This thesis studies cohabitation behavior in the United States, and proposes two answers as to why cohabitation rate increased in the last four decades. In Chapter 1, I propose that the increasing value of economies of scale can contribute to this increase in cohabitation. In Chapter 2, I propose that declining migration could have been a factor. In Chapter 1, I use the National Longitudinal Survey of Youth 1997 to document that even from before the start of cohabitation, cohabitors who eventually marry their partners and cohabitors who do not have different labor supply and wages. I find that those who cohabit and eventually marry work on average six hours more per week than those who cohabit but will not marry. I estimate that every hour worked per week in the year prior to the start of cohabitation increases the probability of transition into marriage by .7\%. Additionally, I find that lower wages prior to the start of cohabitation are associated with a lower probability of transitioning into marriage. I develop a theory of co-residential relationship formation where lower wages increase the value of living together, leading to less selectivity in match quality and thus a lower transition probability into marriage. Singles who have higher hours worked have higher disutility to begin a relationship along with higher income, leading to higher selectivity in match quality and thus a higher probability of transitioning into marriage. I show that lower real wages in the last five decades can explain the trends in cohabitation and marriage for college and non-college graduates. Finally, I suggest that changes to the incentives to form a co-residential relationship create an additional channel where urban poverty programs can affect welfare. In Chapter 2, I compare the migration and cohabitation behaviors between the National Longitudinal Survey of Youth 1979 and 1997 cohorts. I document that for both cohorts, college graduates migrate more and cohabit less than non-college graduates. I hypothesize that having a higher probability of a future migration decreases the likelihood of cohabiting, as cohabitation can be seen as a form of settling down. I find that college graduates are more likely to move out of county or state. I also find that having a big move in the future decreases the odds of cohabiting. This can explain the cross-sectional difference between non-college and college graduates cohabitation rate. Comparing across cohorts and using the Census Bureau Current Population Survey, I find that cohabitation has increased and migration has decreased. This implies declining migration could also have contributed to the increase in cohabitation.