South Africa provides a unique context in which to examine aging because of the confluence of three factors: (1) a steady increase in the proportion of older persons in the population;(2) a decrease in life expectancy brought on by HIV/AIDS and (3) the existence of a state-funded non-contributory pension system for persons aged 60 and over. The proportion of South Africans in the 60+ age group was 7% in 2009 and is projected to double by 2050 (UN 2009). At the same time, life expectancy has dropped from a high of 61.5 in 1991 to 51.6 in 2009, much of it attributable to HIV/AIDS (World Bank 2009;Hontalez et al. 2011). The old age pension has become an invaluable household resource, particularly in rural areas, where it is equal to nearly 50% of mean household income, and benefits not just the recipient but entire households and their networks (Ardington et al., 2010;Case &Menendez, 2007;Schatz &Ogunmefun, 2007). All three factors are likely to be reflected in the living arrangements of older persons. The extant literature suggests that elders in African countries with higher HIV-prevalence are more likely to be living without caregivers (Kautz, Bendavid, Bhattacharya, &Miller, 2010) and providing care to orphaned grandchildren is resulting in increasing numbers of skipped generation households (Amaoteng et al. 2007;Cheng &Slankam 2009). However, emerging evidence shows that pensions may make elders attractive as household members, particularly in households with unemployed adults, thus resulting in multiple generation households (Schatz, Madhavan &Williams 2011). Most analysis of living arrangements of the elderly (Bongaarts and Zimmer 2002;Merli and Palloni 2004), however, use cross-sectional data, greatly limiting our ability to understand how the living arrangements of older persons are changing over time. Moreover, we know even less about how AIDS-related mortality and access to pensions influence older persons living arrangements. In this project, we address these gaps leveraging the power of a unique longitudinal dataset in rural South Africa. We propose to examine changes in the living arrangements of older persons aged 50+ over a ten year period (2000-2010) in a rural, Black community in Northeastern South Africa, a setting that has experienced an increase in the number of deaths attributable to AIDS and where the proportion of pension eligible persons in the population has also increased over a 10-year period. Using longitudinal data from the Agincourt Health and Demographic Surveillance System which covers approximately 14,000 households in the Agincourt sub-district of Mpumalanga province, we will undertake the following analyses: 1) examine the distribution of household types in which older persons reside at three time points (2000, 2005 and 2010);2) calculate the probabilities of older persons'households transitioning from one type to another over two 5-year periods starting in 2000;and 3) analyze the roles of AIDS-related mortality and access to pension in driving changes in older persons'living arrangements.
The findings from this project will contribute to the small but growing research on the living arrangements of the elderly in Africa by focusing on 1) the change over time in living arrangements, 2) the extent of stability in household living arrangements over time and 3) the relationship between AIDS related mortality and access to pensions and living arrangements over a 10 year period (2000-2010) in rural South Africa. This time period is characterized by an increase in the deaths related to AIDS as well the number of people eligible for pension receipt (eligible age - 60). These results will inform intervention efforts and policy development for improving the health and wellbeing of older adults given that living arrangements are one important proximate determinant of physical and psycho-social health.