Working Paper Series

The PSTC Working Papers series aims to disseminate works-in-progress publications that reflect the broad range of research activities at PSTC. We welcome submissions from our associated faculty, postdocs as well as graduate students, but with a letter of support from their faculty advisor. In addition to our Web site, our collection will be archived at the Brown Digital Repository, which is permanent, safe, and accessible, and all papers in the series may be searched and downloaded. The working paper will also be deposited on our section of the Social Science Research Network Web site.

PSTC does not hold the copyright permissions for our Working Papers Series. The copyright remains with the author of the paper. For more information on submitting a working paper, see the PSTC Working Papers Series Submission Guidelines.

Identifying the Hidden Costs of a Public Health Success: Arsenic Well Water Contamination and Productivity in Bangladesh

ISSN 2163-0062 (online)
July 2012
The contamination of groundwater by arsenic in Bangladesh is the largest poisoning of a population in history. In this paper we provide new estimates of: 1) the effects of the consumption of foods grown and cooked in arsenic-contaminated water on individual arsenic concentrations and of; and 2) the effects of the ingestion and retention of inorganic arsenic on direct measures of cognitive and physical capabilities as well as on the schooling attainment, occupational structure, entrepreneurship and incomes of the rural Bangladesh population. We use data from a panel survey of rural Bangladesh households with information on individual’s retained arsenic based on toenail clippings that we collected from over 7,000 respondents. We employ an estimation strategy for identifying the causal effects of retained arsenic that exploits the known genetic linkages in abilities to methylate arsenic among family members who reside in different villages net of the effects of diet and the water quality in their villages. We also use genomic data to verify alternative explanations for our results.