Development Assistance for Health and Experiences of Corruption in Malawi's Health Sector

Background: There is lively scholarly debate about the impact of development assistance on corruption; however, few studies have systematically examined the subject sub-nationally or in the health sector specifically. This study addresses this gap by estimating the association between development assistance for health and reported bribe payments in Malawi’s health sector.

Methods: Data on bribe payments are drawn from individual responses (N=3600) to questions asked in nationally representative Afrobarometer surveys in 2002, 2005 and 2012. District-level per capita aid disbursements for health and HIV/AIDS are measured using geo-coded data from the AidData. Multi-level logistic regression models are then fit with demographic and socioeconomic control variables.

Results: The prevalence of bribe paying for health services is low (3.89%) and declines over time (from 4.67% in 2005 to 3.49% in 2012). There is no association between health aid and reported bribe payment for health services. Payments are disproportionately concentrated among young (OR=0.41 p<0.00), poor (OR=1.04 p=0.05), and rural (OR=1.48 p<0.05) respondents. Those who have paid bribes are also more likely to have used other informal means, such as contacting local leaders, to get help for problems (OR=1.46 p<0.05). Reported payments were strongly and positively related to feelings of ethnic discrimination (OR=1.57 p<0.00).

Conclusion & Implications: The study finds no evidence that aid activity has either increased or decreased bribe payments in Malawi’s health sector. The importance of containing even low levels of bribe payments is underscored by the finding that it affects the poor disproportionately and may intensify feelings of unfairness and discrimination.