Labour market dynamics in the euro area: A model-based sensitivity analysis
Alistair Dieppe,
Jerome Henry () and
Peter Mc Adam ()
Additional contact information Peter Mc Adam: European Central Bank
Authors registered in the RePEc Author Service: Peter McAdam ()
Abstract:
The lack of euro area labour market flexibility is a commonly mentioned issue. In particular, the relatively weak response of wages to high unemployment can pose adjustment problems. We address the issue using extensive simulations of an estimated macro-econometric model for the euro area (the Area Wide Model). The model is characterised by a well-pinned down, unique equilibrium unemployment rate but with relatively slow convergence dynamics. The interaction between unemployment dynamics and the policy conducted is investigated under a number of alternative assumptions on the labour market, using both deterministic and stochastic simulations. Resulting impulse-responses are analysed, as well as the distribution of unemployment (stacked across a large number of simulations), with a particular focus on situations where large deviations from the deterministic steady state occur. The analysis is done, first, with the standard version of the model, to provide a benchmark consistent with its estimated parameters. A similar exercise is then conducted allowing for a stronger wage response to unemployment and for a higher employment elasticity to real wages. In turn, a specifically adverse configuration is analysed, where, due to imposed asymmetries in the Phillips curve, high unemployment results in lower wage responses than those estimated. Such a feature could put the economy in a situation where unemployment remains well beyond its long-run equilibrium value for a protracted period. We then assess the performance of alternative specifications of the Taylor rule under the various labour market structures, with especially, different degrees of aggressiveness and forward-lookingness of monetary policy. One main result is that the differences across labour market configurations are significantly less marked once a forecast-based rule is employed.