Bypassing Health Centers in Tanzania: Revealed Preferences for Observable and Unobservable

Leonard, Kenneth L.; Mliga, Gilbert R.; Mariam, Damen Haile

When patients bypass on one health facility to seek health care at another, strong preference are revealed. This paper advances the view that the patterns of bypassing observed in Iringa Rural district in Tanzania show evidence of patients understanding of various measures of quality at the facilities that they visit and bypass. Importantly some of these measures are "unobservable", meaning that we do not expect patients to be able to evaluate whether or not these types of quality are present just from visiting a center. We use two data sets on various features of health facilities including consultation quality and prescription quality as evaluated by a team of clinicians. This is matched with data collected from health center registers that included the symptoms of patients and the village they traveled from. The register data is transformed into a patient-based sample and we use a multinomial/conditional logit regression on patient choice of provider to show the relationship between patient behavior and objective measure of technical quality in the health facilities. Patients seek facilities that provide high quality consultations, are staffed by more knowledgeable physicians, observe prescription practices and are polite. They avoid facilities that use injections too liberally or over-prescribe medication

Geographic Areas


More About This Work

Academic Units
Department of Economics, Columbia University
Department of Economics Discussion Papers, 0001-02
Published Here
March 7, 2011


April 2001