that ratio, it yielded an estimate of added expense that the presence of residents contributed to the cost of admitting a patient to the hospital. In 1980, Medicare began paying an “indirect medical education adjustment” based on those figures. The resident-to-bed ratio continues in use today, although subsequent study indicates the adjustment it specifies is related less to medical education expenses than it is to patient services. Medicare's handling of graduate education costs is discussed in the Task Force report, "The Future Financing of Teaching Hospitals." Other Payers-Insurers such as Blue Cross historically have paid for graduate medical education by allowing its direct and indirect costs to be included with the hospital's other costs, as Medicare used to do. On the other hand, Medicaid, the state-federal program to care for the poor, reimburses only a portion of hospital costs. A survey of teaching hospitals (Hadley and Tigue, 1982) found that 5.8 percent of them had been turned down by Medicaid on reimbursement for house staff stipends. Commercial health care insurers and other third-party payers who simply remunerate according to hospital charges have given scant attention to costs of education. That situation is changing as everyone becomes more aware of all contributions to the increasing costs of health care. The Veterans Administration is supporting about 12 percent of all the full-time equivalent house staff positions in the United States, with a graduate medical education budget of about $200 million a year. The size of the VA's education financing would give it a strong voice in specialist training policies were it to seek changes. ! ! ISSUES IN GRADUATE The Issue of Cost The prominence of the issue of cost of graduate medical education is an outgrowth of public concern about costs in the entire health care enterprise The concern over education costs may be misplaced, considering that the largest estimate of yearly education costs is hardly two percent of the $40C billion spent nationally for health. However, there is a growing unwillingness simply to accept the cost of education—or the cost of anything else, for that matter-as an unidentified part of the overall hospital bill. Itemúzing the bill may show where costs can be controlled. Such teasing apart o the costs gives a new visibility to the funding of graduate medical education and portends changes. Although Medicare is now paying explicitly for the direct and indirec costs of education, potential alterations are in the offing. The Social Seev rity Advisory Council has recommended that other payment sources be identified by 1987, a move that will, it is hoped, save the Medicare Tius! Fund some $41 billion by 1995 (Aiken and Bays, 1984). Also, the Sena Finance Committee's subcommittee on health is proposing to put limits the Medicare pass-through of education's direct costs. As these nationwide changes are being debated, pressures also are becoming evident at state and local levels: • State budget cuts prompted the University of Washington to reduc state-supported residency slots by 31 percent. • In each of the past three years, California legislators have reduced funding for all graduate-level medical education except primary c residencies. • Blue Cross and Blue Shield of Maryland have instituted a "Selectprogram that directs patients to hospitals that charge less, thus bypassing some hospitals where costs are boosted by education Care" program expenses. Employers seeking to control costs of health benefits are inducing employees to get their care from low-bid providers, which constitutes a refusal to pay the costs of education. The Issue of Size The size of graduate medical education programs has been determined largely by the demands of newly-graduated medical students and the service needs of teaching hospitals. Individual institutions, either hospitals or medical schools, have not taken responsibility for producing the amounts and types of physicians to match the nation's medical manpower needs (Tarlov, 1983). Some observers are predicting an oversupply of physicians in the United States before the turn of the century, but projections that were stated in firm numbers hardly more than a year ago now are couched in less certain terms. Reasons for uncertainty include a shrinking number of students enrolling in medical schools, which will mean fewer new M.D.s seeking residencies a few years from now. In any event, the planners of graduate medical education programs in the future will have to take into account not only the costs of those programs but also the obligations that will result from the number of physicians trained. The Issue of Control The size and design of graduate inedical education programs are influenced, to a greater or lesser extent, by many agents, as was described in detail earlier: the Accreditation Council on Graduate Medical Education and its five parent organizations, 24 residency review committees, 23 specialty boards, 24 specialty societies, hospital directors, medical school deans, program directors, training directors, faculty and house staff. Some of these can be grouped coherently, as shown in Figure 1. The ACGME and the RRCs, for instance, are responsible for accrediting residency programs, whose size and operating characteristics are affected by the hierarchy of people in the teaching hospital, from director to house officer. Playing in a separate ring are the specialty boards, which set up educational criteria for certification of physicians and do the certifying, but have little say in the rest of the action. In yet another ring are the specialty societies, some of which nominate members to some RRCs and some specialty boards-and some of which do not. The main interest of the specialty societies is that graduate medical education produce highly skilled physicians. The lack of dependable and constructive relationships among the players in the three rings is readily apparent. But something else is missing: None of the players is in the chain of funding for graduate medical education. This funding is anchored in the reimbursement of teaching hospitals for patient care. The teaching hospitals have the responsibility of providing graduate education but have little influence on the decisions made elsewhere in the system that determine costs. Two problems arise when separate players can make decisions independently of each other and apart from considerations of funding. One is that, |