Tradition and mortality: Evidence from twin infanticide in Africa

Abstract

Traditions can limit investment in early life health, even if they have been abandoned. We introduce data on historic twin infanticide and merge it with recent birth records from 23 African countries. We use the full sample and a border sample of adjacent societies with and without past twin infanticide. Both samples provide no evidence that past twin infanticide predicts greater differential twin mortality today. This null result is likely a consequence of suppression efforts by Africans, missionaries, and colonial governments. Where these channels were weak, we find evidence of greater twin mortality today.