Friedman, M on registration, certification and licensure

Chapter IX

Occupational Licensure

The overthrow of the medieval guild system was an indispensable early step in the rise of freedom in the Western world. It was a sign of the triumph of liberal ideas, and widely recognized as such, that by the mid-nineteenth century, in Britain, the United States, and to a lesser extent on the continent of Europe, men could pursue whatever trade or occupation they wished without the by-your-leave of any governmental or quasi-governmental authority. In more recent decades, there has been a retrogression, an increasing tendency for particular occupations to be restricted to individuals licensed to practice them by the state.

These restrictions on the freedom of individuals to use their resources as they wish are important in their own right. In addition, they provide still a different class of problems to which we can apply the principles developed in the first two chapters.

I shall discuss first the general problem and then a particular example, restrictions on the practice of medicine. The reason for choosing medicine is that it seems desirable to discuss the restrictions for which the strongest case can be made -- there is not much to be learned by knocking down straw men. I suspect that most people, possibly even most liberals, believe that it is desirable to restrict the practice of medicine to people who are licensed by the state. I agree that the case for licensure is stronger for medicine than for most other fields. Yet the conclusions I shall reach are that liberal principles do not justify licensure even in medicine and that in practice the results of state licensure in medicine have been undesirable.

UBIQUITY OF GOVERNMENTAL RESTRICTIONS
ON ECONOMIC ACTIVITIES MEN MAY ENGAGE IN

Licensure is a special case of a much more general and exceedingly widespread phenomenon, namely, edicts that individuals may not engage in particular economic activities except under conditions laid down by a constituted authority of the state. Medieval guilds were a particular example of an explicit system for specifying which individuals should be permitted to follow particular pursuits. The Indian caste system is another example. To a considerable extent in the caste system, to a lesser extent in the guilds, the restrictions were enforced by general social customs rather than explicitly by government.

A widespread notion about the caste system is that every person's occupation is completely determined by the caste into which he is born. It is obvious to an economist that this is an impossible system, since it prescribes a rigid distribution of persons among occupations determined entirely by birthrates and not at all by conditions of demand. Of course, this is not the way the system worked. What was true, and to some measure still is, was that a limited number of occupations were reserved to members of certain castes, but not every member of those castes followed those occupations. There were some general occupations, such as general agricultural work, which members of various castes might engage in. These permitted an adjustment of the supply of people in different occupations to the demand for their services.

Currently, tariffs, fair-trade laws, import quotas, production quotas, trade union restrictions on employment and so on are examples of similar phenomena. In all these cases, governmental authority determines the conditions under which particular individuals can engage in particular activities, which is to say, the terms on which some individuals are permitted to make arrangements with others. The common feature of these examples, as well as of licensure, is that the legislation is enacted on behalf of a producer group. For licensure, the producer group is generally a craft. For the other examples, it may be a group producing a particular product which wants a tariff, a group of small retailers who would like to be protected from competition by the "chiseling" chain stores, or a group of oil producers, of farmers, or of steel workers.

Occupational licensure is by now very widespread. According to Walter Gellhorn, who has written the best brief survey I know, "By 1952 more than 80 separate occupations exclusive of 'owner-businesses,' like restaurants and taxicab companies, had been licensed by state law; and in addition to the state laws there are municipal ordinances in abundance, not to mention the federal statutes that require the licensing of such diverse occupations as radio operators and stockyard commission agents. As long ago as 1938 a single state, North Carolina, had extended its law to 60 occupations. One may not be surprised to learn that pharmacists, accountants, and dentists have been reached by state law as have sanitarians and psychologists, assayers and architects, veterinarians and librarians. But with what joy of discovery does one learn about the licensing of threshing machine operators and dealers in scrap tobacco? What of egg graders and guide dog trainers, pest controllers and yacht salesmen, tree surgeons and well diggers, tile layers and potato growers? And what of the hypertrichologists who are licensed in Connecticut, where they remove excessive and unsightly hair with the solemnity appropriate to their high sounding title?"¹ In the arguments that seek to persuade legislatures to enact such licensure provisions, the justification is always said to be the necessity of protecting the public interest. However, the pressure on the legislature to license an occupation rarely comes from the members of the public who have been mulcted or in other ways abused by members of the occupation. On the contrary, the pressure invariably comes from members of the occupation itself. Of course, they are more aware than others of how much they exploit the customer and so perhaps they can lay claim to expert knowledge. Similarly, the arrangements made for licensure almost invariably involve control by members of the occupation which is to be licensed. Again, this is in some ways quite natural. If the occupation of plumbing is to be restricted to those who have the requisite capacity and skills to provide good service for their customers, clearly only plumbers are capable of judging who should be licensed. Consequently, the board or other body that grants licenses is almost invariably made up largely of plumbers or pharmacists or physicians or whatever may be the particular occupation licensed.

Gellhorn points out that "Seventy-five per cent of the occupational licensing boards at work in this country today are composed exclusively of licensed practitioners in the respective occupations. These men and women, most of whom are only part-time offiicials, may have a direct economic interest in many of the decisions they make concerning admission requirements and the definition of standards to be observed by licensees. More importantly, they are as a rule directly representative of organized groups within the occupations. Ordinarily they are nominated by these groups as a step toward a gubernatorial or other appointment that is frequently a mere formality. Often the formality is dispensed with entirely, appointment being made directly by the occupational association -- as happens, for example, with the embalmers in North Carolina, the dentists in Alabama, the psychologists in Virginia, the physicians in Maryland, and the attorneys in Washington."²

Licensure therefore frequently establishes essentially the medieval guild kind of regulation in which the state assigns power to the members of the profession. In practice, the considerations taken into account in determining who shall get a license often involve matters that, so far as a layman can see, have no relation whatsoever to professional competence. This is not surprising. If a few individuals are going to decide whether other individuals may pursue an occupation, all sorts of irrelevant considerations are likely to enter. Just what the irrelevant considerations will be, will depend on the personalities of the members of the licensing board and the mood of the time. Gellhorn notes the extent to which a loyalty oath was required of various occupations when the fear of communist subversion was sweeping the country. He writes, "A Texas statute of I952 requires each applicant for a pharmacist's license to swear that 'he is not a member of the Communist Party or affiliated with such party, and that he does not believe in and is neither a member of nor supports any group or organization that believes in, furthers or teaches the overthrow of the United States Government by force or any illegal or unconstitutional methods.' The relationship between this oath on the one hand and, on the other, the public health which is the interest purportedly protected by the licensing of pharmacists, is somewhat obscure. No more apparent is the justification for requiring professional boxers and wrestlers in Indiana to swear that they are not subversive . . . A junior high school teacher of music, having been forced to resign after being identified as a Communist, had difficulty becoming a piano tuner in the District of Columbia because, forsooth, he was 'under Communist discipline.' Veterinarians in the state of Washington may not minister to an ailing cow or cat unless they have first si gned a non-Communist oath."³

Whatever one's attitude towards communism, any relationship between the requirements imposed and the qualities which the licensure is intended to assure is rather far-fetched. The extent to which such requirements go is sometimes little short of ludicrous. A few more quotations from Gellhorn may provide a touch of comic relief.⁴

One of the most amusing sets of regulations is that laid down for barbers, a trade that is licensed in many places. Here is an example from a law which was declared invalid by Maryland courts, though similar language can be found in statutes of other states which were declared legal. "The court was depressed rather than impressed by a legislative command that neophyte barbers must receive formal instruction in the 'scientific fundamentals for barbering, hygiene, bacteriology, histology of the hair, skin, nails, muscles and nerves, structure of the head, face and neck, elementary chemistry relating to sterilization and antiseptics, disease of the skin, hair, glands and nails, haircutting, shaving and arranging, dressing, coloring, bleaching, and tinting of the hair'."⁵ One more quotation on the barbers: "Of eighteen representative states included in a study of barbering regulations in 1929, not one then commanded an aspirant to be a graduate of a 'barber college,' though apprenticeship was necessary in all. Today, the states typically insist upon graduation from a barbaring school that provides not less (and often much more) than one thousand hours of instruction in 'theoretical subjects' such as sterilization of instruments, and thus must still be followed by apprenticeship."⁶ I trust these quotations make it clear that the problem of licensing of occupations is something more than a trivial illustration of the problem of state intervention, that it is already in this country a serious infringement on the freedom of individuals to pursue activities of their own choice, and that it threatens to become a much more serious one with the continual pressure upon legislatures to extend it.

Before discussing the advantages and disadvantages of licensing, it is worth noting why we have it and what general political problem is revealed by the tendency for such special legislation to be enacted. The declaration by a large number of different state legislatures that barbers must be approved by a committee of other barbers is hardly persuasive evidence that there is in fact a public interest in having such legislation. Surely the explanation is different; it is that a producer group tends to be more concentrated politically than a consumer group. This is an obvious point often made and yet one whose importance cannot be overstressed.⁷ Each of us is a producer and also a consumer. However, we are much more specialized and devote a much larger fraction of our attention to our activity as a producer than as a consumer. We consume literally thousands if not millions of items. The result is that people in the same trade, like barbers or physicians, all have an intense interest in the specific problems of this trade and are willing to devote considerable energy to doing something about them. On the other hand, those of us who use barbers at all, get barbered infrequently and spend only a minor fraction of our income in barber shops. Our interest is casual. Hardly any of us are willing to devote much time going to the legislature in order to testify against the iniquity of restricting the practice of barbering. The same point holds for tariffs. The groups that think they have a special interest in particular tariffs are concentrated groups to whom the issue makes a great deal of difference. The public interest is widely dispersed. In consequence, in the absence of any general arrangements to offset the pressure of special interests, producer groups will invariably have a much stronger influence on legislative action and the powers that be than will the diverse, widely spread consumer interest. Indeed from this point of view, the puzzle is not why we have so many silly licensure laws, but why we don't have far more. The puzzle is how we ever succeeded in getting the relative freedom from government controls over the productive activities of individuals that we have had and still have in this country, and that other countries have had as well.

The only way that I can see to offset special producer groups is to establish a general presumption against the state undertaking certain kinds of activities. Only if there is a general recognition that governmental activities should be severely limited with respect to a class of cases, can the burden of proof be put strongly enough on those who would depart from this general presumption to give a reasonable hope of limiting the spread of special measures to further special interests. This point is one we have adverted to time and again. It is of a piece with the argument for the Bill of Rights and for a rule to govern monetary policy and fiscal policy.

POLICY ISSUES RAISED BY LICENSURE

It is important to distinguish three different levels of control: first, registration; second, certification; third, licensing.

By registration, I mean an arrangement under which individuals are required to list their names in some official register if they engage in certain kinds of activities. There is no provision for denying the right to engage in the activity to anyone who is willing to list his name. He may be charged a fee, either as a registration fee or as a scheme of taxation.

The second level is certification. The governmental agency may certify that an individual has certain skills but may not prevent, in any way, the practice of any occupation using these skills by people who do not have such a certificate. One example is accountancy, in most states, anybody can be an accountant, whether he is a certified public accountant or not, but only those people who have passed a particular test can put the title CPA after their names or can put a sign in their offices saying they are certified public accountants. Certification is frequently only an intermediate stage. In many states, there has been a tendency to restrict an increasing range of activities to certified public accountants. With respect to such activities there is licensure, not certification. In some states, "architect" is a title which can be used only by those who have passed a specified examination. This is certfication. It does not prevent anyone else from going into the business of advising people for a fee how to build houses.

The third stage is licensing proper . This is an arrangement under which one must obtain a license from a recognized authority in order to engage in the occupation. The license is more than a formality. It requires some demonstration of competence or the meeting of some tests ostensibly designed to insure competence, and anyone who does not have a license is not authorized to practice and is subject to a fine or a jail sentence if he does engage in practice.

The question I want to consider is this: under what circumstances, if any, can we justify the one or the other of these steps? There are three different grounds on which it seems to me registration can be justified consistently with liberal principles.

First, it may assist in the pursuit of other aims. Let me illustrate. The police are often concerned with acts of violence. After the event, it is desirable to find out who had access to firearms. Before the event, it is desirable to prevent firearms from getting into the hands of people who are likely to use them for criminal purposes. It may assist in the pursuit of this aim to register stores selling firearms. Of course, if I may revert to a point made several times in earlier chapters, it is never enough to say that there might be a justification along these lines, in order to conclude that there is justification. It is necessary to set up a balance sheet of the advantages and disadvantages in the light of liberal principles. All I am now saying is that this consideration might in some cases justify overriding the general presumption against requiring the registration of people.

Second, registration is sometimes a device to facilitate taxation and nothing more. The questions at issue then become whether the particular tax is an appropriate method to raise revenue for financing government services regarded as necessary, and whether registration facilitates the collection of taxes. It may do so either because a tax is imposed on the person who registers, or because the person who registers is used as a tax collector. For example, in collecting a sales tax imposed on various items of consumption, it is necessary to have a register or list of all the places selling goods subject to the tax.

Third, and this is the one possible justification for registration which is close to our main interest, registration may be a means to protect consumers against fraud. In general, liberal principles assign to the state the power to enforce contracts, and fraud involves the violation of a contract. It is, of course, dubious that one should go very far to protect in advance against fraud because of the interference with voluntary contracts involved in doing so. But I do not think that one can rule out on grounds of principle the possibility that there may be certain activities that are so likely to give rise to fraud as to render it desirable to have in advance a list of people known to be pursuing this activity. Perhaps one example along these lines is the registration of taxicab drivers. A taxicab driver picking up a person at night may be in a particularly good position to steal from him. To inhibit such practices, it may be desirable to have a list of names of people who are engaged in the taxicab business, to give each a number, and to require that this number be put in the cab so that anyone molested need only remember the number of the cab. This involves simply the use of the police power to protect individuals against violence on the part of other individuals and may be the most convenient method of doing so.

Certification is much more difficult to justify. The reason is that this is something the private market generally can do for itself. This problem is the same for products as for people's services. There are private certification agencies in many areas that certify the competence of a person or the quality of a particular product. The Good Housekeeping seal is a private certification arrangement. For industrial products there are private testing laboratories that will certify to the quality of a particular product. For consumer products, there are consumer testing agencies of which Consumer's Union and Consumer's Research are the best known in the United States. Better Business Bureaus are voluntary organizations that certify the quality of particular dealers. Technical schools, colleges, and universities certify the quality of their graduates. One function of retailers and department stores is to certify the quality of the many items they sell. The consumer develops confidence in the store, and the store in turn has an incentive to earn this confidence by investigating the quality of the items it sells.

One can however argue that in some cases, or perhaps even in many, voluntary certification will not be carried as far as individuals would be willing to pay for carrying it because of the difficulty of keeping the certification confidential. The issue is essentially the one involved in patents and copyrights, namely, whether individuals are in a position to capture the value of the services that they render to others. If I go into the business of certifying people, there may be no efficient way in which I can require you to pay for my certification. If I sell my certification information to one person, how can I keep him from passing it on to others? Consequently, it may not be possible to get effective voluntary exchange with respect to certification, even though this is a service that people would be willing to pay for if they had to. One way to get around this problem, as we get around other kinds of neighborhood effects, is to have governmental certification.

Another possible justification for certification is on monopoly grounds. There are some technical monopoly aspects to certification, since the cost of making a certification is largely independent of the number of people to whom the information is transmitted. However, it is by no means clear that monopoly is inevitable.

Licensure seems to me still more difficult to justify. It goes still farther in the direction of trenching upon the rights of individuals-to enter into voluntary contracts. Nonetheless, there are some justifications given for licensure that the liberal will have to recognize as within his own conception of appropriate government action, though, as always, the advantages have to be weighed against the disadvantages. The main argument that is relevant to a liberal is the existence of neighborhood effects. The simplest and most obvious example is the "incompetent" physician who produces an epidemic. Insofar as he harms only his patient, that is simply a question of voluntary contract and exchange between the patient and his physician. On this score, there is no ground for intervention. However, it can be argued that if the physician treats his patient badly, he may unleash an epidemic that will cause harm to third parties who are not involved in the immediate transaction. In such a case, it is conceivable that everybody, including even the potential patient and physician, would be willing to submit to the restriction of the practice of medicine to "competent" people in order to prevent such epidemics from occurring.

In practice, the major argument given for licensure by its proponents is not this one, which has some appeal to a liberal, but rather a strictly paternalistic argument that has little or no appeal. Individuals, it is said, are incapable of choosing their own servants adequately, their own physician or plumber or barber. In order for a man to choose a physician intelligently, he would have to be a physician himself. Most of us, it is said, are therefore incompetent and we must be protected against our own ignorance. This amounts to saying that we in our capacity as voters must protect ourselves in our capacity as consumers against our own ignorance, by seeing to it that people are not served by incompetent physicians or plumbers or barbers.

So far, I have been listing the arguments for registration, certification, and licensing. In all three cases, it is clear that there are also strong social costs to be set against any of these advantages. Some of these social costs have already been suggested and I shall illustrate them in more detail for medicine, but it may be worth recording them here in general form.

The most obvious social cost is that any one of these measures, whether it be registration, certification, or licensure, almost inevitably becomes a tool in the hands of a special producer group to obtain a monopoly position at the expense of the rest of the public. There is no way to avoid this result. One can devise one or another set of procedural controls designed to avert this outcome, but none is likely to overcome the problem that arises out of the greater concentration of producer than of consumer interest. The people who are most concerned with any such arrangement, who will press most for its enforcement and be most concerned with its administration, will be the people in the particular occupation or trade involved. They will inevitably press for the extension of registration to certification and of certification to licensure. Once licensure is attained, the people who might develop an interest in undermining the regulations are kept from exerting their influence. They don't get a license, must therefore go into other occupations, and will lose interest. The result is invariably control over entry by members of the occupation itself and hence the establishment of a monopoly position.

Certification is much less harmful in this respect. If the certified "abuse" their special certificates; if, in certifying newcomers, members of the trade impose unnecessarily stringent requirements and reduce the number of practitioners too much, the price differential between certified and non-certified will become sufficiently large to induce the public to use non-certified practitioners. In technical terms, the elasticity of demand for the services of certified practitioners will be fairly large, and the limits within which they can exploit the rest of the public by taking advantage of their special position will be rather narrow.

In consequence, certification without licensure is a half-way house that maintains a good deal of protection against monopolization. It also has its disadvantages, but it is worth noting that the usual arguments for licensure, and in particular the paternalistic arguments, are satisfied almost entirely by certification alone. If the argument is that we are too ignorant to judge good practitioners, all that is needed is to make the relevant information available. If, in full knowledge, we still want to go to someone who is not certified, that is our business; we cannot complain that we did not have the information. Since arguments for licensure made by people who are not members of the occupation can be satisfied so fully by certification, I personally find it difficult to see any case for which licensure rather than certification can be justified.

Even registration has significant social costs. It is an important first step in the direction of a system in which every individual has to carry an identity card, every individual has to inform authorities what he plans to do before he does it. Moreover, as already noted, registration tends to be the first step toward certification and licensure.

MEDICAL LICENSURE

The medical profession is one in which practice of the profession has for a long time been restricted to people with licenses. Offhand, the question, "Ought we to let incompetent physicians practice?" seems to admit of only a negative answer. But I want to urge that second thought may give pause.

In the first place, licensure the key to the control that the medical profession can exercise over the number of physicians. To understand why this is so requires some discussion of the structure of the medical profession. The American Medical Association is perhaps the strongest trade union in the United States. The essence of the power of a trade union is its power to restrict the number who may engage in a particular occupation. This restriction may be exercised indirectly by being able to enforce a wage rate higher than would otherwise prevail. If such a wage rate can enforced, it will reduce the number of people who can get jobs and thus indirectly the number of people pursuing the occupation. This technique of restriction has disadvantages. There is always a dissatisfied fringe of people who are trying to get into the occupation. A trade union is much better off if it can limit directly the number of people who enter the occupation -- who ever try to get jobs in it. The disgruntled and dissatisfied are excluded at the outset, and the union does not have to worry about them.

The American Medical Association is in this position. It is a trade union that can limit the number of people who can enter. How can it do this? The essential control is at the stage of admission to medical school. The Council on Medical Education and Hospitals of the American Medical Association approves medical schools. In order for a medical school to get and stay on its list of appproved schools it has to meet the standards of the Council. The power of the Council has been demonstrated at various times when there has been pressure to reduce numbers. For example, in the 1930's during the depression, the Council on Medical Education and Hospitals wrote a letter to the various medical schools saying the medical schools were admitting more students than could be given the proper kind of training. In the next year or two, every school reduced the number it was admitting, giving very strong presumptive evidence that the recommendation had some effect.

Why does the Council's approval matter so much? If it abuses its power, why don't unapproved medical schools arise? The answer is that in almost every state in the United States, a person must be licensed to practice medicine, and to get the license, he must be a graduate of an approved school. In almost every state, the list of approved schools is identical with the list of schools approved by the Council on Medical Education and Hospitals of the American Medical Association. That is why the licensure provision is the key to the effective control of admission. It has a dual effect. On the one hand, the members of the licensure commission are always physicians and hence have some control at the step at which men apply for a license. This control is more limited in effectiveness than control at the medical school level. In almost all professions requiring licensure, people may try to get admitted more than once. If a person tries long enough and in enough jurisdictions he is likely to get through sooner or later. Since he has already spent the money and time to get his training, he has a strong incentive to keep trying. Licensure provisions that come into operation only after a man is trained therefore affect entry largely by raising the costs of getting into the occupation, since it may take a longer time to get in and since there is always some uncertainty whether he will succeed. But this rise in cost is nothing like so effective in limiting entry as is preventing a man from getting started on his career. If he is eliminated at the stage of entering medical school, he never comes up as a candidate for examination; he can never be troublesome at that stage. The efficient way to get control over the number in a profession is therefore to get control of entry into professional schools.

Control over admission to medical school and later licensure enables the profession to limit entry in two ways. The obvious one is simply by turning down many applicants. The less obvious, but probably far more important one, is by establishing standards for admission and licensure that make entry so difficult as to discourage young people from ever trying to get admission. Though most state laws require only two years of college prior to medical school, nearly 100 per cent of the entrants have had four years of college. Similarly, medical training proper has been lengthened, particularly through more stringent internship arrangements.

As an aside, the lawyers have never been as successful as the physicians in getting control at the point of admission to professional school, though they are moving in that direction. The reason is amusing. Almost every school on the American Bar Association's list of approved schools is a full time day school; almost no night schools are approved. Many state legislators, on the other hand, are graduates of night law schools. If they voted to restrict admission to the profession to graduates of approved schooIs, in effect they would be voting that they themselves were not qualified. Their reluctance to condemn their own competence has been the main factor that has tended to limit the extent to which law has been able to succeed in imitating medicine. I have not myself done any extensive work on requirements for admission to law for many years but I understand that this limitation is breaking down. The greater affluence of students means that a much larger fraction are going to full time law schools and this is changing the composition of the legislatures.

To return to medicine, it is the provision about graduation from approved schools that is the most important source of professional control over entry. The profession has used this control to limit numbers. To avoid misunderstanding let me emphasize that I am not saying that individual members of the medical profession, the leaders of the medical profession, or the people who are in charge of the Council on Medical Education and Hospitals deliberately go out of their way to limit entry in order to raise their own incomes. That is not the way it works. Even when such people explicitly comment on the desirability of limiting numbers to raise incomes they will always justify the policy on the grounds that if "too" many people are let in, this will lower their incomes so that they will be driven to resort to unethical practices in order to earn a "proper" income. The only way, they argue, in which ethical practices can be maintained is by keeping people at a standard of income which is adequate to the merits and needs of the medical profession. I must confess that this has always seemed to me objectioable on both ethical and factuaLgrounds. It is extraordinary that leaders of medicine should proclaim publicly that they and their colleagues must be paid to be ethical. And if it were so, I doubt that the price would have any limit. There seems little correlation between poverty and honesty. One would rather expect the opposite; dishonesty may not always pay but surely it sometimes does.

Control of entry is explicitly rationalized along these lines only at times like the Great Depression when there is much unemployment and relatively low incomes. In ordinary times, the rationalization for restriction is different. It is that the members of the medical profession want to raise what they regard as the standards of "quality" of the profession. The defect in this rationalization is a common one, and one that is destructive of a proper understanding of the operation of an economic system, namely, the failure to distinguish between technical efficiency and economic efficiency.

A story about lawyers will perhaps illustrate the point. At a meeting of lawyers at which problems of admission were being discussed, a colleague of mine, arguing against restrictive admission standards, used an analogy from the automobile industry. Would it not, he said, be absurd if the automobile industry were to argue that no one should drive a low quality car and therefore that no automobile manufacturer should be permitted to produce a car that did not come up to the Cadillac standard. One member of the audience rose and approved the analogy, saying that, of course, the country cannot afford any thing but Cadillac lawyers! This tends to be the professional attitude. The members look solely at technical standards of performance, and argue in effect that we must have only first-rate physicians even if this means that some people get no medical service -- though of course they never put it that way. Nonetheless, the view that people should get only the "optimum" medical service always lead to a restrictive policy, a policy that keeps down the number of physicians. I would not, of course, want to argue that this is the only force at work, but only that this kind of consideration leads many well-meaning physicians to go along with policies that they would reject out-of-hand if they did not have this kind of comforting rationalization.

It is easy to demonstrate that quality is only a rationalization and not the underlying reason for restriction. The power of the Council on Medical Education and Hospitals of the American Medical Association has been used to limit numbers in ways that cannot possibly have any connection whatsoever with quality. The simplest example is their recommendation to various states that citizenship be made a requirement for the practice of medicine. I find it inconceivable to see how this is relevant to medical performance. A similar requirement that they have tried to impose on occasion is that examination for licensure must be taken in English. A dramatic piece of evidence on the power and potency of the Association as well as on the lack of relation to quality is proved by one figure that I have always found striking. After 1933, when Hitler came to power in Germany, there was a tremendous outflow of professional people from Germany, Austria and so on, including of course, physicians who wanted to practice in the United States. The number of physicians trained abroad who were admitted to practice in the United States in the five years after 1933 was the same as in the five years before. This was clearly not the result of the natural course of events. The threat of these additional physicians led to a stringent tightening of requirements for foreign physicians that imposed extreme costs upon them.

It is clear that licensure is the key to the medical profession's ability to restrict the number of physicians who practice medicine. It is also the key to its ability to restrict technological and organizational changes in the way medicine is conducted. The American Medical Association has been consistently against the practice of group medicine, and against prepaid medical plans. These methods of practice may have good features and bad features, but they are technological innovations that people ought to be free to try out if they wish. There is no basis for saying conclusively that the optimum technical method of organizing medical practice is practice by an independent physician. Maybe it is group practice, maybe it is by corporations. One ought to have a system under which all varieties can be tried.

The American Medical Association has resisted such attempts and has been able effectively to inhibit them. It has been able to do so because licensure has indirectly given it control of admission to practice in hospitals. The Council on Medical Education and Hospitals approves hospitals as well as medical schools. In order for a physician to get admission to practice in an "approved" hopital, he must generally be approved by his county medical association or by the hospital board. Why can't unapproved hospitals be set up? Because under present economic conditions, in order for a hospital to operate it must have a supply of interns. Under most state licensure laws, candidates must have some internship experience to be admitted to practice, and internship must be in an "approved" hospital. The list of "approved" hospitals is generally identical with that of the Council on Medical Education and Hospitals. Consequently, the licensure law gives the profession control over hospitals as well as over schools. This is the key to the AMA's largely successful opposition to various types of group practice. In a few cases, the groups have been able to survive. In the District of Columbia, they succeeded because they were able to bring suit against the American Medical Association under the federal Sherman antitrust laws, and won the suit. In a few other cases, they have succeeded for special reasons. There is, however, no doubt that the tendency toward group practice has been greatly retarded by the AMA's opposition.

It is interesting, and this is an aside, that the medical association is against only one type of group practice, namely, prepaid group practice. The economic reason seems to be that this eliminates the possibility of engaging in discriminatory pricing.⁸

It is clear that licensure has-been at the core of the restriction of entry and that this involves a heavy social cost, both to the individuals who want to practice medicine but are prevented from doing so and to the public deprived of the medical care it wants to buy and is prevented from buying. Let me now ask the question: Does licensure have the good effects that it is said to have?

In the first place, does it really raise standards of competence? It is by no means clear that it does raise the standards of competence in the actual practice of the profession for several reasons. In the first place, whenever you establish a block to entry into any field, you establish an incentive to find ways of getting around it, and of course medicine is no exception. The rise of the professions of osteopathy and of chiropractic is not unrelated to the restriction of entry into medicine. On the contrary, each of these represented, to some extent, an attempt to find a way around restriction of entry. Each of these, in turn, is proceeding to get itself licensed, and to impose restrictions. The effect is to create different levels and kinds of practice, to distinguish between what is called medical practice and substitutes such as osteopathy, chiropractic, faith healing and so on. These alternatives may well be of lower quality than medical practice would have been without the restrictions on entry into medicine.

More generally, if the number of physicians is less than it otherwise would be, and if they are all fully occupied, as they generally are, this means that there is a smaller total of medical practice by trained physicians -- fewer medical man-hours of practice, as it were. The alternative is untrained practice by somebody; it may and in part must be by people who have no professional qualifications at all. Moreover, the situation is much more extreme. If "medical practice" is to be limited to licensed practitioners, it is necessary to define what medical practice is, and featherbedding is not something that is restricted to the railroads. Under the interpretation of the statutes forbidding unauthorized practice of medicine, many things are restricted to licensed physicians that could perfectly well be done by technicians, and other skilled people who do not have a Cadillac medical training. I am not enough of a technician to list the examples at all fully. I only know that those who have looked into the question say that the tendency is to include in "medical practice" a wider and wider range of activities that could perfectly well be performed by technicians. Trained physicians devote a considerable part of their time to things that might well be done by others. The result is to reduce drastically the amount of medical care. The relevant average quality of medical care, if one can at all conceive of the concept, cannot be obtained by simply averaging the quality of care that is given; that would be like judging the effectiveness of a medical treatment by considering only the survivors; one must also allow for the fact that the restrictions reduce the amount of care. The result may well be that the average level of competence in a meaningful sense has been reduced by the restrictions.

Even these comments do not go far enough, because they consider the situation at a point in time and do not allow for changes over time. Advances in any science or field often result from the work of one out of a large number of crackpots and quacks and people who have no standing in the profession. In the medical profession, under present circumstances, it is very difficult to engage in research or experimentation unless you are a member of the profession. If you are a member of the profession and want to stay in good standing in the profession, you are seriously limited in the kind of experimentation you can do. A "faith healer" may be just a quack who is imposing himself on credulous patients, but maybe one in a thousand or in many thousands will produce an important improvement in medicine. There are many different routes to knowledge and learning and the effect of restricting the practice of what is called medicine and defining it as we tend to do to a particular group, who in the main have to conform to the prevailing orthodoxy, is certain to reduce the amount of experimentation that goes on and hence to reduce the rate of growth of knowledge in the area. What is true for the content of medicine is true also for its organization, as has already been suggested. I shall expand further on this point below.

There is still another way in which licensure, and the associated monopoly in the practice of medicine, tend to render standards of practice low. I have already suggested that it renders the average quality of practice low by reducing the number of physicians, by reducing the aggregate-number of hours available from trained physicians for more rather than less important tasks, and by reducing the incentive for research and development. It renders it low also by making it much more difficult for private individuals to collect from physicians for malpractice. One of the protections of the individual citizen against incompetence is protection against fraud and the ability to bring suit in the court against malpractice. Some suits are brought, and physicians complain a great deal about how much they have to pay for malpractice insurance. Yet suits for malpractice are fewer and less successful than they would be were it not for the watchful eye of the medical associations. It is not easy to get a physician to testify against a fellow physician when he faces the sanction of being denied the right to practice in an "approved" hospital. The testimony generally has to come from members of panels set up by medical associations themselves, always, of course, in the alleged interest of the patients.

When these effects are taken into account, I am myself persuaded that licensure has reduced both the quantity and quality of medical practice; that it has reduced the opportunities available to people who would like to be physicians, forcing them to pursue occupations they regard as less attractive; that it has forced the public to pay more for less satifactory medical service, and that it has retarded technological development both in medicine itself and in the organization of medical practice. I conclude that licensure should be eliminated as a requirement as a requirement for the practice of medicine.

When all this is said, many a reader, I suspect, like many a person with whom I have discussed these issues, will say, "But still, how else would I get any evidence on the quality of a physician. Granted all that you say about costs, is not licensure the only way of providing the public with some assurance of at least minimum quality?" The answer is partly that people do not now choose physicians by picking names at random from a list of licensed physicians; partly, that a man's ability to pass an examination twenty or thirty years earlier is hardly assurance of quality now; hence, licensure is not now the main or even a ma)or source of assurance of at least minimum quality. But the major answer is very different. It is that the question itself reveals the tyranny of the status quo and the poverty of our imagination in fields in which we are laymen, and even in those in which we have some competence, by comparison with the fertility of the market. Let me illustrate by speculating on how medicine might have developed and what assurances of quality would have emerged, if the profession had not exerted monopoly power.

Suppose that anyone had been free to practice medicine without restriction except for legal and financial responsibility for any harm done to others through fraud and negligence. I conjecture that the whole development of medicine would have been different. The present market for medical care, hampered as it has been, gives some hints of what the difference would have been. Group practice in conjunction with hospitals would have grown enormously. Instead of individual practice plus large institutional hospitals conducted by governments or eleemosynary institutions, there might have developed medical partnerships or corporations -- medical teams. These would have provided central diangnostic and treatment facilities, including hospital facilities. Some presumably would have been prepaid, combining in one package present hospital insurance, health insurance, and group medical practice. Others would have charged separate fees for separate services. And of course, most might have used both methods of payment.

These medical teams -- department stores of medicine, if you will -- would be intermediaries between the patients and the physician. Being long-lived and immobile, they would have a great interest in establishing a reputation for reliability and quality. For the same reason, consumers would get to know their reputation. They would have the specialized skill to judge the quality of physicians; indeed, they would be the agent of the consumer in doing so, as the department store is now for many a product. In addition, they could organize medical care efficiently, combining medical men of different degrees of skill and training, using technicians with limited training for tasks for which they were suited, and reserving highly skilled and competent specialists for the tasks they alone could perform. The reader can add further flourishes for himself, drawing in part, as I have done, on what now goes on at the leading medical clinics.

Of course, not all medical practice would be done through such teams. Individual private practice would continue, just as the small store with a limited clientele exists alongside the department store, the individual lawyer alongside the great many-partnered firm. Men would establish individual reputations and some patients would prefer the privacy and intimacy of the individual practitioner. Some areas would be too small to be served by medical teams. And so on.

I would not even want to maintain that the medical teams would dominate the field. My aim is only to show by example that there are many alternatives to the present organization of practice. The impossibility of any individual or small group conceiving of all the possibilities, let alone evaluating their merits, is the great argument against central governmental planning and against arrangements such as professional monopolies that limit the possibilities of experimentation. On the other side, the great argument for the market is its tolerance of diversity; its ability to utilize a wide range of special knowledge and capacity. It renders special groups impotent to prevent experimentation and permits the customers and not the producers to decide what will serve the customers best.

Notes 1 Walter Gellhorn, Individual Freedom and Governmental Restraints (Baton Rouge: Louisiana State University Press, 1956). Chapter entitled "The Right to Make a Living," p. 106.

2 Ibid. pp. 140-41.

3 Ibid. pp. 129-30.

4 In fairness to Walter Gellhorn, I should note that he does not share my view that the correct solution to these problems is to abandon licensing. On the contrary, he thinks that while licensing has gone much too far it has some real functions to perform. He suggests procedural reforms and changes that in his view would limit the abuse of licensure arrangements.

5 Ibid. pp. 121-22.

6 Ibid. p. 146.

7 See, for example, Wesley Mitchell's famous article on the "Backward Art of Spending Money," reprinted in his book of essays carrying that title (New York: McGraw-Hill, 1937), pp. 3-19.

8 See Reuben Kessel, "Price Discrimination in Medicine," The Journal of Law and Economics, Vol. I (October, 1958), 20-53.

Milton Friedman, Capitalism and Freedom, 1962.