Caregivers' sacrifices: the opportunity costs of adult morbidity and mortality for female pensioners in rural South Africa

Abstract

This paper explores the financial and opportunity costs of adult morbidity and mortality for rural South African female pensioners in the era of HIV/AIDS. As mortality rates from HIV/AIDS and other causes escalate, older women are bearing the brunt of caregiving responsibilities for the sick and orphaned. They often use their state-funded non-contributory pensions to support kin during crises. Interviews conducted with 30 women aged 60–75 years in the Medical Research Council/University of the Witwatersrand Unit (Agincourt) study site in northeastern South Africa revealed that, to cover expenses incurred during crises, older women sometimes forgo spending money and time on their personal needs. They are thus negatively affected as individuals, while contributing positively to the household. Despite the additional household income from pensions, many of the study respondents still found it difficult to recover from the financial impact of these crises.

The authors thank the Medical Research Council/Wits Rural Public Health and Health Transition Research Unit (Agincourt) for encouraging and enabling this project. Agincourt provided the authors with access to the study site, information from the Agincourt Health and Demographic Surveillance System, assistance with sample selection, integration into the site, and collegial support throughout the data collection and analysis. The authors are indebted to the project's three interviewers, Asnath Mdaka, Florence Mnisi and Joyce Nkuna. This work was funded through the Mellon Foundation HIV/AIDS Node (E Preston-Whyte, PI), which is situated at the University of KwaZulu-Natal, South Africa; through the Mellon Foundation African Demography Research and Training Program (J Menken, PI) at the University of Colorado; and a Seed Grant from P30 AG024472-01 National Institute of Aging, Population Aging Center (J Menken, PI). The authors would like to thank Stephen Tollman, Jill Williams and Richard Rogers for comments on an earlier draft of this paper.

Notes

¹We compared this wealth index with the respondents' socio-economic status as measured by the Medical Research Council/Wits Unit in the 2003 household asset index. Although the interviewers found some correlation between the measures, there were many differences. Two things may have caused these differences. First, the AHDSS data were collected in 2003, a year prior to the interviews; thus, changes to household socio-economic status might have occurred over the course of that year. Second, the AHDSS data computed socio-economic status at a household level; although this project's interviewers also looked at the household as a whole, their wealth index focused on the well-being of the respondents. Since many households do not share resources equally among household members, it is possible that some of the discrepancies were because the respondents did not have access to household-wide assets. Thus, we are relying on the interviewers' internal scoring for the wealth index, instead of the AHDSS household asset index.

²As highlighted in Section 3, some deaths were not captured in the household mortality stratification scheme – either because they had occurred before or after the designated period of 2001–2003, or in another household – but the obligation to contribute to the expenses resulting from the illness and death extended to the older woman's household.