Differential Diagnosis on the Labor Market Situation

Is the economy suffering more from an aggregate demand shortfall or more from “structural” issues beyond the capacity of demand-management to solve. I love House and though the show is a poor model for a national health care system, I do think that it offers a useful way of thinking about puzzles. The key conceit of Dr. House’s differential diagnosis method is that you’re looking for a theory that explains all the symptoms. Keep that in mind as you consider this stretch from the just-released January Federal Reserve Open Market Committee minutes:

Participants agreed that recent indicators showed some further gradual improvement in overall labor market conditions: Payroll employment had increased somewhat more rapidly in recent months, new claims for unemployment insurance had trended lower, and the unemployment rate had declined. Some business contacts indicated that they planned to do more hiring this year than last. However, unemployment—including longer-term unemployment—remained elevated, and the numbers of discouraged workers and people working part time because they could not find full-time work were also still quite high. Participants expressed a range of views on the current extent of slack in the labor market. Very high long-duration unemployment might indicate a mismatch between unemployed workers’ skills and employers’ needs, suggesting that a substantial part of the increase in unemployment since the beginning of the recession reflected factors other than a shortfall in aggregate demand. In contrast, the quite modest increases in labor compensation of late, and the large number of workers reporting that they are working part time because their employers have cut their hours, suggested that underutilization of labor was still substantial.

It is profoundly true that the high volume of long-term unemployment is consistent with nondemand stories about the elevation of the unemployment rate. On the other hand, it’s also consistent with the story that demand has been too low for a long time. By contrast, the slow (indeed negative) growth in unit labor costs and large quantity of “part time for economic reasons” workers is consistent only with the demand story. That’s not proof that the demand story is correct. And, indeed, if you made me Policy Dictator and it turned out that I couldn’t get unemployment below 7 percent with demand-side solutions I wouldn’t find that shocking. There are lots of interesting nondemand stories out there some of which may be true. On the other hand, if demand-side solutions alone got unemployment down to 4 or 5 percent I wouldn’t find that shocking either. The point is that the demand story explains the data and until we implement the demand solution we can’t even get any data that would distinguish between demand and non-demand stories.

Something to keep in mind about the quantity and plausibility of structural stories is that structural stuff is always happening in a large dynamic market economy. Right now the smartphone industry is showing explosive growth notwithstanding the general slump. The last boom, by contrast, coincided with the collapse of the traditional film and photocopier industries. This kind of thing matters for the macro picture, but the mere fact that some structural shifts are occuring doesn’t tell us anything in particular. You need to see the worker shortages and rising unit labor costs (a metric, I note, that lends some support to James Bullard’s thesis that 2006 represented an unsustainable level of output) to get a nondemand diagnosis.