Did Maryland All-Payer Global Budgets Change Hospital Margins and Revenue per Discharge? A Comparative Interrupted Panel Analysis, 2010–2023

This registration archives the analysis workflow and non-restricted materials for a comparative interrupted panel evaluation of Maryland’s all-payer Global Budget Revenue (GBR) program implemented in 2014. The study compares acute-care hospitals in Maryland (treatment) and Massachusetts (comparison) over fiscal years 2010–2023 using a hospital–year panel constructed at the CMS Certification Number (CCN) × fiscal-year level. Financial outcomes are derived from RAND’s Hospital Cost Report Information System (HCRIS-derived) files (licensed; not redistributed), and provider characteristics are drawn from CMS’s Provider of Services (POS) File – Hospital & Non-Hospital Facilities (publicly available). The analytic approach is a two-way fixed-effects difference-in-differences design with hospital and year fixed effects and hospital-clustered standard errors, supplemented by (i) an event-study specification to assess pre-trends and dynamic effects and (ii) a placebo falsification test using a fake treatment date (2012) restricted to the pre-period. A COVID-period volatility analysis evaluates whether GBR is associated with differential changes in year-to-year margin volatility during 2020–2022. Expected outputs include regression tables for the main DiD models, robustness checks (including winsorized volatility variants), and figures showing state-year trends and the event-study coefficients. This registration is intended to support transparency and reproducibility by providing code, logs, and non-restricted outputs sufficient to reproduce all results conditional on obtaining access to the licensed RAND data directly from RAND.