By Ching-Yang Lin and Hiroaki Miyamoto
This paper studies how well a search and matching model can describe aggregate Japanese labor market dynamics in a full information setting. We develop a discrete-time search and matching model with productivity and separation shocks and use it as a data-generating process for our empirical analysis. Using Bayesian methods, we estimate the model for data on unemployment and vacancy postings in Japan. We find that the model is successful in matching the volatility in unemployment and vacancies while it does not match the volatility of output and wages. We also find that both productivity and separation shocks contribute to movements in unemployment and vacancies, but productivity shocks more so.
It is interesting to study a very different labor market for once, but I am not sure the model makes the data justice. From my limited knowledge of the Japanese economy, I recall that up to the 1990’s, lifetime employment in the same firm was pretty much guaranteed. This changed after the lost decade of the nineties. The model should at least allow for this change in separation rates, or incorporate some sort of insurance contract the firm offers to its employees (and sometimes in it changes).