VoxEU Column Gender Labour Markets

Commuting time and family labour supply decisions

Commuting time has been regarded mainly as affecting labour supply decisions at the individual level. Previous analyses do not consider the interactions between partners’ commuting times and their labour supply. This column shows that, in response to the husband’s longer commute, the wife’s employment decreases and the husband works slightly more. These results suggest that intra-family interactions need to be considered when evaluating policies that apparently affect one partner only.

Traffic congestion is a serious problem in modern cities and a source of multiple negative externalities. These span from health and environmental issues related to pollution to possible adverse consequences on labour supply, when longer commuting times or distances raise the fixed cost of working. 

So far, the economic literature has mainly focused on the effect of longer commutes on one’s own labour supply (Cogan 1981, Gutiérrez-i Puigarnau and Van Ommeren 2010, Gutiérrez-i Puigarnau and Van Ommeren 2015), neglecting the possibility that this effect may also depend on partners’ interactions. Indeed, there is increasing evidence for the key role of family interactions in explaining individual economic behaviours (Stancanelli and Van Soest 2012, Goux et al. 2014). Moreover, the existence of gender roles (Bertrand et al. 2015) suggests that the burden of labour supply adjustments is likely to be borne by wives. Black et al. (2014) show that women – in particular if married  - but not men, withdraw from the labour market when the average commuting time in their city of residence increases. This aggregate response might be the result of a direct effect of longer commutes, acting through increases in one’s own travel time, and of an indirect effect, acting via intra-household interactions through the increase in the partner’s commuting time. 

In a recent paper (Carta and De Philippis 2018), we aim to assess the relevance of this indirect effect. We estimate the extent to which individual labour supply and the allocation of time within the household respond to changes in the partner’s commuting time. More specifically, we focus on the effect of changes in husbands’ commuting time on their wives’ labour supply, because, given the higher elasticity of married women’s labour supply, this is the case that is likely to generate stronger intra-household responses. 

What the theory predicts: The role of household production

We first show theoretically that the impact of the husband’s commuting time on his wife’s employment and on partners’ time allocation crucially depends on the substitutability between domestically produced goods and those available in the market (i.e. childcare, gardening, cooking, etc.). Intuitively, when the family is indifferent in consuming either home-produced or market goods (since they are perfect substitutes), the family will always consume the ‘cheapest’ available good; this decision will not depend on commuting times. Therefore, household production will not respond to variations in travel times. If the husband’s commuting time increases, he will reduce his working hours and his wife will be more prone to work and commute. Labour specialisation within the couple will unambiguously decrease. 

However, a different mechanism could be at work when the family is not indifferent in consuming market services instead of domestic products (in other words, when the two are imperfect substitutes). In this case, since consumption always requires a combination of market and home-produced goods, the consumption decision does not depend only on relative prices, but also on partners’ overall time availability. Therefore, to offset the reduction of total available time due to the longer commute of the husband, his wife may be induced to stay at home, save her own commuting time, and increase housework, whereas the husband may work more. Labour specialisation within the couple will therefore increase.

This type of response is more likely to take place when wives earn lower wages than their husbands, and in those families not indifferent in consuming market goods instead of home-produced ones. This happens, for example, when childcare is one of the family’s main consumedgood, as the assumption of perfect substitutability between market care (provided in nurseries or kindergarten) and parental care is less likely to hold, relative to other activities such as gardening or cooking (Bernal and Keane 2011).

Empirical evidence on intra-household effects

We provide direct evidence on the relevance and the direction of intra-household interactions by considering only variations in husbands’ commuting distance that are plausibly employer-induced, i.e. those not due to variations in the husband’s job or the household’s domicile location. We focus on distance as a proxy for commuting time.

We find strong cross-partners elasticities: an increase in the husband’s commuting distance increases the specialisation of labour within the couple, by decreasing the wife’s employment probability and inducing the husband to supply slightly more hours of work. In particular, we find that if the husband’s commuting distance doubles, his wife’s probability of working decreases by about 3%. Reductions in wives’ employment are mainly observed in couples in which the wage gap between partners is larger, and in couples with young children (0-5 years old).

Policy implications

Our analysis confirms the existence of sizeable intra-household elasticities and thus implies that a comprehensive evaluation of policies targeting commuters should factor in the indirect effects on their spouses. Female labour supply is particularly sensitive to these indirect effects, as women’s employment is often the margin of adjustment used by couples to absorb the possible costs and benefits of a new programme. Hence, policies aimed at decreasing commuting times may be twice as beneficial for women: beside the existing evidence that they increase women’s labour supply directly by reducing their fixed participation costs, we show that they also increase married women’s employment indirectly, through the intra-household effect generated by changes in their husbands’ commuting times.

Authors’ note: The views expressed here are those of the authors and do not necessarily reflect those of the Bank of Italy.


Bernal, R, and MP Keane (2011), “Child care choices and children’s cognitive achievement: The case of single mothers”, Journal of Labor Economics 29(3): 459–512.

Bertrand, M, E Kamenica and J Pan (2015), “Gender identity and relative income within households”, Quarterly Journal of Economics 130: 571–614.

Black, DA, N Kolesnikova, and LJ Taylor (2014), “Why do so few women work in New York (and so many in Minneapolis)? Labor supply of married women across US cities”, Journal of Urban Economics 79: 59–71.

Carta, F, and M De Philippis (2018), “You’ve come a long way, baby. Husbands’ commuting time and family labour supply”, Regional Science and Urban Economics 69: 25–37. Previous version: Bank of Italy, Working paper 1003. 

Cogan, F (1981), “Fixed costs and labor supply”, Econometrica 49(4): 945–964.

Goux, D, E Maurin and B Petrongolo (2014), “Worktime regulations and spousal labor supply”, American Economics Review 104(1): 252–276.

Gutiérrez-i Puigarnau, E, and JN Van Ommeren (2010), “Labour supply and commuting”, Journal of Urban Economics 68(1): 82–89.

Gutiérrez-i Puigarnau, E, and JN Van Ommeren (2015), “Commuting and labour supply revisited”, Urban Studies 52(14): 2551–2563. 

Stancanelli, EGF, and A Van Soest (2012), “Retirement and home production: A regression discontinuity approach”, American Economic Review 102(3): 600–605.