Improving estimates of job matching efficiency with different measures of unemployment

Traditional measures of unemployment can mask important changes in the labor market across time. We therefore use broader definitions of unemployment to estimate time-varying job-matching efficiency rates that are consistent with vacancies and hiring activity data for the U.S. Our efficiency rates are then modelled along with employment data to study their dynamic, non-linear relationship. We find that including marginally attached workers and part-time workers for economic reasons helps explain the changes in employment patterns observed after the global financial crisis. This finding emphasizes the importance of accounting for labor underutilization, particularly during the latest economic recovery.