This dissertation consists of three studies that evaluate how women form partnerships, leave partnerships, and the economic outcomes of those partnerships. These demographic transitions and outcomes are evaluated in three country contexts with differing political, welfare regimes, social history. I use longitudinal data from Russia to analyze marital status differences and trends in in poverty risk. Contrary to assumptions that unmarried mothers will have higher risks of poverty over time as welfare policy weakens, unmarried mothers and married mothers’ risks of poverty came close to converging in the late 2000s. Second, I use German data to examine educational assortative mating in East and West Germany. I use the Revealed Preference Model (RPM). First, from bivariate analysis of the SOEP, I find that among the people who are partnering, they are doing so mostly homogamously in the East and the West. Highly educated women in the East are still less likely to partner somebody of a lower education status. The RPM estimated parameters then showed that in West Germany and East Germany alike, educationally hypergamous partnerships were most preferable. Though the availability of higher educated partners in East and West Germany are different, the preference for hypergamy remains. Finally, I move on to the United States to estimate the divorce risk of partners of various education levels. I use the Survey of Income and Program Participation, providing accurate representation of the contemporary U.S. The model estimates divorce risk using women’s own education, men’s own education, and their relative education levels. It reveals several persistent patterns. Women’s divorce risk decreases monotonically as education increases, so highly educated women have the lowest rate of divorce. Men’s education, however, is less of a determinant on the risk of divorce. Relative to hypergamy and homogamy, hypogamous unions (woman marrying a man of a lower education status than herself) were more likely to divorce. This study supports past research that finds the female breadwinner model the most volatile when it comes to likelihood of divorce and continued support for this trend into the 2010’s.