A journalism student asked me some questions about the upcoming January employment situation release expected this Friday morning, and I thought I'd reproduce the Q&A's here:
In the reported figures last month the unemployment rate dropped from 7% to 6.7% and the Yahoo finance figures suggest the prediction is that it will drop by another 0.2 percentage in the report on Friday. Does this sound about right to you? Are we part of a drop in unemployment and do you think this trend will continue?
I suspect we will see another reduction in the unemployment rate with the January data, but whether it will be large or small is less clear. A reduction of 0.2 percentage point seems reasonable but might be conservative.
Overall, the economy looks like it is continuing to recover, albeit not very rapidly. So we would expect the unemployment rate to drift downward, back toward a long-run equilibrium level around 5 or 6 percent. In normal times, the unemployment rate is a nonzero number because of the usual pace of job creation and destruction. Exactly where it ends up in the long run is a function of workers' skills relative to employers' demands.
The January number might well be even lower than the current forecast of 6.5% because of the December 31 expiration of extended unemployment benefits. Because people have to declare themselves as unemployed --- out of work but actively looking for work --- in order to receive benefits, the end of the extension program is likely to reduce the incentive to declare as unemployed. It may have prodded people to take jobs they would otherwise not have, or it may cause them to effectively withdraw from the labor force by not actively looking for work. Either way, the number of unemployed seems likely to fall. The labor force may also shrink if the long-term unemployed stop looking for work, but in percentage terms any such decline in the labor force must be smaller than the decline in the number of unemployed through this channel. Thus the unemployment rate would fall, because the numerator would be falling faster than the denominator.
Do you have any sense of where the jobs are dropping or increasing in terms of region or sector?
No, not much of a good sense. The broader story since the Great Recession has been increases in services and in the health sector and decreases in construction. Monthly data tend to be noisy, and there's a lot of data and no clear story. But the December 2013 employment situation news release presents average monthly increases in 2013 by sector, which I would be more inclined to reference.
You can calculate these yourself if you look at Table B-6 in the release and compare Dec 2012 versus 2013 levels of employment by industry and divide by 12.
Do you think the drop in unemployment is more due to people finding jobs or people dropping out of the workforce?
That's a good question, and I think it probably depends from when you start measuring, so whether you're talking about short-run flows into and out of unemployment versus longer-run trends. The Great Recession reduced the employment-to-population ratio by about 4 percentage points, and that effect has been basically permanent so far. So in a way, one could argue that none of the changes in the unemployment rate over the last 6 years reflect more people having jobs. But it is also true that changes in the population age structure due to population aging should have been pushing the employment-to-population ratio downward during this time anyway, so this may not be the ideal way to look at it.
Do you think the expiration of the federally funded unemployment benefits is likely to be playing a role in people dropping out of the work force?
I think there certainly is potential for this channel, but I don't know. On July 1, 2013, North Carolina acted alone in ending long-term unemployment benefits. My reading of the state of knowledge is that economists have struggled to find much clear evidence of any effects stemming from this change, but they have some speculative perspectives. Part of the problem is that there is no clear control group aside from neighboring states; another is that state-level data from surveys tend to oscillate more than national data in part because of reduced sample sizes.
Are there any seasonal fluctuations that I should me aware of that might be affecting the unemployment figures at the moment?
The statistics that we look at are seasonally adjusted, which means that statistical techniques have been applied to remove the quotidian impacts of seasonality. Even so, sharp and irregular snaps in weather can still affect seasonally adjusted data. But I am unaware of anything in particular that is expect to impact the unemployment statistics. We will find out on Friday, when the release typically includes enough armchair quarterbacking and 20/20 hindsight to placate us all!