Methods: Researchers analyzed data from the baseline of a community-randomized controlled trial of cash transfers for households (N=4,274) caring for OVC (N=12,194) in rural Manicaland Province, Zimbabwe. Boys and girls were evenly represented, and ages ranged from 0 to 19 (M=9.6, SD=4.7). HV is measured by a child's having a chronic illness, and/or disability. SV is measured by a child's regular school attendance, and whether the child has birth registration that allows access to community services. An index was created to measure the number of household assets. Other variables included the number of children, orphans, and people in the household, children's gender and ages, the number of adults with disabilities and/or chronic illnesses, and the mean age of adults.
Analysis: Pearson correlations were conducted for all continuous variables, and independent samples t-tests examined mean differences of continuous variables by whether an adult in the household was disabled, or chronically ill, and by child's gender. For children nested within households, two generalized mixed linear models were used to predict health vulnerability (model 1) and social vulnerability (model 2), by number of children in the household; number of orphans; assets; number of members; whether any adults are disabled; whether any adults are chronically ill; mean age of adults in the household; child's age; and child's gender.
Results: Greater assets were negatively correlated with the number of orphans (p<.001), health vulnerability (p<.05), and social vulnerability (p<.001). Children in households with disabled adults had lower health vulnerability (p<.001) but higher social vulnerability (p<.001), and these households had more assets (p<.001).Children in households with chronically ill adults had higher health vulnerability (p<.001), but these households also had more assets (p<.001). For model 1, intraclass correlation=.192 for children within households. Greater household size (p<.05) and older mean age of adults (p<.05) predicted lower health vulnerability. Chronically ill adults in the household predicted higher health vulnerability (p<.001). For model 2, intraclass correlation=.368. Number of children (p<.05) and disabled adults (p<.001) predicted higher social vulnerability, while assets (p<.001) and older age of the child (p<.001) predicted lower social vulnerability.
Conclusions and Implications: Asset ownership has emerged as an important buffer against social vulnerability. Cash transfer interventions may be improved by targeting improvements in educational and social outcomes. Children's health vulnerability is more strongly related to adults' chronic illnesses, whose households rate higher on asset ownership.