Various assessments point to extensive levels of poverty in Cambodia. The poverty headcount index in 1999 was an estimated 36 per cent (Poverty Profiles, 1999). Although comparisons between different poverty measures are difficult, the incidence of poverty appears largely unchanged from 1997. Poverty rates are highest in rural areas where roughly 90.5 per cent of the poor live. The remainder of the poor is located in other urban areas (7.2 per cent) and the capital Phnom Penh (2.3 per cent). The recent Cambodia Socio-Economic Survey (CSES 1999) largely confirmed these patterns. Average annual income in rural areas was less than one-third of Phnom Penh residents (rural US$ 197 per annum, Phnom Penh US$ 691 per annum).

Cambodia compares unfavourably with other Asian countries on broader human poverty indices. For example, using the UNDP Cambodia human poverty index (HPI), a score of 42.5% is reported. In comparison, the average score for Southeast Asia and Pacific developing countries is 25%. Poor performances on primary school completion, health and sanitation service access and child nutrition are major contributing factors to the low HPI in Cambodia. Clearly improved access to basic education services in poor rural areas is a top priority in reducing human poverty rates.

There are also large urban/rural disparities in human poverty. For example, the HPI in urban areas (16 per cent of the population) is 34.2 per cent. In contrast, the figures in rural areas are 44.9 per cent for 84 per cent of the population. Other indicators reinforce these wide disparities. Per capita consumption in urban areas is twice that of rural people. Urban residents have an additional five years life expectancy and higher levels of education attainment. Cost and access barriers appear to be a significant factor since Phnom Penh residents spend between 12 times as much on education as the rural population.

A recent survey (MoEYS 1999) shows some correlation between vulnerability (especially food security) and education participation and attainment. For example, in around 550 vulnerable communes, the literacy rate for 10 - 14- year-olds was around 63 per cent compared to a national average of 68 per cent. The proportion of people never attending school from these vulnerable communes was 50 per cent compared to a national average of 45 per cent. These figures suggest that targeted school feeding programs, especially for upper primary and lower secondary grades, could have positive benefits for enhanced school enrolment and retention of pupils from poor families.

Three key issues emerge from the above discussion. A first priority is to broaden the availability of education services, especially lower secondary provision. A second priority is to alleviate cost barriers to access to primary and secondary education. A third priority is to examine equitable resource allocation policies, possibly including some affirmative poverty indexing in Government spending on basic education services.