Are Interest Rates Responsible for Unemployment in the Eighties ? A Bayesian Analysis of Cointegrated Relationship with a Regime Shift

To what extent can the persistence of very high unemployment rates in most of the European countries be attributed to the presence of high real interest rates ? This question, essentially addressed by the ‘customer market’ price)setting school, was very much debated in Europe these last years. It is empirically analysed for four European economies (Belgium, Denmark, France, Germany) and the USA in our paper. We use a bivariate cointegrating Var model with one endogenous breaking point between unemployment and real interest rate. Within this model and devising a new Bayesian approach, the weak and strong exogeneity of the interest rate is tested. For the four European countries the model is shown to be cointegrating providing a break point is allowed . The four posterior densities of the breaking point are very similar, when the classical estimates give more divergent and counter-intuitive information. For the four countries, the real interest rate is weakly exogenous, providing support to the hypothesis of long run causation of interest rates on unemployment after 1974. Short term causation is verified for only three countries.