DC Accepted Papers Paper:
Employer Market Power and Hospital Inpatient Prices Among the Privately Insured
*Names in bold indicate Presenter
Study Design: Linear regression analyses at the Metropolitan Statistical Area (MSA)-year level were conducted with the log of prices as the dependent variables and the log of the LMC measure as the key independent variable. LMC, like an HHI, was defined as the weighted average within an MSA of the county-level sum of squared firm-level market shares. Average prices were calculated based on a market basket of the 15 most common services (identified by Diagnosis Related Group code [DRG]) in the MarketScan inpatient database in 2016. For each DRG code, average total payment for admissions was calculated within an MSA in 2016, weighted by the overall frequency of the DRG code nationally, and then summed to generate the final price measure. Models were adjusted for MSA-year level market characteristics and demographics and year fixed-effects.
Population Studied: LMC and inpatient prices were studied for 361 MSAs in the United States from 2010-2017. Data from the US Census Bureau County Business Patterns (CBP) was used to calculate LMC. Truven MarketScan® commercial claims are used for the years 2010-2016 to estimate MSA-year level prices. The MarketScan data included private-sector employer-sponsored health data from approximately 350 payers for employees and their dependents, covering roughly 25% of all the commercially insured individuals in the US. The two datasets were combined from the years 2010 to 2016 using MSA identifiers.
Principal Findings: Among the 2,527 MSA-years available in the MarketScan and CBP data, labor markets are highly unconcentrated (mean = 60.76, SD = 48.85) at a county-weighted level. The average price of admission based the market basket definition was $7,264 (SD = 1,730). In the regression analyses, LMC was found to be significantly associated with hospital prices (p-value < 0.05), adjusting for MSA-level market characteristics, demographics, and year fixed-effects. A change from the 1st to 3rd quartile of LMC was associated with a 2.06% decrease in prices.
Conclusions: LMC was associated with slightly lower prices for hospital admissions.
Implications for Policy: Our results suggest that in most markets, where hospital concentration is high, employers are unlikely to have enough market power to negotiate lower prices. The slight relationship between LMC and prices suggests, however, that purchasing alliances between employers might be able to reduce prices in some markets, but further government action may be needed to achieve lower hospital prices broadly.