Employment, uncertainty and fertility in couples: modelling longitudinal clustered data

In recent decades, total fertility has declined in almost all European countries and has reached sub-replacement levels. From the 1990s, and in particular in the aftermath of the Great Recession, economic uncertainty escalated as a fundamental factor in explaining the postponement of family formation and the fertility decline in Europe. The primary goal of this PhD dissertation is to investigate the link between economic uncertainty (using employment uncertainty as proxy), and the transition to the first and to the second child across European countries. An ancillary goal of this research is to scrutinize the effects of the Great Recession on fertility. We use the longitudinal section of EU-SILC (European Statistics on Income and Living Condition), across the years 2004-2015. The survey offers information about individuals and households by collecting a wealth of socio-economic and demographic characteristics. Given that couples are followed for several years and they are nested into countries, we adopt a multilevel discrete-time hazard model of birth conceptions with a random intercept and a random slope on calendar time. For these models, country-specific patterns are obtained via Empirical Bayes predictions. Moreover, according to several goodness-of-fit indicators, the models have a satisfactory fit. To analyse the effect of the Great Recession on fertility we rely on a structural break approach. Our results suggest, in general, that for both partners alternative job typologies than permanent employment depress fertility, especially for women; so, permanent occupations are associated with higher fertility. Furthermore, from the comparison between baseline couple with those in which one of the partners is not active emerges that, for Nordic countries alternative job typologies than permanent employment of either partners lead to a decrease in the hazard of the conception of the first child, especially if the man is not active. Conversely, in Southern countries, if the woman is not active it emerges a positive association with the hazard of the conception of the first child; while if the man is not active the association remains negative. Instead, in the case of the conception of the second child emerges that in Northern and Western countries the couple where the woman is not active presents an increase in the hazard of the conception of the second child, while in Western countries if the man is not active the hazard decrease. Finally, the structural break analysis suggest that the Great Recession had a relevant negative impact on conception of first and the second child in almost all the European countries.