Sub-national distribution and situation of orphans : an analysis of the President"s emergency plan for AIDS relief focus countries

The analysis provides information about the communities where orphaned children reside and about their living situations. Key findings are as follows: (1) Orphan status adversely affects school enrollment in most countries, even after controlling for other confounding factors. Double orphans in Ethiopia, Kenya, Mozambique, and Tanzania were less likely to be enrolled in school than non-orphans. Similarly, maternal orphans in Ethiopia, Haiti, and Rwanda, and paternal orphans in Ethiopia, Haiti, Kenya, and Nigeria were less likely to enroll in school than non-orphans. Unexpectedly, however, paternal orphans in Namibia and maternal orphans in Mozambique and Nigeria were more likely to enroll in school than non-orphans. (2) For countries with two household surveys, the dependency burden for working-age adults tends to have increased nationally (excessively in rural areas, while remaining constant or declining in urban areas) in Haiti, Namibia, Rwanda, Tanzania, Uganda, and Zambia, despite significant fertility declines in these countries. This suggests that one of the broad problems that high HIV/AIDS-prevalence countries face in supporting their young and elderly populations is the shift of the economic and social burdens from urban to rural areas. (3) "Non-relative" child fostering, once an uncommon practice in Africa, is increasing in most African countries. The proportion of double orphans under the care of non-relatives increased in four of six African countries between the two periods of surveys. In Haiti, the proportion of orphaned children living with unrelated heads of households continues to be higher than in Africa, and increased slightly between 1994 and 2000. (3) Consistent with previous studies, there was a significant relationship between children"s school enrollment and their relationship to head of household. Controlling for other factors, in all countries children living in households headed by non-relatives had a significantly lower likelihood of being enrolled in school than those in households headed by a surviving parent. Similarly, children in households headed by relatives other than grandparents or siblings were less likely to be enrolled in school in most countries compared with those in households headed by a surviving parent. (4) As expected, the degree of the orphan problem varies within sub-regions and across countries. The percentage of orphans in the under-15 child population ranged from a low of 4.4% in Guyana to a high of about 28% in Rwanda. This study also identifies the sub-regions within each country that have higher orphan rates than national averages ("hot spots"). Most of these sub-regions were also the same ones with high HIV/AIDS prevalence rates. For countries with two surveys, the total percentage of orphans significantly increased in Kenya, Namibia, Rwanda, Tanzania, and Zambia, and similar trends were observed within the sub-regions of these countries between the surveys. The rates were stable in Haiti and Uganda. (5) Household socioeconomic status and subregion of residence are major determinants of school enrollment. Indeed, the variability in school enrollment was to a large extent explained by household socioeconomic status and sub-region of residence and only to a small extent by parental survival status. (6) Children living in female-headed households had a higher likelihood of enrolling in school than those in male-headed households. In most countries, a higher percentage of orphans live in female-headed households. (7) Everything else being equal, children living with non-working age (elderly and young) heads of households were less likely to be enrolled in school than those living with working-age heads of households in most countries. Also, children living in large families have a higher likelihood to enroll in school than those in smaller households, while rural children were less likely to be enrolled than urban children.