Widowhood and Race

recently bereaved, known as the “widowhood effect,” is key evidence in support of the sociological tenet that social relationships can affect the life chances of individuals (Durkheim [1897] 1997; House, Landis, and Umberson 1988). Dozens of studies across a large number of industrialized countries document that bereaved spouses suffer greater mortality than currently married persons (Hu and Goldman 1990), with evidence mounting that this association is at least partly causal (Lillard and Waite 1995; Lillard and Panis 1996). Comparatively little is known, however, about how the widowhood effect depends on attributes of the affected individuals. Previous studies have tended to focus on the simple presence versus absence of a marital tie while paying little attention to how the characteristics (or actions) of spouses may influence the health outcomes of their union. In network analytical terms, previous work has stressed the existence over the content of social ties, and has underplayed the interaction between the attributes of social actors and the social ties that connect them. What many studies lack, in other words, is attention to effect heterogeneity. Some concern for traits of individuals connected by a marital tie is, of course, reflected in the simple fact that the widowhood effect is typically assessed separately for men and women, but other dimensions of potential variation are almost universally neglected. With one inconclusive exception (Schaefer, Quesenberry, and Wi 1995), research on race differences in the widowhood effect is nonexistent. Effect heterogeneity is important for at least two reasons. First, critical attention to the way widowhood might have different effects in different population subgroups may contribute to an improved Widowhood and Race