Worms: Identifying impacts on education and health in the presence of treatment externalities

Abstract

Intestinalhelminths-including hookworm,roundworm,whipworm,and schistoso- miasis-infect more than one-quarterof the world’spopulation.Studies in which med- ical treatment is randomizedat the individuallevel potentially doubly underestimate the benefitsof treatment,missingexternalitybenefitsto the comparisongroupfrom re- duced disease transmission,and therefore also underestimatingbenefits for the treat- ment group.We evaluate a Kenyanprojectin which school-basedmass treatmentwith dewormingdrugswas randomlyphased into schools, ratherthan to individuals,allow- ing estimationof overallprogrameffects. The programreduced school absenteeismin treatmentschools by one-quarter,and was far cheaper than alternativeways of boost- ing school participation.Dewormingsubstantiallyimprovedhealth and school partic- ipation among untreatedchildren in both treatmentschools and neighboringschools, and these externalitiesare large enough to justify fully subsidizingtreatment.Yet we do not find evidence that dewormingimprovedacademictest scores.