Unemployment, growth and exchange rate dynamics

Chapter 1 presents a modified Mundell-Fleming-Dornbusch model that allows analysis of the effect of changes in long-run output expectations on the exchange rate. The model is used to simulate the evolution of the euro-dollar exchange rate under two assumptions regarding monetary policy: constant-money-supply policy and constant-price policy. The dynamics that follow a change in long-run output expectations under both scenarios are derived and actual data is used to calculate the change in expectations that is needed to match the euro-dollar experience. The quantitative results show that a small change in long-run output expectations is enough to generate the recent euro-dollar dynamics.; Chapter 2 quantifies the importance of Spanish labor market institutions in determining the response of Spanish unemployment to macroeconomic shocks. To do this I develop a model that allows for the analysis of the unemployment impact of a macroeconomic shock under two alternative institutional frameworks: a competitive labor market and a labor market that is characterized by insider wage setting. The difference between the unemployment effect of a macroeconomic shock in the competitive case and the insider case is interpreted as the unemployment effect of the institutional framework. That is, the effect that is not purely due to the macroeconomic shock. The empirical application of this idea is performed with the Differences-in-Differences estimation method.; Chapter 3 shows how the empirical literature on economic growth that is based on per-capita data suffers from omitted variables bias that leads to estimates of the speed of convergence that are too high. This problem arises because the predictions of the neoclassical growth model apply to per-worker data, not per-capita data. The model is reconciled with per-capita data by introducing the distinction between working and non-working population. When this is done, two new variables appear to be relevant for growth during the process that leads to the steady state of the economy, the difference between the growth rate of working and dependent population and the initial dependency ratio. The empirical results suggest that the magnitude of the bias on the estimated speed of convergence β is about .4 to .6 percentage points.