On-the-job Leisure and Work from Home: Measuring the Productivity of Work; Peter Rupert (UCSB)

Abstract

We document a considerable rise in hours worked at home and a small decline in hours not working at work brought about by the 2008 recession. In 2019, workers spent on average 4.5 hours per week working from home and 2.15 hours not working at work. We show that the increases in working from home cannot be accounted for by changes between occupations, but rather by increased computer use within occupations. We also document a substantial increase in the productivity of working from home relative to at the workplace. In 2003, an hour worked at home was about 2% less productive than an hour at the workplace, but in 2019 an hour at home was 12% more productive. The increase in relative productivity can be almost entirely accounted for by less work from home occurring while also providing childcare and more work from home occurring during standard business hours rather than in the early morning or late evening. Finally, 10% of the increase in labor productivity since 2009 can be attributed to the substitution of working at the office to working from home.

Click here to view the paper.