Epidemiological assessment of food aid in the Bosnian conflict, 1994-97.

Surveys in emergency settings are important for the accountability of food aid. Four household surveys conducted between 1994 and 1997 measured the performance of the Bosnia food aid programme, covering a random sample of clusters in Bosnia and Herzegovina and the Republica Srpska. The team calculated coverage, exclusion and inclusion errors, programme misses, and under-supply. Despite intended universal coverage from 1994-96, 15, 19, and 31 per cent, respectively, did not receive food across the three-year time frame. Households categorised as vulnerable were somewhat more likely to receive food. Programme misses were rare, whereas under-supply fell from 30 per cent in 1994 to four per cent in 1997, as the availability of other food increased. Extrapolation suggested that 61 per cent of the food distributed did not reach households. The programme introduced priority categories for targeting in 1997, yet nearly one-half of the highest priority households did not receive food. Incomplete coverage and weak targeting were related to political constraints.