Three essays on labor market volatility, monetary policy and real wage stickiness.

Summary The starting point of this thesis is related to Shimer's criticism, i.e. the inability of the matching model to reproduce the high volatility of the unemployment rate. The introduction of real wage rigidity was seen as the answer to this criticism. Nevertheless, Sveen and Weinke (2008) pointed out that this rigidity would have no effect on volatility when firms choose the number of hours worked per worker. In Chapter 1, we show that this result is related to the way wage rigidity is incorporated. We show that when wage rigidity is introduced by Hall and Milgrom's (2008) "credible bargaining", volatility is strongly amplified, even when firms choose the number of hours per worker. Moreover, a significant trade-off between inflation stabilization and unemployment stabilization is restored. In chapter 2, we point out that credible bargaining, however, produces only a moderate degree of wage rigidity and is only able to fully replicate unemployment volatility for unrealistic values of some parameters. We add an information asymmetry to this model. The resulting increase in wage rigidity allows us to replicate the volatility with a plausible calibration. Finally, we show that it is possible to solve the Shimer critique without resorting to wage rigidities, but by considering a particular calibration of the model with endogenous job destruction. We also highlight a central mechanism of this model, such that the volatility of the destruction rate amplifies the volatility of the return to employment rate.