Working Paper: NBER ID: w23290
Authors: Allison B. Rosen; Kaushik Ghosh; Emily S. Pape; Marcelo Coca Perraillon; Irina Bondarenko; Kassandra L. Messer; Trivellore Raghunathan; Susan T. Stewart; David M. Cutler
Abstract: As health care financing, organization, and delivery innovations proliferate, the need for comprehensive, detailed data on medical spending has never been more apparent. This study builds on previous work to provide a more comprehensive accounting of medical spending at the individual level than has been done in the past. We account for spending by the entire population: the civilian, non-institutionalized population that is the subject of past studies, as well as high medical spenders, the institutionalized, the incarcerated, and active-duty military personnel. We use within-imputation and other adjustments to build a micro dataset and reconcile survey data based on our estimate of medical spending to the National Health Expenditure Accounts (NHEA). The micro dataset we build can be used for more detailed policy evaluations that more closely reflect true national personal health expenditure at the individual level.
Keywords: health care spending; national health expenditures; survey data; micro dataset
JEL Codes: I1
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Adjustments made to survey data (C83) | Accuracy of national health expenditure accounts (E01) |
| Enhanced dataset reconciles survey-reported spending with NHEA (E01) | Accurate representation of health care expenditures across the entire US population (H51) |
| Adjustments made to survey data (C83) | Closer alignment with national totals (Y10) |
| Adjustments made to survey data (C83) | Improved understanding of health expenditure distribution among different demographic groups (H51) |
| Estimates indicate significant underreporting of medical expenditures in existing surveys (H51) | Addressing underreporting through methodological enhancements (C83) |
| Reconciliations improve accuracy of health expenditure estimates (H51) | Facilitate more detailed policy evaluations (D78) |