A study of hysteresis in the open Canadian economy

Abstract

This dissertation analyzes the phenomena of hysteresis in the Canadian economy from an open economy approach which combines studies of hysteresis in the international trade and the domestic labour market. In the theoretical analysis, two models for a small open economy are developed: one for the aggregate demand side and one for the aggregate supply side. The first one presents a hysteretic IS-LM-BP model which shows that trade hysteresis could result in a vertical IS curve and bend the aggregate demand curve until it is infinitely inelastic. A very steep aggregate demand curve explains the demand side hysteresis in the domestic economy. The second model derives an open economy version of the Phillips curve by allowing the exchange rate to enter the right-hand-side of the equation, and provides a theoretical base for empirical testing. The empirical analysis of this study tests trade hysteresis and estimates the Phillips curve. Following the widely used approach of trade hysteresis testing, the unit root and the cointegration tests are performed on the series of net exports and the exchange rate for the data sets of the US, the OECD, and the rest of the world. The results show that both of net exports and the exchange rate followed random walk, and there is no cointegration between the two series which is consistent with trade hysteresis. The Phillips curve estimation includes two steps. Step one uses the causality test to show that changes in the exchange rate cause changes in the inflation rate, but not vice versa, which supports the open economy version of the Phillips curve. The next step is to estimate the Phillips curve, and the degree of hysteresis is calculated from the results of the estimation. In alternative estimations with different measurements of the unemployment rate, different data periods, and different data frequencies, the results suggest that hysteresis has been significant in the Canadian economy.