Health care demand in rural Mozambique

evidence from the 1996/97 household survey

Magnus Lindelow
fcnd discussion paper brief

Despite rapid economic growth in recent years, Mozambique remains a very poor country. Expenditure-based poverty measures are reflected in widespread food insecurity and poor health status. In recognition of these problems, the Government of Mozambique is promoting expanded and improved quality and equity in access to health care as an important component in the global strategy to fight poverty. Given years of colonial neglect and systematic destruction of health facilities during the civil war, recent government policy has focused on expanding the rural health network. However, insofar as the ultimate objective of the provision of curative services is to ensure that those in need of care receive effective treatment, it is also necessary to think beyond supply. Specifically, we need to consider how individuals behave during episodes of illness, and what factors affect this behavior. This paper provides quantitative evidence on the importance of individual, household, and community characteristics on individuals’ care-seeking decisions during episodes of illness. The paper estimates a “flexible” multinomial model of health care provider choice conditional on illness using data from the 1996/97 Mozambique National Household Survey on Living Conditions (IAF). The empirical analysis is underpinned by a basic theoretical framework of utility maximization and household production of health. A number of individual and household characteristics, e.g., age, education, and reported symptoms, stand out as highly significant determinants of health seeking behavior. Also, prices, defined in the model as the composite of user fees and time costs associated with consultations at different providers, are found to be important determinants of choice. The results indicate that the eradication of poverty, independent of improvements in physical access to health care and education, will have only a negligible effect on health care choices.