Unconditional Demand for Health Care in Cote D'Ivoire · Does Selection on Health Status Matter?

Health care demand price elasticities are often estimated from samples conditioned to include only sick people. This paper shows that not only may such estimates be statistically biased, but even when properly estimated they can only be interpreted as short-run price responses. In contrast, unconditional estimates take into account the long-run feedback of prices on health. The paper discusses simple strategies for estimating the long-run unconditional elasticities, which do not depend on controversial exclusion or functional form restrictions. Using the Cote d'Ivoire Living Standards Survey, a test based on the multinomial probit finds that the usual conditional estimates do not suffer from selection bias. However, short-run conditional estimates differ significantly from long-run unconditional ones for several covariates, with conditional price elasticities being about 25 percent larger than unconditional ones.