Abstract
National health accounts combined with data from a nationally representative household survey of health care utilization and expenditures are used to compile a description of the distribution of all health expenditures across a national population. This approach extends the earlier tradition of benefit-incidence studies by using NHA data to constrain the levels of public and private expenditures to allow a true direct comparison of the distributions of each. It overcomes earlier problems in comparing public and private expenditures on health, when information is derived from non-comparable data sources, and also deepens the analysis by using additional secondary
data to account for differences in spending levels within the public sector by geographical area and facility type and level. This is a refinement on the approach used in the COMAC-HSR studies of European and US health care spending.
The methodology requires a comprehensive, nationally representative data set on actual utilization and expenditures. The utilization data are used to allocate public subsidy expenditures for non-market services, while the household expenditure data are used to distribute private expenditures, the total of which is derived directly from the more reliable NHA estimate. The data can be combined with NHA and other public budgetary information in a data set that allocates all known health expenditures across the whole population using appropriate assumptions. The resulting data set, containing 50,000, observations, is then used to directly quantify the distribution of any component of health spending according to the household demographic or socioeconomic characteristics measured in the original survey.
The results indicate that the incidence of overall health expenditures rise with increasing income level. This is due to a distribution of private expenditures which increases with income, as well as a distribution of public expenditures which more modestly also increases with rising income level. The bias in public expenditures in favor of higher income groups is found to be primarily due to the distribution of revenue financing through the social insurance program, as well as through revenue spending at higher quality urban academic hospitals in the public sector, which largely serve higher income urban households owing to reasons of better access through
proximity, as well as the charging of user fees which discourages use by poorer households. In conclusion, the social insurance programs in Egypt and the use of cost recovery in some public sector institutions combine to greater inequality in access to health care resources, both when evaluated from the level of income levels, as well as by gender.
The 1994-95 expansion of social health insurance coverage to children has not improved the distribution of health care spending in favor of lower income households. It is argued that if policy-makers wish to improve the targeting of public expenditures in the health care sector, they should refocus efforts on the main public sector delivery system run by MOHP, and be cautious about expanding financing and delivery through other means. MOHP services while relatively equally distributed across income levels, do contribute significantly to the welfare of lower income households, adding more than 10% to the net income of the poorest quintile, compared
with only 2% for the richest household quintile.