This thesis is a pioneering analysis of the intermarriage and spouse import of immigrants and their descendants in Norway, combining multivariate statistical techniques and mechanism-based theoretical explanations. Using register data on the complete stock of immigrants and descendants who have married in Norway in the period of 1973-2002 (N=55,293), I explore the degree to which national-origin groups differ in their spouse selection, and I model the influence of other individual-level characteristics such as educational level, generation, gender, and age at marriage (and control for place of residence and length of stay before marriage) on both intermarriage and spouse import. Regressions are run for the whole immigrant population, separately for men and women of both generations, and for a subset of national-origin groups: Moroccans, Turks, Indians, Pakistanis and Vietnamese (and for the intermarriage analyses, Danes).
First, I present contingency tables of national-origin and spouse selection outcome. Among both generations, the national-origin groups of Western Europe and North America have the highest intermarriage rates, and Middle Eastern and South Asian groups display the lowest mean intermarriage rates and the highest spouse import rates. In general, men are considerably more prone to spouse import than women. Among immigrants marrying in Norway, women tend to intermarry more than men, but among descendants, men intermarry more than women. Male descendants intermarry more than male immigrants, while female immigrants intermarry more than female descendants.
Second, I report multinomial logit models for intermarriage. The analyses show that the positive relationship between educational level and intermarriage is much lower than that indicated by models from previous research treating the immigrant population as one group, and without controlling for age at marriage. Educational level increases descendants’ probability of intermarriage more than that of immigrants (and particularly for women). Higher age at marriage increases the chances of intermarriage for both immigrants and descendants, to the extent that controlling for age at marriage makes both male and female descendants more likely to intermarry than immigrants. While education has a clear positive bearing on the intermarriage rates of individuals from Vietnam, Turkey and India, and a minimal increase for Pakistanis and Moroccans, the same relationship is negative for Danes. Similarly, while descendants from Vietnam are more prone to marry a majority Norwegian than Vietnamese immigrants, the opposite is seen in the case of Moroccans and Indians.
Third, I report binomial logit models for spouse import among the endogamous. Again, the importance of controlling for age at marriage and national-origin group variation is confirmed. Still, higher educational attainment and age at marriage both decrease the probability of spouse import. On average, descendants are more likely to import spouses than immigrants. Moroccans and Vietnamese are less likely to import spouses the higher their educational level, while Pakistanis, Turks and Indians displayed no significant differences according to educational attainment. However, while descendants from Vietnam and Morocco are more likely to import than immigrants from the same countries, those with Turkish and Pakistani background display less propensity of finding a spouse in their country of origin than women from their parental generation.