Medical care price indexes: theory, construction & empirical analysis of the US series 1927-1990

Abstract

The historical development of price indexes as wage adjustment mechanisms is reviewed, as is the theory of aggregation and methods for dealing with quality and technological change. The construction of the U.S. Bureau of Labor Statistics (BLS) Medical Care Price Index (MCPI) is detailed. ARIMA analysis of the MCPI for the period 1927-1990 indicates that; (i) the MCPI is largely a damped and delayed function of the CPI, with an average lag of 8 months; (ii) medical care prices rose 2-4 percent faster than the all-items CPI since 1950, but not for 1927-1950; (iii) health expenditures are affected primarily by the general CPI, with little independent effect of specifically medical prices. The MCPI is a reliable measure of changes in consumer prices with strong construct validity. However, it was not designed for use as a deflator of medical expenditures, and is misleading when erroneously employed in that unintended role. The price/quantity duality and linear expenditure function which form the basis of Laspeyres price indexes are not applicable to nonconcatenable goods such as insurance or medical care. In these complex transactions, quality dominates quantity, fixed prices are replaced by reimbursement and professional judgement, and the assumption of additive separability required to use the price index as a deflator of health expenditures is not valid.