An analysis of the relationship between marital status and family structure and on-the-job productivity

Abstract

That married male workers earn more than their unmarried co-workers is now well established in the labor economics literature. Traditional estimates of this marriage premium range from 10 to 40 percent. However, the source of this wage difference between married and unmarried men remains obscure. Some economists attribute this wage differential to differences in job productivity between married and single workers. Other economists attribute the wage differential to unobserved characteristics of married workers, i.e., selection effects. This thesis seeks to examine the possible causes of differences in job performance between married and single employers using data on Navy officers. The analysis shows that married male officers receive higher supervisor evaluation scores and promote at higher rates than single male officers. The results also show that there is a positive correlation between supervisor evaluations and promotion. The analysis of the effects of marital status shows that married officers achieved better performance than single officers. Unrestricted line (URL) male officers who have been married longer receive higher performance scores. For both URL and Staff / Restricted Line (STF/RL) male officers performance also increase as the number of dependents increase. OLS regression models also show that male officers who are married have attained more graduate education than single officers. The analysis of selection bias shows that single male officers who will marry in the future perform better than single officers who will remain single in the future. Fixedeffects models that control unobservable individual characteristics support the higher performance of married males. Finally, Heckman style two-step models that control for selection bias due to retention decisions show that the measured effect of marriage is biased upward in single stage models, but that the bias is not large.