Childlessness and education: The case of Italy and Finland

The aim of this paper is to study the role of education as a micro-level determinant of childlessness in Finland and Italy paying attention to the role of union formation. We start from the hypothesis that when modeling the relation between childlessness and education, selectivity processes have to be considered. Our hypothesis is that education can affect childlessness, both directly, and indirectly, through its link with the union formation mechanisms; and that decisions on union formation and on parenthood can be considered as jointly determined. Thus, in order to understand how education can influence childlessness, it is important take into account the possible existence of common, unobserved factors that determine both processes simultaneously. We use Bivariate Probit Models to simultaneously model both union formation and childlessness mechanisms, and to understand which is the role education play in determining such mechanisms. We focus on Italy and Finland, which are characterized by very different institutional contexts, being at different development stages, but showing outstanding level of childlessness with a prevalence over 20% for the cohorts born in the sixties. The similar prevalence observed for the most recent cohort hides interesting differences in terms of childlessness path by education level: Italy shows a persistent positive educational gradient over birth cohort, while in Finland it has reversed, turning out to be negative among the most recent birth cohorts. In a comparative perspective, we are interested in understanding if the cultural and political-institutional context of countries can determine the way childlessness and education are related. We use data from the Finnish Late Fertility Survey 2015 and from the 2009 Multipurpose Italian survey, Family and Social Subjects. Results confirm that a positive education gradient in childlessness exists in Italy, even when taking into account the correlation between the preferences of women with respect to union formation and motherhood processes; while in Finland the negative education gradient reverses when controlling for this correlation as partnership market dynamics strongly influence the educational gradient in childlessness, whose existence is mostly due the difficulty low educated women encounter in finding a partner.