Reflection on Microeconomic Adjustment Hazard Approach

This paper studies the way the adjustment process takes place in labor demand when it is expressed as a Cox proportional hazard model. I use a simulated firm-level panel data based on a threshold model with periods of high and low frequency of employment fluctuations, which is consistent with the infrequent way the adjustment process takes place according to the new theories of adjustment. I model the probability that a firm adjusts its employment level during a time-period as a Cox proportional hazard function dependent on the deviation of its actual employment value variable from its target. I show that the aggregate employment change, based on a high proportion of firms experiencing large employment fluctuations, could be very well represented by the Cox proportional hazard and also could be very well approximated by the empirical mean of the product of the hazard function and the deviations. On the other hand, I show that the aggregate employment change based on a very low proportion of firms facing large employment adjustment can be well represented by a quadratic (nonlinear) adjustment hazard. Finally, I try to conclude that in order to construct a measure of deviation from the target level (which is the state variable of the model) the regression of the employment fluctuation on the wage fluctuation could be more helpful than the regression of the employment fluctuation on the hours fluctuation.