Many studies have found an increase in the percentage of workers working 50 or more weekly hours in the second half of the 20thcentury; however, few studies extend this analysis into the 21stcentury, and few have analyzed these patterns for women in depth. This paper provides an analysis of long work hours for men and women from 1979 to 2017. I investigate how workers who differ in education level, presence and age of children, salary type, and occupation gender-mix classification (for managerial/professional occupations), differ in their likelihood to work long hours. Using a linear probability model, I determine that those most likely to overwork include highly educated men and women, men with children, women without children, salaried workers, and workers in historically male-dominated managerial/professional occupations. Finally, using a Oaxaca decomposition, I find that changes in observable characteristics can account for between 52.28% and 72.62% of the 2 percentage point decrease in long work hours seen for men between the 2000-2002 time period and the 2015-2017 time period.

No description available.