Impacts of the ACA Medicaid expansion on health behaviors: Evidence from household panel data

Abstract

A motivation for increasing health insurance coverage is to improve health outcomes for impacted populations. However, health insurance coverage may alternatively increase risky health behaviors due to ex ante moral hazard, and past research on this issue has led to mixed conclusions. This paper uses a panel of household purchases to estimate the effects of the recent state‐level Medicaid expansions resulting from the Affordable Care Act (ACA) on consumption goods that present adverse health risks. We utilize within‐household variation to identify whether increases in Medicaid availability impacted household purchase patterns of alcohol, nicotine‐related, snack food, and carbonated beverage products. Overall, we find little evidence that the ACA Medicaid expansion led to ex ante moral hazard across any of these products, but we find compelling evidence that the Medicaid expansions reduced cigarette consumption and increased smoking cessation product use among the Medicaid‐eligible population.

Public Insurance Expansions and Smoking Cessation Medications

“…We find that some of this increase is due to cost‐shifting of ongoing medication use from private insurance and self‐paying uninsured patients to Medicaid, most of the increase reflects new prescription medications. We find no evidence that Medicaid expansion increased over‐the‐counter nicotine replacement therapy medications (NRTs) financed by Medicaid, in contrast to a study by Cotti, Nesson, and Tefft () using household scanner data and OTC purchased without a prescription. Our study of smoking outcomes does not suggest that Medicaid expansion lead to reductions in smoking among likely Medicaid beneficiaries, thus contributing an estimate to a mixed body of literature (Cawley, Soni, and Simon ; Cotti, Nesson, and Tefft ; Courtemanche et al , ; Koma et al ).…”

“…Indeed, pre‐ACA research documents that enrollee knowledge of Medicaid coverage for cessation services was poor (McMenamin et al ). The pattern of no effect in Medicaid‐financed claims and large positive effects for overall OTC purchases shown by Cotti, Nesson, and Tefft () using ACA Medicaid variation in eligibility is in line with this explanation. Finally, we note that, while not statistically different from zero, our coefficient estimate implies a 10% increase in Medicaid‐financed OTC cessation medication prescriptions postexpansion with the top tail of the 95% confidence interval implying an even larger increase.…”

“…However, a follow‐up study by these authors using 3 years of postexpansion data found more limited effects on smoking (Cawley, Soni, and Simon ). On the other hand, Cotti, Nesson, and Tefft () find that cigarette purchases decline postexpansion using household scanner data, and Koma et al () show increased smoking cessation among low‐income smokers postexpansion. Finally, Courtemanche et al (, ), using a triple difference estimator, find no evidence that these expansions affected smoking outcomes.…”

“…In terms of behavioral health outcomes (substance use disorders [SUD] and mental illness), expansion increased Medicaid-financed prescriptions for SUDs by 55% to over 100% (Maclean and Saloner 2019;Meinhofer and Witman 2018;Wen et al 2017) and mental illness by 21% (Maclean et al 2018a). In terms of nonprescription medication utilization, Cotti, Nesson, and Tefft (2019), use Nielson scanner data and document that the probability that a household purchased an OTC cessation medication increased 50% postexpansion.…”

“…Smoking cessation prescription medications are important to study, separate from other classes of medications, given that smoking is distinct from general health (e.g., cancer, hypertension, diabetes) and other behavioral health (e.g., mental illness, SUDs) outcomes in terms of addiction, the associated costs to the individual and society (for instance, SUDs are linked with crime and violence [Caulkins, Kasunic, and Lee 2014], which are not typical features of nicotine addiction), utility derived from engaging in the behavior (e.g., smoking, unlike many general health conditions such as diabetes, provides utility to the individual consuming the product), and treatment costs (the ACA requires zero cost sharing for smoking cessation medications for newly eligible population but this is not the case for all medications). Finally, prescription medications, unlike the OTC medications examined by Cotti, Nesson, and Tefft (2019), require a consultation with a healthcare provider to secure a prescription, which adds to the financial and time cost of these medications for enrollees.…”