The long-term impact of maternal leave duration on smoking behavior

Maternal leave policies aim at protecting the health of new mothers. However, the impact of such policies on precursors of health, such as smoking behavior, is both theoretically and empirically understudied. We investigate the effect of maternal leave duration on the long-term smoking behavior of mothers across 14 European countries by combining survey data on health behaviors with retrospective information on birth and employment histories, and link these with maternity and parental leave policies between 1960 and 2010. To identify the causal impact of maternal leave duration, we exploit between and within country variation in mothers’ exposure to statutory leave duration policies in an instrumental variable framework. We find that a one-month increase in maternal leave duration increases the probability that a woman smokes in the long run by 2.3 percentage points. Similarly, a one-month increase in leave duration increases the lifetime duration of smoking by 13 months. We document non-linearity in this effect for the first time, showing that shorter leave durations have a protective effect, while very long maternal leave promotes harmful health behavior. Suggestive evidence shows lack of financial support from spouse around childbirth as a mediator of the observed effects, while employment and other socio-demographic characteristics play no role.