The Census Bureau reports that the 1997 official U.S. poverty rate for married women who were 65 years and older was 4.2 percent, whereas the poverty rate for widows was 17.9 percent. Such disparities have been the impetus of much economic research. At the same time, there is much dissatisfaction with the official U.S. poverty because there are many reasons to believe it mis-identifies households that suffer low consumption. For example, a National Academy of Sciences panel suggests that the official poverty measure does not rely on an appropriate adjustment for household size. This problem is particularly relevant to understanding widow poverty because, if the adjustment for household size were not correct, the poverty rate for widowed and married women would not be directly comparable In addition, various authors have suggested that part of the relatively high poverty rate among widows is due to preexisting poverty. To improve our understanding of the well-being of widows, we propose a series of analyses that attempts to overcome many of the concerns about poverty measurement. First, we propose to examine consumption directly because the ultimate concern is that increased poverty might imply reduced consumption. Second, we propose to examine consumption and changes in consumption around the time of widowhood to better understand the direct impact of widowhood. We will rely on the various data sets that are collected as part of the Health and Retirement Study (HRS) and the Asset and Health Dynamics (AHEAD) surveys for this research.
