Core Unemployment

All through my working years, I kept hearing it repeated that a certain level of unemployment was unavoidable in any free economy. It was loosely referred to as “core unemployment.” It was said to be somewhere between 4 and 4.5 percent. The technical name of this kind of unemployment is “frictional” unemployment—meaning that it takes people who lose their jobs to find new ones—and for employers who lose their employees involuntarily to replace them.

I found the following useful definition of that adjective in a Bureau of Labor Statistics paper here, attributed in a quote to Frank C. Pierson of the W.E. Upjohn Institute for Employment Research, 1980:

Unemployment can be said to be structural in nature if aggregate demand is high enough to provide jobs at prevailing wages for everyone seeking work but job openings remain unfilled because of a persistent mismatching of skills or geographical locations. If the mismatching is resolved voluntarily through mutual search by workers and employers in a reasonably short period of time, say 8 or 10 weeks, the resulting unemployment falls in the frictional category.

Frictional unemployment is not reported by BLS as such, but the agency does provide unemployment by duration. With that in mind I plotted unemployment for the period 2001 through early 2011 by categories of duration: less than 5, 5 up to 15, and 15 weeks or longer. Here is the result:

As always in these contexts, unemployment is nonfarm unemployment. (The reason why farm employment gets a pass is itself an interesting subject I might take up one of these days.) Data are from this BLS facility. The graphic presents an interesting picture. Short-duration unemployment is essentially flat, and asking for a trend line for this period, I found it absolutely flat. It is affected by recessions, shown in shaded areas, but not a great deal. Unemployment at longer durations, particularly the gut-grinding 15-weeks and over, are a much better indicator.

Frictional or core unemployment—thus the more or less unavoidable kind, is minimally described by the black line. In this 123-month period, it produced an average unemployment for the period of 1.9 percent. The 5-to15 category, averaged 2.6 percent, the two together 4.4 percent. Perhaps there is something to the common wisdom that “core unemployment” runs about 4.5 percent. In this same period, the longest duration (15-weeks or greater) averaged 1.7 percent unemployment. And that category, signals genuine problems: a shrinking of available jobs.

I find it amusing that Frank C. Pierson (above) described structural unemployment as employers unable to find employees in 1980—when, in the year before, unemployment averaged 5.9 percent—whereas I see structural unemployment in 2011 as employees seeking jobs that simply aren’t there—when, in the year before, I see unemployment averaging 9.6 percent.