Children of unknown fathers: Prevalence and outcomes in Sweden

Anders Björklund1, Karin Hederos Eriksson2, Marianne Sundström 1
1Stockholm University, Sweden, 2Stockholm School of Ecnomics, Sweden

This paper examines the background and outcomes of children of unknown fathers, a group of whom very little is known and for whom intergenerational mobility can not be analyzed. We used a unique data set based on the Swedish Multigenerational Register combined with information from the Censuses as well as educational and income registers at Statistics Sweden comprising a sample of almost 427,000 individuals born in 1955-67 to Sweden-born parents. We measure outcomes as educational attainment, annual earnings and longevity. We find that these children, on average, have very young mothers, that a large fraction of them are given away for adoption and that boys are overrepresented among them. Our results show that they had higher mortality, that is, a higher risk of dying before the age of 45, which is quite an extreme outcome. Also, they were overrepresented among those who had missing values on years of schooling and annual earnings. Among those who were alive in 1999 we do however not find any educational disadvantages among those with unknown fathers, but we do find that they have lower earnings. Taken together, these findings suggest that the fact that we cannot analyze intergenerational mobility in education or earnings for this group bias the estimates upwards. In addition, as it is likely that many of these children would not have been born if abortion had been legal, the results are consistent with the hypothesis of Donohue & Levitt (2001) that legalized abortion reduces crime.

View full paper