Marriage, Money and Migration

The thesis consists of a summary and four self-contained papers. Paper [I] examines the effects of interregional migration on gross earnings in married and cohabiting couples. In particular, we examine the link between education level and income gains. We find that pre-migration education level is a key determinant of migration and economic outcomes and is also a determinant of the effect of migration on income distribution within the household. The positive average effect on household earnings is largely explained by income gains among highly-educated males. Females generally experience no significant income gain from migration in absolute terms. Paper [II] analyzes the effect of the spouse’s education on individual earnings. In this study, we control for time-invariant heterogeneity that may be correlated with the spouse’s education level and use a rich data set that includes observations of individuals when they are single and when they are married. The results support the hypothesis of cross-productivity for both males and females. Furthermore, couples with education within the same field experience even larger effects. In Paper [III] we aim to study how the spouse’s productivity in the labor market affects one’s own individual earnings when married. Using longitudinal data on individuals as both single and married allows us to estimate the spouses’ productivity as single persons and thereby avoid problems of endogeneity between the two spouses’ labor market performances. Productivity is approximated with residuals from estimates of pre-marriage earnings equations. Results indicate that there are negative effects of the spouse’s productivity on individual earnings for both males and females, and that this effect appears to be enhanced by the duration of the marriage. Paper [IV] studies spousal matching on earnings for females in secondorder marriages. We aim to follow women who marry, divorce, and subsequently remarry compared with females who marry and stay married over the course of the study interval. Overall, we find significant positive correlations for all three of the marital partitions. The correlation tends to be smaller for the first of a sequence of marriages for women who divorce than for women who marry and stay so. For the second of the successive marriages, however, the correlation of the residuals is larger than that for women who marry but once.