Time-diary data from four countries suggest that differences in market time between the unemployed and employed represent additional leisure, not increased household production. In areas where unemployment is cyclically high, however, reduced market work is offset by additional home production.

We link ATUS 2003-04 to March CPS data and examine market work's effects on non-market activity. In a structural model fixed time costs of work generate a utility-equivalent of 8% of income to be overcome before working becomes optimizing.