Are benefit costs increasingly driving the cyclicality of employment?

Here is the job market paper of Grace Weishi Gu, from Cornell:

The Cost of Benefits and Employment Dynamics in Recent U.S. Recoveries

Abstract: I document (1) the slow rebound in U.S. aggregate employment following recent recessions, despite recoveries in output, as well as (2) a rising trend in per worker benefit costs, the cyclicality of those costs, and a positive correlation of the cyclical benefit costs with employment growth cycles. I show how these two phenomena are related. Then I develop a DSGE model that includes firms’ dynamic benefit costs, financial conditions (i.e., borrowing capacity), and the tradeoff between extensive and intensive labor margins to explore the effects of these features on post-1990 employment dynamics. I find that the benefit costs’ rising trend, cyclicality, and interactions with financial conditions have contributed significantly to the observed slow employment growth following the three most recent recessions in the U.S. This paper offers two main improvements over the standard model, which lacks such arrangements: (1) delivering 2-to-7-quarter delays relative to NBER business cycle troughs for the employment recoveries from the 1990, 2001, and 2007 recessions while generating no delay for the pre-1990 period, which is in line with the data; and (2) explaining 40-90 percent of employment volatility and harmonizing with that of output and per worker hours.

I would like to see more testing of this against alternative explanations, but still this is provocative and important work.