“In Sickness and in Health”

Using data from the 1994-95 (Canadian) National Population Health Survey (6,494 women, 5,368 men), we investigated the impact of cohabitation on a range of physical and mental health indicators, controlling for self-selection into cohabitation and other relevant factors. Uncontrolled results indicate that the physical and mental health of cohabitors tends to fall between that of the married and the divorced/separated, widowed, and single/never married. However, when other factors are controlled, health differences between cohabitors and the currently married become nonsignificant. Self-selection, into cohabitation and into marriage, initially appears to play a significant role in accounting for variations in health, but with controls added to the models, selection mostly becomes nonsignificant. We concluded that self-selection at most may explain a small proportion of the variation in health but that protection effects are more likely to explain the positive health advantages of marriage and cohabitation.