That’s the topic of a new FEDS Note that I just published with Hie Joo Ahn. Here’s what we discuss:
The U.S. unemployment rate averaged 8.4% during the first five years of recovery from the Great Recession of 2007-2009, the weakest recovery on record. But as the expansion continued, unemployment continued to decline and by 2018 reached the lowest levels in almost half a century. Why did unemployment remain so high for so long, and what factors contributed to the recent lows?
Economists rack their brains to come up with new explanations of the slow recovery, even inventing new classes of workers to blame it on.
It’s pretty simple. For the first time in recent history a recession was accompanied by simultaneous fiscal austerity and contraction of government employment.
By the way, and I won’t mention any names, but certain economists near here were very distraught at stimulus spending in the midst of the worst recession since the Great Depression.
“The U.S. unemployment rate averaged 8.4% during the first five years of recovery from the Great Recession of 2007-2009, the weakest recovery on record (see Figure 1). But as the expansion continued, unemployment continued to decline and by 2018 reached the lowest levels in almost half a century. Why did unemployment remain so high for so long, and what factors contributed to the recent lows?”
What follows is an interesting discussion from the perspective of individuals. Let me just add a macroeconomic comment. We could have gotten back to full employment more quickly had we: (a) listened to Christina Romer’s call for a much more vigorous fiscal stimulus in early 2009; and (b) helped the states with Federal revenue sharing so we could have avoided their unfortunate move to fiscal restraint in 2011.
JDH Couple of questions:
(1) Are you saying that once an individual falls into one of the two categories, that particular individual’s hazard rate remains more or less constant throughout?
(2) Have you applied this approach across different regions or states, or just at the national level? In other words, do differences in states’ safety nets have any effect?
(3) Can you explain why the PL probabilities in Figure 4 don’t seem to change regardless of the state of the economy? You do mention that they tick up a bit during recessions, but it looks like they oftentimes also tick up a bit during expansions I’m not smart on labor economics, but my understanding is that most unemployment duration models have some kind of covariate for economic activity; e.g., industrial production index or whatever.
2slugbaits: (1) Our specification allows the probabilities for a given group of individuals to change a little every month, so that the circumstances facing any given person could be different from the month before. But as you observe in (3), a surprising finding is that the probabilities for a given group don’t change that much even during a recession. What seems to happen instead is that there are more inflows into the pool of unemployed of those who have trouble finding new jobs. (2) There aren’t enough data to do this at a state level, though we have looked at broad categories based on a number of observable characteristics, such as age, gender, industry, and reason for unemployment. We have found that within every category, there are still tremendous differences across the experience of different people that can be well represented by these two groups. The one observable category that matches most closely with our designation of those in group L is those who indicate that the reason they are unemployed was a permanent involuntary separation from their previous job. (3) We conclude that the primary thing that changes in a recession is the kind of people who lose jobs and the circumstances under which they lose those jobs. We interpret this to mean that the driving factor behind unemployment during recessions is increased layoffs of individuals who, at least in those circumstances, have great difficulty finding new work.
As Mr Spock used to say Fascinating