Unobserved Variables and the Elusive Added Worker Effect

Consistent with recent empirical work, this study finds no evidence of the "added worker effect" among married couples in the U.S. It is hypothesized that unobserved variables may be obscuring this added worker behavior. This hypothesis is rejected. The labor supply behavior of married women is influenced by the permanent, and not the transitory, nature of their spouses' unemployment. Wives with frequently unemployed husbands have lower reservation wages. Yet, apart from their measured human capital, these same women face persistently lower wage rates and higher unemployment propensities in the labor market. As a result, married women with unemployed spouses are no more likely to be in the labor force, and are actually less likely to be employed than observationally equivalent married women with employed spouses.