Financial Crises and Labor Market Recoveries : A Bayesian Evaluation

Average hours worked in the US recovers much faster than the unemployment rate following financial crises. Using an identified vector autoregression framework with nine quarterly US time series from 1984 to 2014, I find that an adverse financial shock leads to a fall in economic activity with a persistent increase in the unemployment rate but a transitory decrease in average hours worked. I then embed labor market frictions and financial frictions into a New Keynesian model to explain this stylized fact. The model introduces a relatively new financial shock the default cost shock, which captures the financial market’s inefficiency. I estimate the model using Bayesian methods and find that the default cost shock explains a significant portion of the variations in aggregate variables. In particular, this shock accounts for the business cycle features of a financial crisis better than the other relevant shocks do