Housing quality, neurotoxins and human capital acquisition

Abstract

Much existing literature on age–earnings variation has focused on identifying the connection between ability and wage and salary income. This is often done through the careful selection of an instrument, believed to be correlated with ability but not with earnings, and typically for data on individuals older than 18. Recent efforts have discussed the path of human capital acquisition from younger ages, including discussions of the relevance of early childhood education on the long-term path of human capital. This paper approaches the issue by selecting a more plausibly exogenous factor in human capital acquisition – infant exposure to household neurotoxins (e.g. lead, mercury, PCBs, etc.). I capitalize on the time varying usage of paint and PCBs in household constructions to use housing vintage (year of construction) as a proxy for exposure. I measure this exposure at the state level. State level variation in the age of housing stock is found to explain a large portion of the individual level age–earnings profile. Living in an old vintage house with a high probability of neurotoxin exposure translates to loss of earnings equivalent to about six years of schooling. This supports the medical literature's view that neurotoxins affect both mental capacity as well as the ability to learn.

Acknowledgements

Many thanks are due to Steven Durlauf, Rody Manuelli, Ananth Seshadri and Kenneth West for ongoing advice and encouragement. The author is grateful as well for comments on an earlier draft from John Kennan and Ritesh Banerjee. The author further acknowledges financial support from the Institute for Research on Poverty and the Wisconsin Department of Economics. All errors remain the author's.