Correlation of Traits in Married Couples: Assortative Matching or Just Who you Meet?

There is a positive and significant correlation of many traits, such as age, religion, socioeconomic status, and education, among spouses. Becker (1973) shows that positive assortative matching – which results in a perfect correlation of traits in spouses – is optimal if the traits enter the household production function as complements. This paper investigates a second reason for these correlations – that they arise from imperfect markets that restrict one’s pool of potential partners to those similar to oneself. Understanding what causes likes to marry likes has important policy implications. If the observed sorting is due to preferences, then policies promoting diversity are unlikely to be very effective. If, however, the primary reason for similarities among spouses is differential meeting rates, then providing incentives for less segregation becomes important -- not only to reduce contemporary social stratification, but also to mute the intergenerational transmission of inequality