School Closure Policies and Infectious Diseases (Job Market Paper)

COVID-19 has drawn people’s attention to public schools’ influence on infection rates, but understanding its magnitude remains highly imperfect. To deepen our knowledge of factors affecting the transmission of infectious diseases with properties similar to COVID-19, I study influenza, pneumonia, and respiratory infection to uncover intertemporal, within-family, and across-age cohort infection patterns. The key policy question is the extent to which changing school vacations, opening, and closing dates affects infection transmission, which affects not only School-age children but also preschool, college, and adult populations. I combine patient information and diagnoses from the Merative® (formerly IBM) MarketScan® Commercial Database between July 1. 2010, and June 30, 2019, with MSA-level weekly school data previously collected by the author with coauthors documenting school opening and closing dates over the same pre-pandemic period (Chen et al., 2021). I use linear probability models that also include weather and other MSA-level control variables on a sample of 122,487,230 individuals and their weekly diagnostic data. I find that within- family infection rates of pneumonia, influenza, and respiratory infection, especially high school students’ infection rates, rise as the number of days schools are open. Infected primary and high school students are the main introducers of pneumonia, influenza, and respiratory infections. School boards and local governance can use the methodology and results of this study to shape school closure policies that improve student welfare and limit the spread of infectious diseases.

Synthetic Index of Medicaid Dental Benefit Generosity among Older Adults (with Astha Singhal, DDS)

The high cost of dental care remains a major barrier for low-income older adults. Apart from the absence of a universal measure that determines Medicaid dental benefits generosity across states, the employment of annual maximum limits (AML) by Medicaid programs makes dental services less affordable. In this study, we first develop an unbiased synthetic measure of Medicaid dental benefits coverage for a nationally representative sample of 4219 older adults from the 2019 Medicare Current Beneficiary Survey (MCBS). This measure calculates the proportion of dental procedures that each state’s Medicaid dental policy can cover. We then use this measure to estimate Medicaid payment generosity by examining the proportion of older adults whose dental payment is under the AML of each state. Results show the most common dental services are exams, x- rays, and cleanings, which sum up to more than 70 percent of 19,950 dental services. We calculate that states that excluding Medicaid coverage for these common dental services, frees up 10 percent of their dental payments to use by Medicaid for covering other services while remaining below the same AML. Understanding this tradeoff may be helpful to policymakers examining state variation in Medicaid dental coverage and payment generosity, informing future policies that improve the quality of life for older people.

Topcoding, reinsurance, and outlier adjustments to the Diagnostic Cost Group (DCG) risk adjustment payment model (with Corinne Andriola)

Andriola et al. (2023) just completed a major project that build on the rich Diagnostic Items (DXI) classification system of Ellis et al. (2022) by developing a new machine learning algorithm that enables researchers and policymakers to calibrate ready-to-use risk adjustment payment formulas that achieve high predictive power, avoid underpaying rare diagnoses, minimize use of vague and gameable diagnostic information, and groups together DXIs with similar cost implications so as to reduce incentives for upcoding and keep the model parsimonious in the number of parameters used. This Diagnostic Cost Group (DCG) framework was developed and thus far has only been evaluated for predicting only one outcome: concurrent year total health care spending topcoded at $250,000. In this paper, we use the DCG clustering algorithm on concurrent year untopcoded total spending and prospective year spending with and without topcoding. Also, drawing upon the work of Tom McGuire and Richard Van Kleef, we examine the performance of the DCG framework when combined with mixed payment formulas, outliers, and reinsurance strategies in order to evaluate its performance relative to the existing payment formulas.