Inflation-unemployment Trade Offs in an Open Economy: Theory and Evidence for Australia

It is now widely accepted that inflation is generated by a combination of excess demand and inflationary expectations. The main objective of this paper is to present a simple combined model of wage and price inflation that incorporates these two features in a consistent manner. The model is constructed so as to pay particular attention to open economy features. The general characteristics of the model are derived from the static neoclassical theory of firms and households. To this structure we add the following dynamic adjustment assumptions: tradeable prices adjust immediately to world tradeable prices adjusted for the exchange rate, while wages and non-tradeable prices adjust to existing and anticipated excess demands in their respective markets. This procedure leads to a pair of excess demand-expectations equations in wages and non-tradeable prices. Empirically, we measure expectations about any given variable at any point in time as an optimal extrapolative function of past observations on itself. The system, as constructed, therefore contains three levels of information about the wage-price relationship: past information; information contained in the model itself; and, any additional information contained in the contemporaneous covariance of wages and non-tradeable prices. All three levels of information are used in estimating the system by a variant of Zellner’s seemingly unrelated regressions technique. In general terms the simple model appears to fit the data quite well and the theoretical properties of the system are supported by the estimates. This framework allows at least two interpretations of the long-run Phillips curve relationship between wage inflation and unemployment. The first concerns a situation in which the exchange rate is fixed. In this case, the domestic cycle of wages and non-tradeable prices has a built-in stabiliser around the trend of world tradeable prices. Of course the world price trend may itself be destabilising in some instances. Nevertheless, taking the world price trend as given we can define a long-run trade off relative to this trend. We call this the fixed-exchange-rate trade off. Attempts to exploit this trade off to lower unemployment would involve domestic inflation of non-tradeable prices at a rate in excess of the world tradeable price trend. The second concerns a situation in which the exchange rate moves over time in such a way as to preserve purchasing power parity. We call this the fixed-relative-price trade off. It could equally be called a variable-exchange-rate trade off. The data for Australia in the 1960’s and 1970’s suggest that Australia faces a vertical long-run fixed-relative-price trade off but a negatively-sloped long-run fixed-exchange-rate trade off. Thus, if the authorities could maintain, with a fixed exchange rate, a long-run inflation rate for non-tradeables that is divergent from the world inflation rate for tradeables, then some long term gain in unemployment could be achieved at the expense of a higher rate of inflation. While this assumption would be untenable to most economists, the essence of this result is that the fixed exhange rate decreases the slope of the trade off over the time horizon relevant to policy-makers.