Occupational and job mobility in the US

We propose a new methodology to measure worker mobility across occupations and jobs in the US, building on the limited longitudinal dimension of monthly CPS data. For the period 1979-2006, we find that about 3.5% of male workers employed in two consecutive months report different three-digit occupations. This rate is procyclical, mildly rising in the 1980s and falling after 1995. We also revise upward current estimates of aggregate job-to-job mobility since 1994, from 2.7% to 3.2% of employment per month. Despite extreme similarity of average levels and time-series behavior, occupational and job mobility are only weakly correlated.