Health care financing in Kenya: An empirical analysis

Abstract

This study was conducted against the background of concern regarding the welfare impact of introduction of user fees in government health facilities in Kenya. The introduction in 1989 of new and revised user fees in public health facilities represented a major policy change from a policy of "free" health services for all at the time of Independence. This change, together with concerns about the welfare loss it would likely engender, led in a large part to a lack of its acceptance by the public as well as a lack of political will to implement it. The Thesis uses panel data to estimate a demand model for health care in Kenya and to assess the impact on welfare of the health care financing reform just noted. The data used covers a lG-year period - 3 years before (1986-88) and 7 years after the cost sharing reform (1989-95). Thus the data used for the analysis captures changes of the demand factors both over time as well as over the cross-section of health facilities included in the sample, which makes it possible to estimate" parameters of the Kenyan medical care demand function more precisely than has been done in previous studies. Further, the data permit a comparison of the rate of health service utilization in public and private health facilities before and after cost sharing, which in turn permits a direct inspection of the effect of the cost sharing reform on medical care use The sample statistics show that service utilization was lower in both government and in private health facilities after the implementation of cost sharing. Demand for health services in government health facilities fell by around 33% compared to a drop of 17% in non government facilities. The drop in medical care use was steepest in government hospitals (44%) where the fee increase was the largest. In government health centres, where fees were moderate, demand fell by 12%, while unexpectedly, in government dispensaries, where service provision continued to be free of charge, demand fell by 69%. These results suggest that interventions in one sub-sector of a medical care system can have quite unexpected effects on other parts of the system. The regression results show that in overall terms, the net effect of cost sharing on health service utilization was negative, a finding that is consistent with what the sample statistics reveal. Implementation of cost sharing led to an increase in medical care demand (due perhaps to improvement in service quality), but this increase was insufficient to offset a relatively large, negative demand effect of fees (the usual price substitution effect). Distance between competing health facilities, emerges in this study, as a major determinant of demand for health services: demand for medical care is greater in facilities that are clustered together than in those that are far apart. This is perhaps because proximity among facilities enhances quality via competition and/or via the sharing of medical expertise. Proximity also increases households' knowledge about the available health services. These findings are important in a variety of ways. For policy, they imply that caution should be taken when introducing user fees since they can lead to significant reductions in utilization of health services, particularly among the poor. They also suggest that in settings in which densities of alternative health care providers are high, such as in urban areas, the negative demand effects of user fees can be greatly compensated for by the positive effects of quality improvement that appears to be associated with the proximity among providers. The results further show that proper management of user fee revenues and attention to quality improvements have the effect of increasing the demand for health services. Moreover, changes in fees and in other determinants of demand in one region or district can have large effects on service quality and service use in a contiguous area. An increase in the number of doctors in Kiambu for instance (an area contiguous to Nairobi) reduces the number of visits to health facilities in Nairobi. Thus demand effects of regional spillovers of user fee changes and of other policy changes need to be considered for effective implementation of health care reforms.