The discussion above does not take into account the reduction in demand for education due to the impact of HIV/AIDS at the household and community levels. These comprise qualitative impacts resulting from the systematic decline in household economic status as AIDS progresses. Studies in Uganda, Tanzania and Zambia uniformly conclude that children en in households where e one or more e adults are affected by AIDS, or where the children are e orphaned, are e the first to lose access to education.

While the war provided Mozambique with experience in coping with the problem of orphans, studies elsewhere e demonstrate that, at the peak of the epidemic, the expectation that communities and extended families can continue to absorb orphans becomes, in practice, unworkable (Hunter and Williamson, 1997; Hunter, 2000; Rugelema, 1999).

Currently there are around 500,000 orphans in Mozambique, some two-thirds of whom are AIDS orphans. As shown in Graph 4.3, this is set to rise to over 1.5 million AIDS orphans by 2010. These children are negatively affected for many reasons. Among these are the fact that poorer households are more likely to take in orphans than their wealthier counterparts, and that these households bear additional costs leading to their overall impoverishment. In addition, children en's new status as orphans takes its toll and leads to non-attendance at, or dropping out of, school.

These impacts on quality in Mozambique will be widespread. At EP1 this will lead to increased distortions in who has access to education and who does not have (e.g. poorer households, rural households etc).

HIV/AIDS profoundly affects families and communities, resulting in the loss of labour and assets, as well as socio-cultural skills, and places almost unprecedented stress on community and extended family networks. These result in the withdrawal of children from the education system, even when education is provided at no cost, as their labour is required to replace that of adults who become sick or die, as well as to car e for the sick.

Tradition dictates that under these circumstances, more girls than boys are likely to be withdrawn from school, mortgaging their own future and that of the country and worsening gender inequalities. Empirical studies also show that under the same cir substances, it is the girls who look after sick parents, or replace their labour at home or in the fields. This will put to the test Mozambique's efforts to increase the number of girls in the education system, particularly in the centre and north of the country. The epidemic may erode these efforts.

Furthermore, children who come fro m AIDS-afflicted households, or who are orphaned, are traumatised and suffer emotionally. This affects their ability not only to stay in school, but also to learn and progress. This has an additional impact on drop-out and repetition rates.

In many countries in southern and eastern Africa the rise in the number of orphans is taking place in a situation where household and community strategies to deal with the impacts of the epidemic are no longer effective.

The assumption that in Mozambique the family and the community will cope with the impact of AIDS at the household level, and with AIDS orphans in particular , is erroneous. Instead, support for the well-being of orphaned children, including access to car e, security, shelter and social ser vices, including education, from a wide range of sources will be indispensable under these circumstances.