First childbearing at higher ages in Sweden and Poland

Livia Olah, Stockholm University
Ewa Fratczak, Warsaw School of Economics

In the last decades, the mean age at first birth increased significantly in most European countries, while fertility has been (much) below replacement level. While there are theoretical explanations for the postponement of first birth, it is less clear why individuals who have not become a parent by their mid- or late twenties would do so later on in their lives. In our exploratory study, we seek to shed more light on the mechanisms of first childbearing at higher ages in the period of the mid-1970s - late 1990s. We focus on Sweden and Poland given the differences in their family formation patterns and family-level gender relations as well as the similarities regarding female labour force participation and policies that support the combination of employment and family responsibilities. In our theoretical framework we rely on the life-course approach taking into account that normative pressures for parenthood have weakened and competing demands of education, paid work, etc. as well as new opportunities render childbearing to be a choice among many other choices individuals make in their lives. We analyze data extracted from the Swedish Level of Living Survey (2000 & 1991) and the Polish Retrospective Survey of 2001, using non-parametric methods (Kaplan Mayer estimation) and hazard regression incl. individual background characteristics, variables of partnership behaviour as well as measures of current situation (education, employment, calendar period). Both the main effect models and the interaction terms show that the most important mechanisms of the late transition to parenthood work via partnerships, both former and current ones, and via age given the increase of fecundity impairment with increasing age (biological aspect) as well as the increasing importance of social and/or cultural obstacles (normative aspect). Also educational attainment and employment status are important but there are differences in the effects by country and gender.

  See extended abstract

Presented in Session 31: Gendering family dynamics network 1