In Sickness and in Health: Marital Status and Adult Health
Posted: 19 Jun 2007
Date Written: June 2007
There is a well-established connection between marriage and various measures of health. Rising rates of obesity among both adults and children in the U.S. have prompted a close look at the determinants of obesity. At the same time, marriage rates have fallen, particularly among low-income adults. In this paper, we aim to identify systematic and predictable changes in the Body Mass Index (BMI) that are associated with changes in union status. We examine five different union possibilities: Never and not Cohabiting, Never Married and Cohabiting, Married, Divorced and Not Cohabiting, and Divorced Cohabiting. Our work is prompted by policy objectives aimed encouraging marriage. Using longitudinal U.S. data and fixed effects models to net out time invariant individual heterogeneity, we examine three hypotheses: the Marriage Selection Hypothesis which states that healthy people are selected into marriage and unhealthy selected out, the Marriage Protection Hypothesis which states that married people are healthier because their spouse will take care of them when ill and discourage risky behavior and the Marriage Stability Hypothesis which states that married people are less healthy as they are no longer in the marriage market and reduce their investment in physical fitness or that married people have additional obligations that may keep them from investing in their health. Our results indicate BMI increases for both men and women during marriage. This finding is consistent with the stability hypothesis and may be a result of family responsibility over personal health or of personal complacency. For each year a woman is married there is significant weight loss, which may be an indication of the marriage protection hypothesis. We find similarities between marriage and cohabitation in that weight gain also occurs in cohabiting relationship for women but it increases with the duration of the relationship. Union dissolution has the opposite effect. Our results suggest a significant weight loss for men. This loss is largest in the year of divorce but the weight is regained over time. We also estimated our models separately by education status and by race, and find some significant differences.
Keywords: Marriage, health
JEL Classification: I12,J12
Suggested Citation: Suggested Citation