Logo: to the web site of Uppsala University

This thesis consists of an introduction and four self-contained essays.

Essay I (with Susanne Dahlberg) estimates cohort size effects on earnings profiles and further assesses whether these profiles are affected by the individuals' position in the baby boom. Using a rich individual based panel data set, we follow the Swedish baby boomers of the 1940's and the following baby bust of the 1950's from 1968 to 1999. We find significant cohort effects on the earnings profile which are fairly consistent across gender but not across education levels. Large cohorts have a higher overall earnings level than small cohorts. Cohorts born in an upswing of a boom have a higher earnings level than cohorts born in a downswing. The effects on return to experience vary across education and experience levels.

Essay II explores the relationship between income inequality and growth using panel data on Swedish counties from 1960-2000. Compared to standard methods of estimating this relationship, yearly regional level data are used, and inequality is al-lowed to be endogenous. We find a significant positive impact of inequality on growth, but the magnitude of the effect decreases with the length of the growth period studied. When allowing income inequality to be endogenous, using panel 2SLS IV estimation techniques, we find a positive effect of inequality on 1 to 5-year growth rates, when significant, whereas the effect on 10-year growth rates is not clear cut.

Essay III examines labour supply responses to spousal sickness absence (SSA) using a Swedish longitudinal panel data, from 1996-2002. The overall results clearly indicate a decrease in labour supply in response to spousal sickness absence. The effect on labour supply increases with spousal earnings level. Women react stronger than men, and more often respond to current shorter term SSA, whereas men mostly react to longer term SSA.

Essay IV contributes to the literature on income and health by providing a detailed investigation of the family income/child health relationship using matched parent – child survey data from the Swedish Survey of Living Conditions (ULF). This study focuses on both physical as well as on the psychosocial health of the child. Furthermore, the effects of both parental socioeconomic background as well as the liquidity constraint problems the household faces are explored. We find little evidence of an income gradient or effect on children’s physical and psychosocial health. However, our study suggests that the occurrence of liquidity constraints in the household in-creases the likelihood of the child having a lower psychosocial health status.