The Effect of Government Health Expenditure on the Income Distribution: A Comparison of Valuation Methods in Ghana

Government spending on services affects the level and distribution of welfare, but measuring its value is a challenge. To assess how publicly funded in-kind health care affects the income distribution, we must estimate its monetary value to beneficiaries. We describe and compare three approaches to measuring the distributional consequences of government health spending: average cost of provision, willingness-to-pay, and health outcomes. In addition, we estimate the value of financial risk protection from insurance, which is a benefit of health spending that can be added to each of the aforementioned approaches. Average cost is the standard method used in benefit-incidence studies (Lustig, 2018). This method values utilization of each unit of care at the government’s average cost of provision, calculated with national accounts data and administrative records. Willingness to pay uses revealed preference to estimate compensating variations for health care subsidies. The health outcomes method estimates the effect of government health spending on mortality and values those mortality reductions in monetary terms. We provide example applications for each of these methods using a national cross-section from Ghana for 2012/13. We estimate a willingness to pay model for outpatient services and find that, on average, users value those services at less than what the government pays for them. The estimated marginal effect of health spending for outpatient care on inequality are modest and somewhat smaller than those for the average cost approach. In contrast, the health outcomes method finds that the marginal effects of health spending for three causes of death and five health interventions are very large. Health interventions to reduce malaria mortality such as indoor residual spraying and distribution of insecticide-treated bed nets are strongly progressive and the averted mortality from providing anti-malarial medication dwarfs the distributional effects of any other public expenditure or tax in Ghana. Adopting the health outcomes approach dramatically changes our assessment of how public spending in Ghana affects the welfare distribution. The benefit of financial risk protection from Ghana’s National Health Insurance Scheme equals 0.25% to 0.5% of income for the three poorest quartiles and between 0.5% and 1% of income for the wealthiest, yet insurance is still distributed somewhat more equally than income itself.