Despite many cost control measures implemented in the 1980's, Medicare outlays to physicians continued to grow at double-digit rates. With the passage of the Omnibus Budget Reconciliation Act of 1989 (OBRA89), the Congress approved a new system of paying physicians. Under the new system, Medicare pays physicians on a free schedule based on resource costs instead of the much-maligned customary, prevailing and reasonable payment system. OBRA89 and OBRA90 have reduced Medicare fees for 245 """"""""overvalued"""""""" procedures by as much as over 30 percent. the fee reductions were substantial in magnitude and varied across both procedures and geographic areas. As a result, Medicare fee reductions create an excellent opportunity for studying physician response to such exogenous fee changes along with multiple specialty groups and geographic comparison groups. This situation is similar to a """"""""natural experiment"""""""" where a methodologically strong estimate of volume response can be made. Due to the lack of a generally agreed upon theoretical model and appropriate data sources, none of the previous studies could explicitly reveal the magnitude of income and substitution effects of a fee change in a multi-payer environment. The objective of the proposed study is to examine the magnitude and the determinants of volume response, not only for Medicare services, but for those services provided to privately insured patients as well. There are two specific aims of the proposed study: (1) to examine the impact of Medicare fee reductions on physician behavior; and (2) to test the model of physician behavior in response to fee change in a multi- payer system formulated by McGuire and Pauly with data from both Medicare and private payer sources. The proposed study will use data sources from both the Medicare program and private insurers that have not been made available before. The analytical data file consists of (1) detailed discharge data from approximately 250 hospitals provided by the Commission on Professional and Hospital Activities (CPHA); (2) physician pricing data file provided by MEDSTAT Systems, Inc. which contains summary claims records from over 100 different insurance companies other than Medicare and Medicaid; (3) information regarding Medicare fee reductions and the price charged for physician services by the Physician Payment Review Committee (PPRC) compiled from the BMAD file; (4) American Hospital Association's Annual Survey of Hospitals; and (5) the Area Resource File. There will be 45 months worth of data covering a two-year period before OBRA89, a 9 months period after OBRA89 and a 12 months period after OBRA90. As the study uses pooled cross-sectional observations over time, a fixed-effects or random-effects model will be used to test the magnitude of the income and substitution effects.
