Protecting the poor under cost recovery: the role of means testing

Abstract

In African health sectors, the importance of protecting the very poor has been underscored by increased reliance on user fees to help finance services. This paper presents a conceptual framework for understanding the role means testing can play in promoting equity under health care cost recovery. Means testing is placed in the broader context of targeting and contrasted with other mechanisms. Criteria for evaluating outcomes are established and used to analyze previous means testing experience in Africa. A survey of experience finds a general pattern of informal, low-accuracy, low-cost means testing in Africa. Detailed household data from a recent cost recovery experiment in Niger, West Africa, provides an unusual opportunity to observe outcomes of a characteristically informal means testing system. Findings from Niger suggest that achieving both the revenue raising and equity potential of cost recovery in sub-Saharan Africa will require finding ways to improve informal means testing processes.

PIP: This article reviews the African experience with means testing for the poor who are unable to pay user fees or would suffer financial hardship by paying fees. A framework is given for evaluating and comparing outcomes of various schemes for protecting the poor from financial hardship with and without official cost recovery. Informal means testing in Niger is used to illustrate cost recovery outcomes. Means testing is differentiated from targeting, which is the channeling of health care to a target population. Means testing is only one way of targeting a population. Means testing is specific to individuals, who are determined to be eligible or not according to income criteria. Proxy-based means testing determines eligibility based on other criteria related to income, such as housing characteristics or education. The four possible outcomes of means testing are identified as patients correctly classified and correctly given or not given benefits, and patients incorrectly classified and benefits were not given to the deserving poor or benefits were incorrectly given to the non-poor. The incidence of particular error types determines accuracy. One caveat is that a more accurate targeting mechanism may not optimize resources. There is an efficient amount of leakage to the non-poor that can be allowed. The literature review of 56 targeting experiences and the 14 means-tested health programs in Africa show that in Africa policy offers little guidance for practices. Health sector means testing occurs at the point of service. In Ethiopia waived fees were paid from community association funds. Local community associations issued free health care certificates but did not issue them freely. There was low leakage and possibly inadequate coverage among the poor. In Zimbabwe, leakage was high, because the income threshold was high and set by central concerns. Facilities in rural areas must determine eligibility since pay slips are not used. African systems generally have high leakage. In other regions means testing outcomes are influenced by environmental factors and design factors. However, constraints to means testing in Africa are not known.