Effects of Electoral Success on the Family Life and Fertility of Men and Women

• In Event: Women in Politics: Understanding Female Political Participation

Abstract

Evidence from numerous countries has shown that female parliamentarians are more likely to be divorced and less likely to have children than their male colleagues. In this paper, we develop further insights into the gendered personal costs of winning a parliamentary seat. We analyze several costs related to the work-life balance of politicians, most importantly the probabilities to be become married, to get a divorce, or to have (more) children.
A central challenge of studying the consequences of a political career is to isolate the consequences from holding the job from the characteristics of men and women who seek out those jobs. If pressures are harder for women, being single or childless could be a pre-condition for candidacy rather than a consequence of holding specific positions. We address this empirical challenge by exploiting close election around marginal seats on electoral ballots. In this setup, we can observe candidates for the job as parliamentarian -- the persons who narrowly won a seat and those who narrowly lost a seat -- and observe them over time after the election. A difference in family structure between these persons, who were both candidates, allow us to assign events that occur in the years following the election to (some aspect of) the political job. Such potential mechanisms are the focus of a descriptive analysis.
We apply our method to rich administrative data from Sweden. The data covers the universe of electoral ballots since 1982-2014, including the party, list rank and personal identification code for every candidate. Identification codes are mandatory and can be used to link individuals to their relationship histories, captured by 30 years of annual administrative data. This data has no missing values, is not self-reported, and covers every marriage, divorce, or child birth through record keeping by Swedish regulatory agencies. In addition to a host of socioeconomic data, our database includes the geographical code of every politician's home, in addition to individual-level data for politicians' spouses. These variables allow us to study the results more in detail by splitting the sample in various ways, getting a better understanding of the mechanisms behind the main results.
The results show that winning a seat in parliament, comes at high personal costs for women but not for men. Electoral success reduces women's likelihood of having a spouse or cohabitant, a combination of relationship dissolutions and a reduction in the rate of finding a new intimate partner. It also reduces the likely of having (additional) children. As for the divorce effect among women politicians, we find divorces occurs more often in couples with a larger age gap and a less equal division of leave. It also occurs in households in which her promotion shifts the division of earnings (further) away from the norm of male dominance. No divorce effect is found in couples that are more gender-equal in terms of having a smaller age gap and a more equal division of parental leave. In an extended analysis, we separate the sample by the commuting distance between the politician's home town and the parliament. While personal costs do not differ by distance, the types of men and women who become candidates sheds additional light on our main results. For example, the longer the commuting distance from the parliament, the higher the proportion of single and childless women among the candidates.

Authors

• Olle Folke, Uppsala University

• Johanna Rickne, Stockholm University