Some analysis of this claim by my CBPP colleague Paul Van de Water and me, over at Politico.
Update: Evan Soltas gets the same finding we do with much the same data. That’s always good to see. BTW, Evan correctly notes the volatility of the Household Survey, the source for monthly data on part-time employment. Do you know how many jobs added or subtracted in a given month it takes to reach statistical significance from these data? You don’t? Really? 400,000 (i.e., monthly gains or losses that are less than this are statistically indistinguishable from no change at all. For the payroll survey, the comparable number in 100,000). That’s a big number relative to the usual monthly result–the average monthly employment change in absolute value over the past decade is about 260,000.