Health and mo­bil­ity im­pacts of in­dus­trial chem­i­cal ac­ci­dents on Medicare ben­e­fi­cia­ries (Job Market Paper)

Industrial chem­i­cal ac­ci­dents pre­sent sig­nif­i­cant threats to pub­lic health and com­mu­nity sta­bil­ity, of­ten dis­pro­por­tion­ately af­fect­ing mi­nor­ity and other dis­ad­van­taged pop­u­la­tions. Medicare ben­e­fi­cia­ries, peo­ple aged 65 and older, and long-term dis­abled are par­tic­u­larly vul­ner­a­ble to chem­i­cal ac­ci­dents be­cause of their age and pre­ex­ist­ing health con­di­tions. This pa­per in­ves­ti­gates the long-term health and mo­bil­ity im­pacts of se­vere chem­i­cal ac­ci­dents on Medicare ben­e­fi­cia­ries in the US. Using ad­min­is­tra­tive Medicare data on over 3 mil­lion el­derly and long-term dis­abled ben­e­fi­cia­ries, I find that ac­ci­dents in­crease the mor­tal­ity rate of ben­e­fi­cia­ries by 5% and in­crease rates of in­pa­tient hos­pi­tal vis­its and ER vis­its. Back-of-the-envelope cal­cu­la­tions sug­gest the health ben­e­fits of pre­vent­ing chem­i­cal ac­ci­dents ex­ceed the costs of en­hanc­ing risk-man­age­ment prac­tices in in­dus­trial fa­cil­i­ties. Moreover, ben­e­fi­cia­ries’ mo­bil­ity rate in­creased by 4% af­ter ac­ci­dents. Disparities rise as Black ben­e­fi­cia­ries are more likely to move, and they sys­tem­at­i­cally sort into places with higher lev­els of PM2.5. The pol­lu­tion ex­po­sure gap be­tween Black and non-Black ben­e­fi­cia­ries could be widened by in­dus­trial chem­i­cal ac­ci­dents.

This pa­per es­ti­mates the short-run and long-run im­pacts of early child­hood lead ex­po­sure from drink­ing wa­ter on ed­u­ca­tional out­comes, the spa­tial and de­mo­graphic dis­tri­b­u­tion of these im­pacts, and the wel­fare im­pacts of lead abate­ment poli­cies. I merge data from the U.S. Environmental Protection Agency on lead vi­o­la­tions un­der the Safe Drinking Water Act with data on in­di­vid­ual stan­dard test scores, ed­u­ca­tional at­tain­ment, and wages from re­stric­tive-use Texas data. I also match lead con­cen­tra­tion in drink­ing wa­ter with Blood Lead Levels (BLLs) data, ed­u­ca­tion and la­bor-mar­ket out­comes for a sub­group of in­di­vid­u­als for whom we can iden­tify their drink­ing wa­ter provider in Texas to un­der­stand the dose-re­sponse im­pacts of lead in drink­ing wa­ter. I find that lead ex­po­sure at birth from drink­ing wa­ter has significant neg­a­tive im­pact on stu­dents’ 3rd grade stan­dard test scores. In the long run, it also significantly re­duce the high school grad­u­a­tion rate.

Bringing Hydrologic Realism to Water Markets, with Danielle Grogan, Matthew D. Lisk, Shan Zuidema, Karen Fisher-Vanden4, Richard B. Lammers, Sheila M. Olmstead, Lara Fowler and Alexander A. Prusevich (In sub­mis­sion)

Water scarcity in arid re­gions is chal­leng­ing cur­rent wa­ter man­age­ment in­sti­tu­tions. Water mar­kets are im­por­tant adap­ta­tion tools as wa­ter stress in­creases be­cause they move wa­ter from lower- to higher-val­ued uses. However, analy­sis of wa­ter mar­kets lack­ing phys­i­cal wa­ter sys­tem dy­nam­ics can­not ad­e­quately cap­ture mar­kets’ po­ten­tial or lim­i­ta­tions. We de­velop a novel in­te­grated model of wa­ter rights trad­ing and phys­i­cal hy­drol­ogy to eval­u­ate wa­ter mar­kets in the U.S. West, and find that po­ten­tial agri­cul­ture-to-ur­ban trad­ing may gen­er­ate net eco­nomic re­turns two or­ders of mag­ni­tude larger than ex­ist­ing, lim­ited wa­ter mar­kets. Where ground­wa­ter rights are not reg­u­lated, mar­kets ex­ac­er­bate ground­wa­ter de­ple­tion, but dif­fer­ences in con­sump­tive wa­ter use al­low agri­cul­ture to re­tain un­ex­pected lev­els of wa­ter re­sources, even with strict ground­wa­ter con­straints.

Harmonized Database of Western U.S. Water Rights, with Matthew D. Lisk, Danielle S. Grogan, Shan Zuidema, Robert Caccese, Darrah Peklak, Sheila Olmstead, Karen Fisher-Vanden, Richard B. Lammers, and Lara Fowler (In Submission)

In the arid and semi-arid west­ern U.S., ac­cess to wa­ter is reg­u­lated through a le­gal sys­tem of wa­ter rights. Individuals, com­pa­nies, or­ga­ni­za­tions, mu­nic­i­pal­i­ties, and tribal en­ti­ties have doc­u­ments that de­clare their wa­ter rights. State wa­ter reg­u­la­tory agen­cies col­late and main­tain these records, which can be used in le­gal dis­putes over ac­cess to wa­ter. While these records are pub­licly avail­able data in all west­ern U.S. states, the data have not yet been read­ily avail­able in dig­i­tal form from all states. Furthermore, there are many dif­fer­ences in data for­mat, ter­mi­nol­ogy, and de­f­i­n­i­tions be­tween state wa­ter reg­u­la­tory agen­cies. Here, we have col­lected wa­ter rights data from 11 west­ern U.S. state agen­cies, har­mo­nized ter­mi­nol­ogy and use de­f­i­n­i­tions, for­mat­ted them con­sis­tently, and tied them to a west­ern U.S.-wide shape­file of wa­ter ad­min­is­tra­tive bound­aries. We demon­strate how these data en­able con­sis­tent re­gional-scale west­ern U.S. hy­dro­logic and eco­nomic mod­el­ing.

Suicide and Lithium in the Public Water Supply of 870 US Counties, with Paul von Hippel and Sheila Olmstead

Several stud­ies have re­ported that sui­cide rates are lower in ar­eas with higher con­cen­tra­tions of lithium in drink­ing wa­ter. Some au­thors have rec­om­mended adding lithium to the pub­lic wa­ter sup­ply. We es­ti­mate the as­so­ci­a­tion be­tween sui­cide and drink­ing-wa­ter lithium in the largest dataset yet used for this pur­pose. In 870 US coun­ties, we regress county sui­cide rates on lithium con­cen­tra­tions in ground­wa­ter and sur­face wa­ter. We con­trol for spa­tial cor­re­la­tion and county-level cor­re­lates of sui­cide rates, in­clud­ing eco­nomic con­di­tions and re­li­gious, racial, and eth­nic com­po­si­tion. With or with­out co­vari­ates, we find no significant as­so­ci­a­tion be­tween county sui­cide rates and the con­cen­tra­tion of lithium in the wa­ter sup­ply. Our re­sults do not sup­port a pol­icy of adding lithium to the wa­ter sup­ply.