20/5/15: Effects of commuting times on couples’ labour supply

"You’ve come a long way, baby. Effects of commuting times on couples’ labour supply" by Francesca Carta and Marta De Philippis, (Banca d'Italia, number 1003 - March 2015: http://ssrn.com/abstract=2600874) [comments within quotes are mine]"…explores the effects of husbands' commuting time on their wives' labour market participation and on family time allocation. We develop a unitary family model of labour supply, which includes commuting times and household production. In a pure leisure model [model where there is a binary choice: work or leisure; as opposed to work, leisure or work at home] longer commuting time for husbands increases their wives' labour market participation and reduces their own working hours. However, a model that includes household production might determine the exact opposite result."

So far so good for the theory: the paper shows that when the non-principal earner is faced with a choice of either enjoying leisure only or working only, absent household production, the second earner will opt, on average, for work and the principal earner will take less effort in their own work. However, once household production is an option or a requirement (as in the case of, say, a family with children or other dependents), then the secondary earner will more likely opt for staying out of the workforce and devoting their effort to increased household production, while the primary earner will apply more effort in their own work.

That's in theory. But empirical application is a bit less straight forward:  "We then examine the sign of these effects by using data from the German Socio-Economic Panel from 1997 to 2010. Employer-induced changes in home to work distances allow us to deal with endogeneity of commuting times. We find that a 1% increase in a husband's commuting distance reduces his wife's probability of participating in the labour force by 1.7 percentage points, 2% over the mean. Moreover, it increases his working hours by 0.2 hours per week. The average effect masks substantial heterogeneity: lower participation rates are concentrated in couples with children and where the husband has higher levels of education."