Commuting and its effect on work decisions

We suggest a method that enables the use of cross-sectional dataset for estimation of substitution effect and income effect as well as the effect of commuting costs over hours worked. In the relevant economic literature such estimation is only possible if longitudinal data set is available (which are pretty rare). We evaluated the difference between potential and actual hours worked (hour gap) and the difference between potential and actual hourly wage (wage gap). Potential variables were estimated as a function of personal characteristics. Estimation of hour gap as a function of wage gap, income variable and time needed for getting to and from work we identified a positive substitution effect and a negative income effects on hours worked. Time needed getting to work has a negative effect on all working hours, except for agents working more than 40 hours a week and is not significant for agents working less than 10 hours a week.