Research
Working Papers
Get Fit with MA? Estimating the Impact of Medicare Advantage on Physical Activity Paper
(with Li J.)
Medicare Advantage plans increasingly offer fitness benefits, yet the evidence on their behavioral effects remains limited. This paper studies whether Medicare Advantage enrollment increases physical activity among older adults using a regression discontinuity in difference-in-difference design centered on the age-65 Medicare eligibility threshold. Using nationally representative survey data on workout behavior with insurance coverage information, we find that the discontinuity in the probability of daily exercise at age 65 is significantly larger for MA enrollees than for non-MA beneficiaries. Estimates for ages below 65 show no evidence of differential pre-trends, while the post-65 pattern indicates that the effect is concentrated in the short run. The response is larger among women, non-White, and individuals without a college degree. Robustness checks, including placebo tests and a donut-hole analysis yield consistent results. These findings suggest that insurance design can affect preventive behaviors and highlight supplemental benefits as a potentially important policy margin for promoting healthy aging.
Move to Move? Neighborhood Fitness Culture and Exercise Among Older Adults
This paper studies whether neighborhood fitness culture affects physical activity among older adults. Using the Health and Retirement Study from 2004–2022 linked to restricted ZIP-code identifiers, I measure neighborhood fitness culture as the ZIP-level prevalence of physical activity among older adults. I first estimate a mover-only design that relates changes in individuals’ exercise behavior to the difference between destination and origin fitness culture. Moving to a more active neighborhood is associated with higher moderate physical activity, conditional on demographic and health controls, built-environment measures, and origin and destination selection residuals. To address residual sorting, I then estimate an event-study design comparing weather/climate movers to never-movers. As these movers relocate for climate reasons rather than fitness amenities, the resulting change in neighborhood fitness culture is more plausibly incidental. The event-study evidence suggests that workout participation falls after moves to less active neighborhoods, with weaker positive responses among movers to more active neighborhoods. These results indicate that local fitness culture shapes preventive health behavior and point to the potential value of place-based policies for healthy aging.
Where Seniors Sweat: Geographic Variation in SilverSneakers Uptake
(with Rovniak L.)
Abstract forthcoming.
Cross-national Analysis of the Determinants of Dementia Prevalence: Evidence from the United States and Caribbean Countries
(with Li J. & Dow W.)
Abstract forthcoming.
Predictors of Discontinued Participation in a Fruit and Vegetable Voucher Program: The Role of Community Organizations in Engaging Non-English Speakers
(with Chaparro, P., Wallace, J., Baquero, B., Jones-Smith, J., & Knox, M.)
Healthy food voucher programs (HFVPs) are an important fiscal tool for addressing inequities in healthy food consumption. Fresh Bucks, a HFVP in Seattle, Washington, USA, provides US$40/month for fruit and vegetable (FV) purchases to low-income households, encouraging long-term participation and utilizing a community-supported enrollment model to engage non-English speaking households. Previous research demonstrated that Fresh Bucks participation reduced food insecurity and increased FV consumption; yet 46% of participating households chose to discontinue participation when requested to re-apply ~2 years post initial enrollment. We sought to identify factors that predict discontinuation of Fresh Bucks participation, particularly among non-English speakers
Do Unionized Firms Have Better Management Practice? Paper
(with Basu A)
Using restricted data from the World Management Survey (WMS), we examine the relationship between labor unions and management practices across private manufacturing firms in Latin America, North America, and Europe. While prior research shows that management quality is strongly correlated with firm performance, the role of unions in shaping managerial practices remains unclear. We find that unionized firms are significantly better managed overall. However, this aggregate relationship masks substantial heterogeneity across different management domains. Firms with higher union rate score lower on People Management practices, reflecting constraints on hiring, promotion, and dismissal, but score markedly higher on Operations, Monitoring, and Target-setting practices. These patterns suggest that unions are associated with the adoption of more structured performance systems while limiting managerial discretion in personnel management