Medical education is certainly an important and worthy objective. American consumers of all ages benefit from having highly trained and skilled medical professionals, including physicians, nurses, and allied health professionals. Medical education is, and has the potential to become, an important export sector of the U.S. economy as we train medical professionals from all over the world. But the public policy issue today is not the value of medical education, but who should bear the burden of financing this education.

The health and safety of consumers depend on many types of education. We need well-trained engineers to design and build safe bridges and automobiles, computer programmers to write programs that safely control Metro cars and the air traffic system, and trained individuals to distribute and process our food supply to protect us from contaminated food. Just because an area of education is important to our health and well-being does not mean it requires a public subsidy or a federal trust fund to bring it about. It is not required that the Congress pass a special tax on automobiles to assure that there is an adequate supply of engineers.

In other words, I explicitly disagree with those who argue that academic medical training requires public subsidies to produce the optimal amount of medical training, research, or patient care. There may be a public goods rationale for the public support of basic science, the type we now fund through the National Institutes of Health (NIH) and the National Science Foundation (NSF), but applying the public goods argument as a rationale for Medicare funding to academic health centers and teaching hospitals is misleading as a guide to public policy.

Currently, we now have a system that is characterized by an open-ended entitlement program financed directly from the Part A trust fund. This "entitlement" goes t o large teaching institutions rather than individuals. It exacerbates the overall financial problems of Medicare and imposes a set of inefficient economic incentives on almost all of the players in medical markets.

There are more efficient ways to subsidize the "social missions" of academic health centers. Several rationales are given to justify Medicare subsidies for graduate medical education (GME) including medical training, research, and patient care. Each of these objectives has some benefit for some people, so the issue is not the desire to bring about worthwhile outcomes, but what is the most efficient way to achieve that objective.

GME funding probably has some positive effect on each of these objectives, but it is exceedingly difficult to identify and measure these effects. The money flows from Medicare to medical teaching institutions, but measuring the effects in terms of the quantity and quality of the medical training, the specific results of research, or the amount and kind of patient care for the poor is impossible to determine. In effect, Medicare puts the money, currently about $7 billion per year, on the stump and relies on the preferences of the management of each medical institution to decide how to spend the money. Some of the managers of the approximately 1,200 institutions that receive Medicare payments may spend some of this money on additional research or care for the poor, but there is no assurance that the money will be spent for these purposes.

A new study by Sean Nicholson at the Wharton School entitled, "The Unintended Consequences of Subsidizing Teaching Hospitals," studies the effects of Medicare GME subsidies using data on 843 teaching hospitals. In this ongoing study, Nicholson did not have the data to determine if individual teaching hospitals responded to the GME subsidies by increasing their research or charity care. He does find evidence that the subsidies have had little effect on hospitals’ use of inputs, such as number of residents or beds. My conclusion from his study is that the teaching hospitals are indeed using the money for unintended and unidentifiable purposes. With no definitive proof that Medicare’s GME subsidies actually affect research or charity care, neither the Commission nor the Congress should take the stated rationales at face value.

For each of the present rationales, there is a more efficient approach to achieving the objective than is provided by the present system of granting GME payments based on historic costs and Medicare volume formulas. With respect to medical training, a medical training voucher could be made available to qualified medical graduates to be used at any of a set of approved teaching hospitals or other medical institutions. This would give teaching institutions strong incentives to compete for medical students by adjusting their curricula and tuition costs to meet the demands of students and to be concerned more about the cost and effectiveness of their medical training (rather than using residents as a source of cheap labor).

The usual objection to a voucher system is that it would be too complicated to set up and run. In my view, it would be no more complicated that the present system, and it would provide for a much more efficient system of medical education. Congress could establish income (or any other criteria) for eligibility for the vouchers if it wanted to ensure that poor students had greater opportunities for medical careers. We now have a system that spends approximately $7 billion per year to subsidize the training of a group of professionals that become some of the wealthiest individuals in our society.

With regard to research, we already have systems for sponsoring medical research in teaching hospitals and medical schools through the peer-reviewed systems at the NIH and the NSF, among others. These systems could be utilized to expand any line of research that the Congress felt would be beneficial to Medicare or other patients. Such research mechanisms could be held more accountable than through the present system.

And finally, with respect to patient care, a more direct way to ensure access by the poor or presently uninsured would be to provide vouchers or refundable tax credits that could be used by individuals to buy health insurance or enroll in a health plan.

The present system of providing Medicare subsidies for medical training is a hangover from a previous time. It is no longer in the best interests of either Medicare recipients or taxpayers to continue this system. There is no evidence that it is producing the results that are claimed for it, and it is not affordable given Medicare’s financial condition. To the extent that Congress wishes to promote any of the stated goals of GME funding, in each case a more direct approach would achieve the objective more efficiently. The present system of subsidizing GME through an open-ended entitlement program is expensive and inefficient and denies Congress the opportunity to weigh the objectives of GME funding against other public policy goals.

Robert B. Helms is a resident scholar at AEI.