Improving RPP Estimation Using Commercial and Medicare Claims
The U.S. Bureau of Economic Analysis (BEA) produces regional price parities (RPPs) to measure differences in the price level of goods and services across states and metropolitan areas in the United States. Health care, which accounts for approximately 16 percent of personal consumption expenditures (PCE), presents a longstanding measurement challenge for the RPP program: because geographically detailed and nationally consistent health care price data have historically been limited, many health care categories are assigned national price levels in the RPP estimation process. New claims-based data sources create an opportunity to enhance this treatment by introducing state-level price variation for health care.
A recent BEA working paper develops state-level health care price inputs using two large claims sources: the Merative MarketScan Commercial Database and Medicare fee-for-service (FFS) claims from the Centers for Medicare & Medicaid Services. Authors Calvin Ackley, Kun Li, and Kyle K. Hood estimate price levels for six categories of medical services that align with the PCE expenditure classification. Their study demonstrates a coherent approach for incorporating health care price information into the RPPs and allows for constructing an expenditure-weighted health care aggregate price level that summarizes geographic variation in medical prices across payer types and service categories.
The authors first organize claims-based health care price estimates by PCE expenditure category rather than by claim type. By classifying claims into PCE-aligned categories, their estimates can be incorporated into the RPP framework while preserving the underlying expenditure classification. Second, the authors estimate comparable price levels using both commercial and Medicare FFS claims. This allows the authors to distinguish payer-specific geographic price patterns and to construct payer-weighted aggregate measures that better reflect the composition of health care spending (chart 1). Third, they show how incorporating these medical price inputs affects state all-items RPPs and supports construction of a standalone health care RPP.
[View larger chart]
The resulting estimates behave as expected. Large differences in health care service prices across regions produce greater dispersion in RPPs, and the magnitude of these effects aligns with the substantial role of health care within PCE. States where health care is relatively more expensive may shift in their relative rankings once these prices are included. Price variation is notably larger among private payers, whereas Medicare prices remain comparatively stable. Overall, the resulting regional price levels exhibit consistency over time.
