Date of Award

Spring 1-1-2011

Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Lori M. Hunter

Second Advisor

Jason D. Boardman

Third Advisor

Jane Menken


This project examines socioeconomic vulnerability to household shocks at the Agincourt Health and Demographic Surveillance Site (AHDSS), a research site in rural Mpumalanga Province, South Africa. The AHDSS site is managed by the Rural Health and Health Transitions Research Unit of the University of the Witwatersrand and the South Africa Medical Research Council. I focus on the shock of prime age adult mortality, a critically important shock in rural South Africa in light of the Southern African HIV/AIDS epidemic. Household socioeconomic status is an important dimension on which to assess vulnerability to shocks in rural, developing settings. Although it is generally assumed that more impoverished households are more prone to experience shocks and are therefore in greater need of assistance (Drimie and Casale 2009; Mishra et al. 2007), relatively wealthier households in these areas are often found to be vulnerable as well (Baringhausen et al. 2007; Gregson et al. 2001; Paumgarten and Shackleton 2011). Likewise, an inability to cope with shocks is often evidenced by socioeconomic declines. The following research questions are addressed: (1) To what extent is household socioeconomic status associated with the experience of future shocks? (2) To what extent is the experience of a shock associated with subsequent changes in household socioeconomic status? This project uses data from the AHDSS site annual census and a bi-yearly asset topical module to examine these relationships. Cause of death is determined via verbal autopsy (Kahn et al. 1999) allowing for examination of whether AIDS related deaths are distinct from other types of adult mortality. Household socioeconomic status is measured by (1) holdings of consumer goods, (2) livestock ownership, and (3) the extent to which the household uses electricity for tasks such as cooking and heating, in lieu of more traditional power sources such as fuelwood and paraffin. Results indicate impoverishment - in terms of owning fewer consumer goods and using traditional fuel sources in lieu of electricity - is associated with a greater likelihood of experiencing an adult death. These associations are stronger for women than for men, suggesting impoverishment is particularly dangerous for prime age women. Poverty prevention campaigns may therefore prove beneficial in arresting the HIV epidemic, in conjunction with campaigns to empower rural South African women. Households that experience the death of a head are most severely disadvantaged in the wake of a mortality shock - exhibiting a reduced growth in consumer goods, the liquidation of both large and small livestock, and a greater propensity to dissolve than their unaffected counterparts. The death of a male head is generally a worse shock than that of a female head, and AIDS related mortality appears to be a more severe shock than deaths due to other causes. This work speaks to the utility of Robert Chambers' (1989) distinction between internal and external vulnerability in developing contexts. The most internally vulnerable households at the AHDSS site are not, in fact, the most vulnerable to an adult death - even though more impoverished households of a given size and dependency ratio are more externally vulnerable than their less impoverished but comparably composed counterparts. This highlights the fact that these two types of vulnerability are not synonymous and should be treated as distinct concepts. This project also speaks to the prevalence of erosive and non-erosive coping at the AHDSS site (Donahue 1998; Drimie and Casale 2009; Seeley 1993). Households are reluctant to liquidate consumer assets in response to mortality shocks, evidenced by the overall increase in ownership of these items over the study period regardless of mortality experience. The depletion of livestock is, however, evidenced among death-impacted households - particularly those experiencing the AIDS related death of a household head. Erosive coping may endanger household viability, suggesting households experiencing this type of shock may be in particular need following the death. Evidence of non-erosive coping is found among households experiencing the death of a female head, as these households shift to less labor intensive energy sources. Overall, these findings contribute to the current debate on socioeconomic vulnerability to HIV/AIDS, providing support for the characterization of HIV as a "disease of poverty" (Fenton 2004) due not only to its propensity to target more impoverished households, but also its association with subsequent socioeconomic declines. This speaks to the urgency of developing interventions capable of assisting rural South African households coping with the HIV/AIDS epidemic.

Included in

Sociology Commons