According to Paul Krugman in today's (Apr. 14) NY Times: The official unemployment rate may be relatively low — but the percentage of prime-working-age Americans without jobs, which isn’t the same thing, is historically high.
According to a comment on this earlier CD post about Krugman and Don Boudreaux's response, Krugman was referring to the top chart above in the April 12 NY Times article by Floyd Norris "Many More Are Jobless Than Are Unemployed," which claims that "Men in the prime of their working lives are now less likely to have jobs than they were during all but one recession of the last 60 years. Most of them do not qualify as unemployed, but they are nonetheless without jobs."
Norris uses a "jobless rate," or "proportion of people without jobs," which can be calculated as: 1 - Male Employment Ages 25-54/Population. Using employment/population data for men, women and all workers aged 25-54 from the BLS (via Economagic), the "jobless rates" are calculated and displayed in the bottom chart above. The middle chart above shows this calculation for males aged 25-54, which matches the Norris graph at the top.
If Krugman did refer to that article as his source, there are a few problems:
1. Krugman says "the percentage of prime-working-age Americans without jobs is historically high," which is clearly not accurate. It would be more accurate to say that it is close to being historically low (see middle brown line above for "All Workers"). Krugman may have used Norris' data, but then mistakenly discussed the jobless rate for all workers aged 25-54 being high, when he should have been discussing men only aged 25-54.
2. When the data are displayed over a range from 0% to 70% (bottom graph) instead of a more narrow range from 2-16%, it's much clearer that the jobless rate for men aged 25-54 has been relatively stable at about 12% for the last 25 years. Further, the jobless rate for all workers aged 25-54 has been relatively stable at about 20% for the last 25 years, and jobless rate for women has been stable at about 28% for the last 20 years, and is close to an historical low.
Update: After exchanging emails with Norris, he calculated his "jobless rate" by first finding the number of men with jobs aged 25-54 in the household employment survey, and he then compared it to the civilian noninstitutional population for the same age group. Then he subtracted the employment/population ratio from 1 to calculate the "jobless rate." That calculation can be replicated by using the Employment/Population ratio from the BLS, and then subtracting that ratio from 1, see the middle chart above - it replicates Norris' top chart exactly.