April 19, 2010
Data Based Assessment of the Unemployment Insurance Impact on UE
Although economists have shown that extended availability of UI benefits will increase unemployment duration, the effect in the latest downturn appears quite small compared with other determinants of the unemployment rate. Our analyses suggest that extended UI benefits account for about 0.4 percentage point of the nearly 6 percentage point increase in the national unemployment rate over the past few years. It is not surprising that the disincentive effects of UI would loom small in the midst of the most severe labor market downturn since the Great Depression.
The Valletta/Kuang empirical test compares the unemployment rates for quits/entrants (not eligible for UI) against job losers (eligible for UI), and finds only a small difference. This point is highlighted in the following graph.
Figure 2: from Valletta and Kuang (2010)
For earlier discussion of UI, see this post.
Posted by Menzie Chinn at April 19, 2010 02:48 PMdigg this | reddit
No one wants to sit around all day with nothing to do. Only people who are well off can afford to be unemployed for long without getting incredibly bored, or running up credit card debt.
Posted by: Armon Ketabian at April 19, 2010 11:05 PM
Why did they use total job losers when they could have used job losers not on temporary layoff? Their job loser measure is contaminated with a significant number of job losers on temporary layoff who will have much lower unemployment durations. Temporary layoffs will be recalled regardless of the nature of unemployment benefits.
Posted by: tj at April 20, 2010 08:09 PM
One can work part-time and still keep unemployment. It gets pro-rated and one comes out 'ahead' , as opposed to just staying at home.
I suspect a large percentage of people on unemployment are working part-time. The lazy meme doesn't make any sense.
Posted by: purple at April 21, 2010 02:17 AM
A few points:
1. I’m not sure I agree with their control groups as I think there is likely to be some significant share of people that will identify themselves as having lost their job “through no fault of their own” when money is on the line, but will still chose to identify themselves as “voluntary job leavers” rather than “involuntary job losers” when no money is involved.
2. Taking a look at the insured unemployment numbers published by the BLS shows that the number of insured unemployed peaked in the second quarter of 2009 at a level rough 2-3/4 times its level prior to the recession. Since then it has fallen. This timing lines up well with their figure 2 which suggest the biggest difference between the red and blue lines occurred in early 2009. So while I think Valletta and Kuang's numbers may have some downward bias as to the effect of the extensions of unemployment insurance on the increase in the unemployment rate, the general timing that their method gives probably works.
3. Nevertheless, I don’t think their results differ that much from the general view that extensions of unemployment benefits probably don’t have that much impact in times of significant labour market slack, but extensions do have an effect when the labour market is good. (This is because when there are a lot of people looking for jobs and few jobs then each job will probably find a person to fill it quickly regardless of the search intensity of the unemployed. On the other hand, when there are few people looking for jobs and many jobs then the search intensity of the unemployed will matter.)
Posted by: AKD at April 21, 2010 05:46 AM
Let's see - UI based on a 6% rise accounts for .4% so we say it is nothing to worry about. A simple back of the box extrapolation then tells us that almost 3/4% of our total unemployment problem then is directly attributable to congress passing UI. Hmmmmm!
Then what about that last paragraph?
Despite the relatively minor influence of extended UI, it is important to note that the 0.4 percentage point increase in the unemployment rate represents about 600,000 potential workers who could become virtually unemployable if their reliance on UI benefits were to continue indefinitely. This in turn would raise the minimum attainable unemployment rate by a similar amount, a problem that has been noted as an outcome of the generous UI systems in some advanced European countries. Given the experience with the elimination of extended UI benefits during previous U.S. economic recoveries, a permanent increase in the U.S. unemployment rate is unlikely.
Or, back of the box again, almost 1 million of the total unemployed are because congress passed UI!
Posted by: RicardoZ at April 22, 2010 02:28 PM