inflation have a natural representation in the time series model as restrictions on the number of common generating the correlation among national inflation rates. Thus, we are able to test, for example, whether the inflation that simultaneously occurred in many western economies during the 1960s was the result of one common disturbance such as the exogenous growth rate in the world money (Genberg and Swoboda, 1975; Parkin 1976, 1977a) or, more specifically, excessive growth in the U.S. money stock (Johnson, 1972; Logue and Sweeney, 1975). This latter proposition, that U.S. monetary policy may be responsible for worldwide inflation, is a widely held view as evidenced by Parkin' s (1974) summary of the literature: