Hassles and Environmental Health Screenings: Evidence from Lead Tests in Illinois

Lead paint, a harmful environmental hazard, can still be found in millions of homes in the United States. Due to high inspection and clean-up costs, prevention programs target intervention to the relatively few homes where small children test positive for lead poisoning. Because children have to visit a doctor to get tested, only households willing to undergo this hassle self-select into screening. Is self-selection an effective targeting mechanism? I study screening take-up by analyzing geocoded 2001-2016 lead screening data on 2 million Illinois children. My empirical strategy exploits variation in travel costs due to healthcare providers’ openings and closings. I find that travel costs reduce screening among lowand high-risk households alike, without improving targeting. Consistent with low poisoning rates, high-risk households are only willing to pay $4-29 more than low-risk households for screening. Despite poor targeting, screening incentives may be cost-effective because of the externalities of lead exposure. ∗Department of Economics, University of Chicago. Email: lgazze@uchicago.edu. I am indebted to Michael Greenstone for his mentorship throughout my postdoctoral scholarship. Ali Abbasi, Marcella Alsan, Alex Bartik, Fiona Burlig, Thomas Covert, Catie Hausman, Michael Kofoed, David Meltzer, Rebecca Meyerson, Jack Mountjoy, Nick Sanders, Tommaso Sonno, Dan Waldinger, and seminar and conference participants at EPIC, Indiana University, Urban Labs, APPAM, ASHEcon, the H2D2 Research Day, and the 4th Marco Fanno Alumni Workshop provided helpful comments and suggestions. I am also extremely grateful to the staff at the Illinois Department of Public Health for sharing the data for this analysis as well as their insights and expertise in interpreting the results. The conclusions, opinions, and recommendations in this paper are not necessarily the conclusions, opinions, or recommendations of IDPH. This project would not have been possible without the generous support of the Joyce Foundation. Bridget Pals and Xiyue (Iris) Song provided excellent research assistance. All remaining errors are my own.