This paper is published as part of the Hutchins Center on Fiscal and Monetary Policy’s Productivity Measurement Initiative.
More than two decades ago a well-known study provided evidence from heart attack treatments suggesting that prices in medical care were actually declining, when appropriately adjusted for quality. The topic has only grown in importance in the past two decades, as the share of the gross domestic product (GDP) devoted to medical care rose substantially. Health experts believe that much of the growth in this sector is driven by new technologies that improve treatment in the long run and recent work has shown that new innovations have been a key factor behind the rapid growth in expenditures over this period for many conditions, including rheumatoid arthritis, cancer, hepatitis, and HIV. Meanwhile, life expectancy at birth in the United States has increased. While innovations are a key contributing factor to the growth in spending for medical care, changes in the quality of medical care are not reflected in U.S. national statistics, leading official statistics to overstate inflation in this sector.
This paper revisits this subject looking at a large number of conditions and more recent and more comprehensive data sources to compare alternative methods of quality adjustment. The goals of this paper are to establish a framework relating the different methods, to illustrate the differences between them empirically, and to demonstrate that price declines are found in a different time period and over a wider set of conditions than previously studied.
The main contribution of this paper is that it shows that the method of constructing quality adjusted price indexes matters theoretically and empirically. Applying a consistent methodology of utility-based quality adjustment across a wide range of studies (e.g., studies that differ on a variety of dimensions such as how they measure cost and quality and applying widely different data sources) produces surprisingly consistent results of quality adjusted prices declining. These estimates have important implications for the measurement of output and productivity growth. The Bureau of Labor Statistics (BLS) estimates multifactor productivity growth for the hospital and nursing home sector to be negative over the 2001-2014 period, with an annual decline of 0.3 percent. Under the strong assumption that our conservative utility-based measure of quality adjustment for our three conditions studied with the Medicare data is representative of the hospital sector more broadly, the authors apply the adjustment to the output price index. We find that it implies a multifactor productivity growth rate of 2.8 percent, holding inputs constant.
This paper provides comprehensive evidence that innovations commonly lead to quality-adjusted price declines in the medical care sector. We find that applying the appropriate quality-adjustment methodology is critical for obtaining a meaningful quality-adjusted index. The utility-based COLI price index whose quality adjustment is based on the monetized value of the increase in the health benefits of treatment. Important differences can arise between the utility-based method and other indexes when the marginal valuation of life differs from the average price per unit of health produced. These differences are found to be of great empirical importance for the thousands of cost-effectiveness studies in the Tuft’s CEAR database and for the three acute conditions studied using Medicare claims.
While the authors are able to show that there may be substantial quality-adjusted price declines from new innovation, more work is needed to incorporate this information into annual disease-based price indexes. It will be important for academic researchers and statistical agencies to continue research to build a consensus around quality adjustment methods that may be applied to the health care sector more broadly. Until a consensus is formed, it may be important to report a range of estimates for the quality-adjusted prices, rather than applying a single method or set of assumptions. There is considerable promise for further development of quality-adjusted price indexes for medical conditions as measurements of quality of life are improved, more detailed data become available, and valuations of health become more certain.
The authors did not receive financial support from any firm or person for this article or from any firm or person with a financial or political interest in this article. None is currently an officer, director, or board member of any organization with an interest in this article.
Report Produced by The Hutchins Center on Fiscal and Monetary Policy