Hazed and Confused: Air Pollution, Financial Decision Making, and Health Outcomes among the Elderly

In this article we develop a dynamic structural model of the life cycle from in utero to death. Accumulated exposure to air pollution and other environmental toxins affects physical capital and human capital, which affect wages and medical expenditures.  Workers purchase goods, accumulate wealth, and make location decisions which, in turn, affect their exposure to pollution.  Human capital also affects the costs of market transactions, so that ceteris paribus people with lower human capital are more likely to leave money on the table when purchasing market goods. As workers age and develop chronic medical conditions they may increase the quantity and/or quality of their lives by purchasing prescription drugs and they may lower their costs of drugs by purchasing health insurance. One contribution of the model is that it provides a unified framework to formally connect several disparate strands of research in the fields of health, labor, and environmental economics.  Another contribution of the model is to highlight several potentially important parts of this problem that have received little to no attention in the prior literature.

We use these insights to guide an empirical investigation into the spatial joint distribution of air pollution, Alzheimer's and other chronic illnesses, expenditures on prescription drugs, and financial decision making at late stages of the life cycle.  We accomplish this by combining CMS administrative data with the MCBS data for 2005-2010. The administrative data indicate Medicare beneficiaries’ housing locations, demographics, chronic conditions, drug claims and drug spending, set of available prescription drug plans (PDPs) and PDP choices. We also use a cost calculator developed in our prior work to estimate the costs that each person would have incurred under every PDP available to them. The MCBS data shed further insight on the individuals’ demographics, effort to search for information about health insurance, decision making processes and knowledge about the PDP market. We combine these CMS data with data on housing transactions and ambient concentrations of local air pollutants (including particulate matter, ozone, sulfur dioxide, carbon monoxide, nitrogen dioxide, and EPA’s NATA data on localized cumulative cancer and non-cancer risks from 139 specific air toxins) to look at the effects of air pollution on the health of Medicare beneficiaries and their financial decision making, which we observe from their choices in the PDP market and housing market. Finally, we consider whether particular drugs or drug classes offer protective effects against any negative consequences of pollution.