On Wednesday, I blogged about some research by economists contending that changes in the unemployment rate were better explained by an unwillingness of firms to hire during a downturn, rather than the conventional view that the primary driver of unemployment was job losses.

But a reader points us to some good work from three University of Michigan economists that dispute some of this. The strongest claim in one of the earlier papers by University of Chicago's Robert Shimer was that job losses (i.e., flows into unemployment) didn't appear to react much to changes in the business cycle.

Using Shimer's own data, the three Michigan researchers -- Michael Elsby, Ryan Michaels, and Gary Solon -- show that this is not the case. While hiring decisions were still a major driver of the unemployment rate (accounting for about 65% of the changes in the unemployment rate), the Michigan trio also found a substantial role for job losses, which drove 35% of the changes. They add that the contributions of job losses to unemployment are especially pronounced at the beginning of a recession (meaning they probably explain a lot more than 35 percent of the unemployment rate at that point).

--Zubin Jelveh