Crowding out of Government Health Spending: Evidence from Bangladesh

Abstract

While public spending on health in developing countries has increasingly come from domestic sources, existing studies show that government spending has been crowded out by foreign aid in the health sector. However, current research has mostly focused on cross-country studies. Cross-country studies present results that cannot be generalized to an individual country. Some countries may not experience a displacement of government health spending when there is an inflow of aid, while others may respond to health assistance by reducing their own spending in health. It may be worthwhile to look at country-specific cases in order to investigate whether health assistance over time crowds out purely government expenditure in health. To that end, this paper studies the displacement effect of foreign aid in Bangladesh. We specifically assess if government health spending in Bangladesh has been crowded out by health aid channeled to the government during the period 1995-2008. We identify pure government health expenditure net of aid (GHES) and health aid disbursed to governments (DAHG) and evaluate if GHES is affected by changes in DAHG. We use data from National Health Accounts (NHA) of the World Health Organization (WHO) and from the Institute for Health Metrics and Evaluation (IHME). We utilize Ordinary Least Squares (OLS), Generalized Method of Moments (GMM) and the Three-Stage Least Squares method (3SLS) to measure the crowding-out effect of aid in Bangladesh. We further address the causal relationship between domestic health spending, health aid and gross domestic product in the framework of the simultaneous equations model. We find that a one percent increase in health aid displaces Bangladesh's government health spending by at least 0.15 percent and at most 0.30 percent according to our 3SLS estimates. Our OLS and GMM estimates suggest that that aid inflows into the public health sector causes government health spending to fall between 0.14 and 0.23 percent. The government appears to reduce its own spending in response to aid inflows in the public health sector. Past spending in the health sector does not determine current health spending – the government policy is that of maintaining a steady expenditure. Our findings suggest that health-related projects are highly substitutable against donor aid. It may be easier to cut spending in health and reallocate the money to other priority sectors. There is a possibility that such displacement along with volatility in donor funding may have negative effects on the allocation of resources. To reduce this crowding-out effect, there should be an alignment between donor aid policies and priority sectors that governments want to focus their spending on.