Social connections are important determinants of health, education, nutrition and other dimensions of well-being—particularly in youth and old age.
The most common measure of social connection in population and public health research is co-residence, which has critical limitations. Failure to attend to the full range of social relationships limits our ability to understand the social context of health and well-being.
Non-household relationships are particularly important under conditions of extreme social or epidemiologic transition, as in the post-Apartheid South African context of high rates of unemployment, labor mobility and HIV/AIDS.
To redress this failure, we are using existing ethnographic data and longitudinal demographic data from rural South Africa to develop new measures of social connection that can be used in social surveys.
The ethnographic data are from an intensive fieldwork investigation of children's networks of support in the Children's Well Being and Social Connection project (CWSC), which we conducted from 2002 to 2004.
The quantitative data is the result of 14 rounds of a population-wide census in a sub-district population of 70,000 people conducted by the Agincourt Health and Demographic Surveillance System (AHDSS).
We are using histories of prior coresidence, union or parenthood status to reconstruct the constellation of relationships affecting an individual. We will then use these indicators to test for improved explanatory power in models of the social determinants of nutrition, schooling and residence.
This research will increase the utility of the AHDSS and other population-based data systems for understanding the social dimensions of well-being.
This project is a collaboration among Nicholas Townsend of the Brown University Population Studies and Training Center, Sangeetha Madhavan of the University of Maryland, and Mark Collinson and Paul Mee of the Agincourt Health and Population Unit.
It is funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development and further supported by the University of Colorado at Boulder African Population Studies Research and Training Program—which is supported by the William and Flora Hewlett Foundation.