An article in today's New York Times discusses trends in unemployment in New York City. The article begins by stating a pretty unassailable point, that unemployment in the city largely started rising after it had risen elsewhere. Then it discusses the spread of unemployment up the educational spectrum.
The article cites some statistics quoted by Lawrence Mishel, the president of the Economic Policy Institute, that suggest the burden of rising unemployment is higher among the college educated. His numbers are indeed correct: the number of unemployed who are college educated has risen 75% since March 2007, a faster clip than the 62% among all workers 25 and older.
But I think the more appropriate statistics to examine are these groups' unemployment rates. Among the college educated during this period, unemployment has risen from 1.8% to 3.1%, a rise of only 1.3 percentage points. Among all workers 25 and older, unemployment rose form 3.5% to 5.5%, up 2 percentage points. The unemployment rate is the unconditional probability an individual worker will be unemployed, and a smaller rise in that probability suggests greater insulation against recession. The percentage increase in the number of unemployed is a statistic that cannot account for the very different equilibrium frequencies of unemployment among educational groups, which makes it less meaningful.