ORCID

https://orcid.org/0000-0003-0560-7744

Date of Award

Fall 2025

Language

English

Embargo Period

11-14-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

College/School/Department

Department of Economics

Program

Economics

First Advisor

Pinka Chatterji; Chun-Yu Ho

Committee Members

Baris K Yoruk

Keywords

Physician Prescribing; Knowledge Spillover; Medicare Part D Accessibility; Disability Inequality; Job Demands

Subject Categories

Econometrics | Health Economics | Public Economics

Abstract

The first chapter examines relationships between physicians exposed to new drugs’ clinical trials and prescribing. Clinical trials benefit patients directly by providing novel treatments, but little is known about the indirect effects of clinical trials on physician prescribing through localized knowledge spillover. This study examines whether exposure to a clinical trial of a new drug in a physician’s local geographic area affects the physician’s propensity to prescribe the drug, i.e. a localized knowledge spillover from clinical trial sites to physician. Utilizing the Medicare Part D prescribing patterns of more than 10,000 physicians across 29 new cancer drugs approved by the US Food and Drug Administration between 2014 and 2019, I find that an exposure to clinical trials of new cancer drugs increases the likelihood that physicians prescribe these drugs by 0.18 percentage points, representing a 14% increase relative to the average prescribing rate. Notably, the localized knowledge spillovers are more pronounced when the proximate clinical trial site hosts the leading researcher of the clinical trial and the physician is affiliated with the proximate clinical trial site. There is no clear difference on the localized knowledge spillover between trials before and after FDA approval. Nonetheless, the localized knowledge spillover is stronger for physicians graduating from higher-ranked medical schools, and with more experienced physicians, and male physicians.

The second chapter investigates the geographical accessibility of preferred pharmacies in the Medicare Part D market (PDPs and MAPDs) from 2010 to 2024, quantifying disparities in average and additional distance to nearest preferred pharmacy, and cost-savings across ZCTAs stratified by rurality and racial/ethnic composition. While preferred pharmacy networks have expanded, particularly in the MAPD market, a significant access gap remains, with 31.8% of ZCTAs lacking a preferred pharmacy in 2024, disproportionately affecting areas with high white and American Indian/Alaska Native (AI/AN) populations. For beneficiaries with access, the cost-benefit trade-off is highly favorable: the average cost saving of approximately $5.09 outweighs the minor average additional distance of 0.87 miles. Nevertheless, a persistent and significant rural-urban disparity was confirmed by multivariate analysis, showing that rural ZCTAs face up to 1.5 times longer additional distances to preferred pharmacies (P < 0.01) compared to urban areas. Furthermore, ZCTAs with higher proportions of Hispanic or Black residents experience slightly greater additional distances, suggesting gaps in network inclusion, while AI/AN areas face the longest overall distances but the highest cost-saving potential. These findings highlight that while preferred networks lower costs for many, the resulting access and travel burdens are unevenly distributed, underscoring the critical need for policymakers to monitor network structures to ensure equitable access to affordable medications across all Medicare beneficiary groups.

The third chapter investigates the evolving income gap experienced by people with disabilities (PWD) among the aging US workforce (1996–2018). It moves beyond static analyses by employing a dynamic Recentered Influence Function (RIF) - Oaxaca-Blinder decomposition in conjunction with the Erreygers Index (EI) to systematically partition the change in income-related disability inequality over two decades. The core innovation is the simultaneous inclusion of subjective (HRS-based) and objective (O*NET-based) job demands as covariates. Findings reveal a growing, pro-poor income inequality in disability status over the period, driven primarily by the Structure Effect—changes in the economic penalty associated with job requirements—rather than the Composition Effect (changes in the distribution of workers). Specifically, structural volatility is maintained by an intense, offsetting battle between the valuation of occupational demands, where changes in the returns to Objective Physical Effort and Stooping/Kneeling/Crouching requirements are the dominant structural forces. Contrasting subjective and objective measures reveals that while workers’ self-assessments of demands are often higher, it is the structural valuation of specific, objectively measured physical demands that overwhelmingly dictates the dynamics of income-related disability inequality.

In summary, this dissertation investigates the interconnected factors required for achieving health equity by analyzing three dimensions: information flow, geographic access, and labor market structure. Chapter 1 establishes that clinical trial sites act as key knowledge hubs, demonstrating how information about new medical technologies, specifically drug adoption, primarily flows through geographic networks. Chapter 2 shifts focus to resource distribution, highlighting the geographic unevenness of access to affordable drugs, with rural and minority communities facing specific barriers. Finally, Chapter 3 examines the relationship between work and disability, arguing that widening inequality is driven by labor market structures that disproportionately value certain job demands over others, thereby shaping long-term health and economic outcomes. Collectively, these findings underscore that attaining health equity is a multi-dimensional challenge requiring targeted interventions across information transmission networks, equitable geographic access to healthcare resources, and structural reform of the labor market that mediates health and economic well-being.

License

This work is licensed under the University at Albany Standard Author Agreement.

Download

Share

COinS