Empty Bed Blues: Australian Health-Care Policy

FOREWORD

People who expect to pay thousands of dollars to repair their houses do not expect to pay the same money to repair their bodies.  They have been encouraged to expect this service free or for little cost at the point of delivery.

Afraid of crippling expenses or inadequate care, most people insured before health insurance was compulsory.  For various reasons, in spite of heavy subsidies, some did not.  Some of these suffered the consequences of not insuring, others became a cost to the taxpayer.

When a friend pulled an arm off in a tractor accident he asked his surgeon how long he was to be kept in hospital.  Instead of telling him directly his surgeon asked who was paying.  His surgeon was thinking of the effect on my friend's attitude but the answer affected both their attitudes.

Public funding removes the most important incentive not to waste health care resources.  Instead bureaucracy and queueing ration medicine.  Pity the doctor who can be sued by his patient for not carrying out a full diagnostic appraisal and virtually excluded from practice by the health department for "over-servicing".

There is still not an adequate safety-net to protect against the financial disasters of major illnesses while the taxpayer pays for trivial medical problems.  Of course the cost is out of control.

The failure of the publicly-funded system is many-sided.  Public hospitals are plagued with industrial problems, sometimes even in life-threatening situations, and public patients lose their choice of doctors.  The result, as the author says, is a shambles.

Richard Wood provides an economic analysis of health care.  Like all good economics it is concerned with human behaviour in the face of incentives and disincentives.

He builds his argument for economic efficiency from the basic, clear, politically-acceptable objective:

That no Australian resident suffering sickness or injury should ever be forced to forego treatment and care by the lack of means to pay for it.

The problems of the health-care industry are attributed to the pre-eminent position of governments in the provision of hospital care, and laws and regulations which inhibit competition in both the delivery of medicine and insurance.

He sets out reasoned suggestions for reconstructing health care policy.

John Hyde

---

INTRODUCTION

The Australian health system is a shambles.  Public hospitals are being run down for want of funds to carry out the repairs and maintenance needed to keep them functioning properly, and to make advanced medical technology available to doctors and patients who use them.

There is said to be a critical shortage of nurses, and this is blamed for the closing of many hospital beds -- even whole wards -- which has led to the rapid growth of waiting lists for "elective" ⁽¹⁾ surgery.  Industrial trouble, involving everyone from cleaners to surgeons, is commonplace and shows no sign of abating.

The health insurance system, which plays an important part in determining the nature of the health delivery system, has been going through a seemingly interminable series of changes -- to better fit the ideological standpoint of the government of the day, or to try to patch over problems not foreseen in the previous version, or in response to political pressures exerted by one interest group or another.

The system seems open to abuse, and attempts at control not particularly fruitful.  As with tax "avoision", those who seek loopholes to exploit always seem ahead of those whose job it is to close them.

Nor has the compulsory insurance scheme (Medicare) been able to fulfill what was promised of it, explicitly and implicitly, in the early 1970s.  It is true that access to doctors has been made very cheap for all Australians, and free for a significant proportion of the population, through the practice of bulk-billing.  But the cost of consulting a doctor is not really what people have in mind when they are thinking of their desire for protection against unaffordable health expenditures.  After all, even in the absence of insurance it would cost considerably less to see a doctor than to call out a plumber to unblock a sewer, or to have the car serviced, or to repair the washing machine.

The really big health expenditures that most of us have good reason to fear are those which require hospitalisation for anything more than a day or two.  A single day in hospital generates actual costs of the order of $300, and the length of stay is typically around 6 days.  A single occasion of hospitalisation could therefore easily cost $1,800, whereas in the normal course of events, the average person sees a GP about four times per year -- the total cost of which is only about $60-$70, plus whatever medication might be prescribed.  For the compulsory insurance scheme to be considered successful, therefore, it is far more important that it should provide us with adequate cover against the need to be hospitalised.

Medicare, however, does not do this.  It provides cover for treatment in public hospitals only, but -- as has already been noted -- these hospitals seem to be beset by worsening chaos.  It is not much comfort to be insured against a stay in a public hospital if one has to wait several months or more to gain admission, or if the institution is liable to be plunged into another industrial dispute during one's stay, or if the standard of care is likely to be inferior because of nurse shortages.  Add to this the very significant shortcoming that Medicare does not allow the patient his own choice of doctor, nor the option of private room rather than shared ward accommodation, and we are left with very poor quality insurance indeed.

These are the sorts of problems which have provided the stimulus for this study:  surely, we are capable of designing a better health system than the one we have at present.

It needs to be emphasised at the outset that greater expenditure on health does not necessarily make the population more healthy.  The US spends roughly one-third more on a per capita basis than the average for Australia, Canada, France and Germany as a group, but its health, as indicated (very roughly) by life expectancy at birth and infant mortality, is virtually the same.  Sweden spends a little more than the US, and has appreciably better health.  Japan spends less than half what Sweden does, but is just as healthy.  In the UK, the Government-dominated health system spends even less on health than is the case in Japan, and the population seems to be the least healthy of all the countries mentioned here.  Clearly, the problem of achieving better health is very much less one of finding more resources for this purpose than of finding ways to use those resources efficiently. ⁽²⁾

---

CHAPTER 1
THE ROLE OF ECONOMIC ANALYSIS

This paper is grounded in the belief that the insights of economists, and the methods of analysis they use, can and should be brought to bear on questions relating to the design of policies for the health industry -- policies which very largely determine how effectively the industry operates.  Those who feel that "health is different from hamburgers" because of its "social" implications (whatever that might mean) and therefore outside the realm of economics, should, before consigning these pages to the waste basket, consider the following:

• More money spent on health by governments (as is certain to be required if existing arrangements are not substantially modified) is likely to mean less spent on things like education, roads, sewers, water supply, environmental protection, public housing, unemployment benefits, employment generation, law and order, defence, child-minding facilities, pensions, public transport and so on.

  Just about everyone who favours more expenditure on health also favours at least some of these other kinds of expenditure, and I have yet to meet the person who favours higher taxation (except for other people!), so it is important to appreciate that more spent on health implies less for other things held dear.

  This is one reason health policy is clearly within the province of economics:  probably the most fundamental principle of economics is that of "opportunity cost" -- that if more of a given amount of resources is used for one activity, less will then be available for all other activities.

• Whatever social objectives may be considered desirable in the health field, there will always be choice about the ways we try to achieve them.  It is important that the right choice be made:  the one which meets those objectives with least cost of resources of manpower and capital.

  This is the bread and butter of economics:  the pursuit of efficient means of achieving objectives.  Not the least contribution which the economist can make here is to warn against equating means (policies) with ends (objectives).  For example, there is a common tendency to wrongly regard the policy of maintaining the existence of an institution like Medicare as an objective.  The fundamental objective, presumably, is to try to ensure that everyone can have access to health care when it is needed, and there are many policies compatible with this objective, some of which do not require the existence of Medicare.  This paper aims to help find better policies for achieving this kind of social objective (which is stated more precisely below).

• The political reality is that neither of the major parties is prepared to go on making extravagant promises of what they, in government, will do for us in the health field. ⁽³⁾  In years gone by, it seemed good politics in the short term (that is, until the next election) to promise bigger and better health handouts of one kind or another -- subsidised health care of various kinds, higher health insurance benefits and so on.  Now, the longer-term taxation implications of these vote-buying exercises are becoming more and more apparent, and successive governments will be forced to prune health outlays in their struggle to bring huge budget deficits under control.  But poorly-conceived policies of cost-containment in health may cause immense damage to the industry, just like the earlier unthinking growth of government intervention.  Tightening hospital budget allocations which close beds in the face of long waiting lists for admission is but one example.  Economics has a lot to say about getting better health care more cheaply.

  I have emphasised the important role of economists in helping to design policies for the health industry because I believe far too little attention has been given to their views in the past.  Specialised training enables engineers to design bridges, scientists to explore outer space, and doctors to treat illness.  So too the training of economists fits them to talk about how the desire for good health can be balanced against all the other material wants of mankind, given limited physical and financial resources.

  It is a matter for grave concern that the most influential arm of government in the health sphere -- the Commonwealth Department of Health -- appears to have little understanding of the views of economists.  In its recent monograph Health Care and the Consumer, for example, we find the assertion that:  "Economists agree that the market for health services does not behave like other markets" (p.4).  Nothing could be further from the truth.

When I studied Economics at University not long ago, I impressed upon my lecturers that the same basic techniques of analysis could and should be applied to all markets, regardless of whether the good or service being traded was bananas, ice creams, oil, labour, education, health, or whatever.

This is not to say that there are no economists who believe that "health is different", but my guess is that these are a small minority who seem over-represented amongst government policy advisors.  I hope this study will help redress the imbalance in the spectrum of economists' views to which health industry policy-makers are exposed.

CHAPTER 2
THE PROXIMATE CAUSES OF HEALTH INDUSTRY PROBLEMS

Most or all of the problems in our health industry noted above, and others less noticeable, can be attributed to one or more of three factors:  a poor insurance system;  the pre-eminent position of governments (rather than the private sector) as providers of hospital care;  and laws and regulations which inhibit competition in both health care and health insurance.  These factors are discussed in turn below.

1) THE HEALTH INSURANCE SYSTEM

The purpose of all varieties of insurance is to protect the insured against financial loss in the event of some particular occurrence.  Insurance usually does not reduce the aggregate loss incurred by all insured parties combined -- indeed, it may tend to increase it.  Its great contribution is merely to spread these losses around, so that the misfortunes of the few are borne by the many, and thus greatly mitigated.

For most kinds of insurance, there are two significant problems to be faced by the insurer.  One is that, once insurance has been arranged, the insured has less incentive to avoid the occurrence of the event insured against.  We tend to take fewer precautions like locking doors and keeping fire extinguishers close at hand once we have purchased house and contents insurance.  Therefore insurance often has a tendency to increase the likelihood or extent of loss.

The second is that the insurer is often vulnerable to fraud in relation to the insurance contract.  The value of buildings and their contents may be deliberately overstated and then set on fire;  a ship's cargo might be surreptitiously unloaded before the ship itself is scuttled;  a panel-beater or a builder might overcharge for repair work he undertakes if he knows the insurance company is paying (and may split the profit with the insured, in return for the latter's acquiescence);  an unhappily married woman might gain both her freedom and the wherewithal to enjoy it if she is bold enough to arrange her husband's demise.

These problems are of such great importance in the insurance field that a phrase has been coined to describe them:  namely "moral hazard".  Although the term seems fitting enough for the second aspect, the first is really not an issue of morality, but rather, simply a rational reaction to the change in economic circumstances which accompanies purchase of insurance. ⁽⁴⁾  In this paper, the term "moral hazard" will be taken to refer to the second aspect only;  for want of a better term, I shall refer to the first aspect as "weakened avoidance incentive".

In health insurance, the dimensions of the first problem are considerably widened.  Not only is there less incentive to avoid ill-health and accidents, but there is also less incentive to delay seeking medical advice in the hope that the body's natural recuperative powers will prevail.  Moreover, there is little incentive to resist a doctor's advice to take expensive medication, to seek specialist advice, to undergo costly diagnostic testing, to submit to surgery, to be hospitalised for a lengthy period (or at all), to have more consultations, and so on.

It needs to be stressed that the difference here is one of degree, not one of substance.  Keeping the problems of weakened avoidance incentive and moral hazard under control is one of the challenging management tasks which is a distinguishing characteristic of the insurance business -- not only in health insurance, but in most other insurance fields as well. ⁽⁵⁾  Furthermore, the means of control sometimes have much in common in these various fields.  Some of the best known and most relevant to health insurance are as follows:

1. Requirement for the insured to bear a certain proportion (20%, say) of costs incurred

   This is often referred to as "co-insurance".  In the health field, the share borne by the insured is known as the "patient moiety".  With co-insurance, the insured has an incentive both to avoid occurrence of the insured event, and to keep down costs if it does occur.  This incentive is especially strong in the case of large cost items.

2. Requirement for the insured to meet all costs incurred, up to a certain level

   This provides the strongest possible incentive to avoid costs up to the cut-off level (the "excess", or the "front-end deductible" (FED) in the health field), by making them entirely the responsibility of the insured.  This type of control is most effective for small costs, and so it may make good sense to combine co-insurance with front-end deductibility if potential costs are spread continuously over a wide range -- as is the case with health costs.  Such an insurance contract would give the patient the incentive to control costs at both ends of the health expenditure spectrum.

   Front-end deductibility would be highly beneficial in certain kinds of insurance, including health, even in the absence of weakened avoidance incentive and moral hazard.  If a very high proportion of the insured population can be expected to incur some health costs, but relatively few are likely to incur high health costs, then the administrative costs of the insurer can be greatly reduced.  First, the number of claims will be much less.  Second, the average size of those claims which are submitted will be much greater, so if we assume the cost of processing a claim does not vary much with the size of the claim, then administrative costs as a percentage of claims paid will also fall.

   In short, recognising that there are real costs involved in the provision of insurance (that is, the costs of running the insurance organisations themselves), it is clear that front-end deductibility is a way of substantially reducing these costs.  Of course, the insured has less than complete cover, but what is important is that the really big risks are covered.  Small health outlays can be accommodated within the normal household budget, in just the same way that we accommodate to the need to occasionally repair the car or the washing machine, or take the cat to the vet,

3. Offer of discount in return for claims restraint

   Here, the patient who claims relatively little is rewarded with a reduction in his future premiums.  To put it in reverse fashion:  those who claim heavily are penalised.  This variant is often referred to as the "low-claim bonus".

4. Provision of insurance in kind

   This is a much more radical solution to the problems we have been discussing, which involves the insurer in providing health care, rather than merely paying for it.  In one variant -- the Health Maintenance Organisation (HMO) -- health care and health insurance are unified in a single organisation, which guarantees to treat its members' health problems as they arise.  The membership fee is effectively an insurance premium, and clearly such an arrangement makes it in the interests of the (profit-oriented) organisation to keep costs down to the minimum consistent with effective health care.  (There is no incentive to go too far in this direction, because this would only result in the loss of members -- one way or another!)  My understanding is that HMOs are growing very rapidly in importance in some states in the US by comparison to more conventional forms of health insurance.  This may reflect a better ability to keep downward pressure on health care costs, at least in some circumstances.

   In a second variant, the insurance company contracts with health care providers -- in particular, doctors and hospitals -- to treat patients to whom it sells insurance cover.  There may, for example, be an agreed schedule of rates charged for specific medical conditions.  Or, charges may relate to the number and duration of consultations, or the length of hospital stay.

   This second arrangement would be conditional on the doctors and hospitals concerned agreeing to provide adequate documentation on each case (regarding symptoms, tests, diagnosis, treatment, complications, outcome, and so on), to participate in quality assurance programmes, to allow other doctors to give "second opinions" when the insurer thought this justified, and so on.

In short, contracting doctors and hospitals would have to agree to allow the insurer to do the same kinds of things it would do if it were a well-managed HMO to keep costs as low as would be consistent with proper health care.  In turn, by monitoring their performance in respect of both cost and health outcomes, the insurer is then in a position to request improvement on the part of poor performers, and to withhold supply of patients if this is not forthcoming.

Clearly, the extent of both weakened avoidance incentive and moral hazard is greatly curtailed in the case of insurance in kind.  Indeed, the incentive to cost reduction here may be even stronger than when there is no insurance, because it is the health care provider rather than the patient who has the incentive to resist unnecessary expense.  The fact that the former specialises in health care presumably means it is more likely to be able to find ways to economise, consistent with satisfactory health outcomes, than the latter.

To summarise:  problems of weakened avoidance incentive and moral hazard are a characteristic of most kinds of insurance.  One of the attributes of a well-conceived and well-managed insurance scheme in any field, including health, is its ability to keep these problems under control.  Failure means payouts on claims will be higher than otherwise, which in turn means higher premiums and therefore makes insurance less attractive.

How do Australia's health insurance arrangements measure up?  It will be helpful to consider first the compulsory Medicare system, and then to look at voluntary insurance provided by the health funds.

a) Medicare

As far as care provided in public hospitals is concerned (and remembering the patient's loss of choice of doctor), Medicare eschews the use of co-insurance, front-end deductibles and low-claim bonuses.  As far as I am aware, it does not involve itself in monitoring public hospitals' performance with a view to improving their efficiency.  In short, it appears to do nothing to try to offset the problem of weakened avoidance incentive or to reduce inefficiency in public hospitals.

For medical services, Medicare does incorporate a form of co-insurance, but it is quite different from the kind described above.  Rather than reimbursing the patient with a fixed percentage of the doctor's fee, it reimburses a fixed amount, dependent on the type of service provided.  These fixed amounts are listed in the Medical Benefits Schedule (MBS), which is adjusted from time to time within a process of arbitration rather similar to the arbitration of wage claims.

The system is made confusing by first listing the fees which the arbitrator divines as acceptable (by a process which would appear to have no economically rational basis), and then calculating the benefits payable as 85% of these fees, up to a maximum of $10 per occasion of service.  Since the doctor is free to charge whatever he likes, the "acceptable" fee is largely irrelevant;  what matters to the patient -- and thus, indirectly, to the doctor -- is the Medicare benefit.

There are at least four significant weaknesses in these arrangements.  First, if doctors are happy enough to charge an amount equal to the Medicare benefit, the cost to the patient is zero.  In other words, there is no financial disincentive to seeing the doctor except the cost of getting to the doctor's rooms, and it should come as no surprise, therefore, to hear of people -- especially those with lots of spare time on their hands -- wasting doctors' time under such circumstances.  Despite its protestations about the cost to the taxpayer of "over-servicing", the Government continues to encourage "bulk-billing" -- delivering services at zero cost to the patient.

Second, doctors wishing to attract more patients by reducing their fees are limited in the extent they can do so. ⁽⁶⁾  Once their fee falls to the level of the Medicare benefit, the cost to the patient is zero;  any further reduction would be of no consequence to the patient, since the benefit may not exceed the actual fee charged.  Hence it would be impossible to attract new patients in this way.

It is sadly amusing to contrast the extent to which governments go to oppose "fee" (i.e benefit) increases in the arbitration process with the continuing effective refusal to allow doctors to reduce their fees below the Medicare benefit level!  Instead of competing on price, doctors are thus forced to compete on "quality", in the sense of providing longer consultations to create the illusion of more intensive consideration of the patient's problem, providing services at times more convenient to the patient and less to the doctor, furnishing the waiting rooms more lavishly, and so on.

These kinds of things of themselves may well be in the patient's interests, but efficiency demands that there should be freedom for both parties to find the most desirable price/quality combination;  they are stopped from doing so by Medicare.  That this is a real problem is shown by the fact that such a large proportion of GP consultations (which are by far the most important of all medical services in terms of aggregate cost) are bulk-billed -- some 53% in 1984-85. ⁽⁷⁾

Third, again there is no front-end deductible.  This means, for example, that a hypothetical (and by no means atypical) young bachelor earning $20,000 p.a., visiting his GP on average three times a year, and having no other health problems, will incur total medical fees of around $45 -- equivalent to two tanks of petrol for his car -- and have at least 85% of this reimbursed by Medicare.  With all due respect to the well-meaning people who gave us the scheme, this is a travesty of what insurance is really all about -- namely, protection of individuals and families against the unexpected need to incur unaffordably high expenditures in order to maintain or regain good health.

Fourth, not only does the patient moiety fall to zero if the doctor bulk-bills:  it does so as well at the margin for any service rendered at a fee equal to or greater than $67, 15% of which is only just over $10 -- the maximum amount payable by a patient covered by Medicare.  Small wonder that the Government has, for example, resisted private sector installation of Magnetic Image Resonators -- very technologically advanced pieces of diagnostic equipment costing hundreds of dollars per diagnosis to operate. ⁽⁸⁾  The increased incentive to utilise such equipment resulting from artificial reduction in the cost to the patient to only $10 is not difficult to imagine.

To say that Medicare leaves something to be desired, from the point of view of sensible insurance programme design, would therefore be an understatement.  It seems to turn most sound principles of insurance on their head.

Having given patients the incentive to disregard the cost of the use of health resources, the Government has then found doctors convenient scapegoats to blame for the inevitable increase in the amount of health care demanded.  The doctor seems expected to bear in mind the true cost of the medications, tests, specialist consultations and hospitalisations he might recommend, when exercising his professional judgement, while the Government by its own actions encourages people to think of them as virtually costless.  This does seem more than a little unreasonable,

I do not deny that some medical practitioners exploit the system.  The method which has gained most attention has been to bring in large numbers of patients, using the bulk-billing method preferred by the Government, then prescribe extensive diagnostic tests to be carried out by related practices -- also at no cost to the patient, but at considerable cost to Medicare.  Whether or not the tests are actually carried out has sometimes been called into question.

The fee-setting process on which Medicare relies appears to bear much -- perhaps most -- of the responsibility for this state of affairs.  In pathology, in particular, advances in automated testing techniques have allowed the cost of many tests to be dramatically reduced, if the equipment is operated on high volumes.  But the process of fee arbitration -- unlike that of market price adjustment -- has been slow to determine fees more in line with economic reality.

It is hardly surprising that undesirable practices emerge around the opportunities created by the combination of bulk-billing, fees which incorporate huge profit margins for modern, high-volume testing practices, and wide scope for judgement in deciding on the desirability of various diagnostic tests.

In his report on the 1985 Medical Fees Enquiry, Mr Deputy President McKenzie showed his awareness of the issue (pp.29-32), but effectively declined to do anything about it, at least until one of the parties involved provided him with some hard evidence on which to base a determination.

Bearing in mind that moral hazard (which is what is involved here) is an almost inevitable characteristic of most kinds of insurance, it would seem incumbent upon a government which decides to impose a compulsory insurance scheme on the community to design it in such a way as to keep this problem under control.  Having thrown away its most promising means of doing so (in meaningful co-insurance and front-end deducibility), the only remaining avenue of defence of the hapless taxpayer is direct surveillance of medical practitioners.

But even here, the approach adopted is quite unsatisfactory.  The procedure is to collect statistics per practitioner on the extent of his prescriptions of diagnostic tests.  Comparisons are made between practitioners (or practices), highlighting those with a strong propensity to prescribe.  This is useful for providing advice to doctors whose prescribing is excessive compared with their peers, and in this way encouraging doctors who are unknowingly over-prescribing to change their ways, but appears largely useless in preventing wilful over-prescription.

Statistical measures like these are unlikely to stand up in court as sufficient evidence of intent to defraud.  Recall the wide scope for professional judgement relating to health problems:  how could a prosecutor prove that Doctor A did not believe there was some possibility that his patient was suffering Condition B, for which Test C was needed for proper diagnosis?

Perhaps the Health Insurance Commission takes a more optimistic view, but I prefer to be guided by the facts that in the eleven months from July 1984, only twenty-four practitioners had been referred to over-servicing committees, that there had been but thirteen determinations of over-servicing, requiring an amount of $144,000 to be repaid, and that $117,000 of this was still subject to review. ⁽⁹⁾

It may be, of course, that fraud and over-servicing is not the multi-million dollar problem it has been said to be, but if the problem is of that order of magnitude, it seems unlikely to be brought under control using the approaches currently favoured by the Government.

Another serious defect of Medicare's hospital cover is the patient's lack of freedom:  to choose his own doctor, to opt for private room accommodation in public hospitals, and to seek treatment in private hospitals.  This defect could be removed at a stroke, simply by restoring these freedoms.

Patient choice of doctor, roughly speaking, would involve no additional resource cost to the community, since one doctor's time costs much the same as another's (within any given specialty). ⁽¹⁰⁾  The penalty for exercising this choice, however, is at least $85 per day for hospitalisation, plus that part of the doctor's fee not covered by Medicare.

Private room accommodation does involve additional cost, but the Medicare penalty for choosing it -- $160 per day -- appears to be vastly in excess of this.  Total cost of hospitalisation is very roughly of the order of $300 per day, by far the largest component of which is labour cost, not the capital cost of the hospital building.  Besides, there is no reason why people could not be given this option, provided they agree to make supplementary contributions.  Alternatively, they could simply elect to have private accommodation if and when they are hospitalised, and be directly charged for the extra cost at that time.

Freedom to opt for private hospital treatment, on the other hand, would result in substantial savings if private hospitals (or at least, the better ones which patients would be more likely to choose) are more efficiently operated than many of the public hospitals -- a point elaborated later, but on which there seems to be wide agreement in the hospital trade.

Bearing these comments in mind, and remembering the continuing industrial turmoil, nurse shortages, long waiting lists, inadequate maintenance, and reluctance to replace outdated equipment in public hospitals, it is clear that the insurance that has been forced upon the taxpayer is unlikely to be very highly valued by him.

One is left with the conclusion that either:  the Government, having made much of having given us Medicare, does not really want us to use it too much, because it does not want to have to raise the taxes to pay for it;  or the ideology that holds publicly provided health care is better because it is untainted by the quest for profit has won out over the less paternalistic view that people may decide for themselves what kind of health care they want.

Clearly, there is inequity here, in the sense that individuals who feel it important to be able to select the doctor who treats them and their families, who put a high value on privacy, and who believe they might be better looked after in a private hospital, are taxed in order to subsidise those who are prepared to entrust their well-being to a doctor unknown to them, in a shared ward in a public hospital.

It would be just as fair to make it nominally compulsory to drink beer, and to tax everyone in order to make beer freely available.  The compulsion could not be made to work, but there would certainly be a redistribution of income from wine buffs and teetotallers to beer drinkers.  We can be forcibly insured by Medicare, but, so far, we cannot be forced to consume health care we believe to be second-rate.

Unfortunately, there is a substantial cost involved in obtaining the private cover for hospitalisation which many believe to be essential, because the nature of Medicare is such that private cover is not complementary to Medicare cover, but has to be a substitute for it.  If we want to choose our doctor, accommodation and hospital, we must take out full private insurance.  Yet we are still forced to "purchase" Medicare cover -- even though we don't use it -- since it is financed by taxes, not premiums.

b) Private Health Insurance (including Medibank Private)

There are two broad reasons for believing that private health insurance in Australia may not be developed to its full potential.  The first is concerned with the health funds' corporate form, and the second with the nature and extent of regulation to which they are subject.

The funds are not ordinary limited liability companies.  They have no shareholders or shares, and there is therefore no share price to reflect how well they are managed.  There are no profits, just a "surplus on operations" (or a deficit), and these surpluses are added to reserves, to be available to meet future claims.

In such circumstances, economists would expect the incentives to innovate (for example, to develop new products to better suit the preferences of their contributors), to improve internal efficiency, and to monitor the performance of health-care providers (with a view to putting pressure on them so as to reduce the size of contributors' claims), to be less than with a profit-seeking company owned by shareholders.

This is because in the latter case profit and share price are such simple yardsticks, and because there are likely to be people and organisations amongst the shareholders with enough at stake to be willing to invest the time and effort into keeping well-informed about their company, and to push management to continually improve its performance.  Moreover, if the value of the shares falls below the level justified by their earnings potential, this makes the business a takeover prospect, since a major shareholder could buy into the business cheaply, install better management, and guide future development along more profitable lines.

In short, the quest for profit in a competitive environment is a strong motivator to consistently seek excellence in management.  In theory, the incentive to manage efficiently and creatively should be weaker in "non-profit" organisations like health funds.

Here, there is almost no way the "owners" of the funds (that is, their contributors) can directly get rid of poor managers:  the effort involved in actually discovering that management is inadequate, of trying to convince others of this in order to gain support, and then of trying to do something about it, is much too great by comparison with the modest gains this could bring to an individual contributor.

Nor are takeovers by outsiders a possibility.  One can only "buy into" a fund by becoming a contributor, and this does not confer any significant ability to have a major impact on management policy.  Thus, poor management is largely protected against dismissal, since "owners" have virtually no control.  If contributors are unhappy about poor management (which manifests itself mainly in the form of too-high premiums and insurance packages not in tune with contributors' preferences), their only recourse is to take their business elsewhere.

This, of course, does apply some pressure to management to perform well:  if the volume of business falls, managers' career paths and their scope for upgrading their working conditions are constrained.  Some of the less successful funds have in fact gone out of business, yielding their contributors to other funds.  Survival is a prime motivator for all business managers, and there is no reason to suspect that the managers of the funds are any different.  Nevertheless, the particular stimulus to better performance provided by an actual or potential major shareholder is precluded in organisations like Australia's health funds.

A recent reflection of the management of the health funds has been their reaction to the government's decision (in September, 1985) to allow them to offer FED insurance to their contributors.  There are good theoretical grounds for believing that this kind of contract will be preferred by consumers of insurance to complete coverage. ⁽¹¹⁾  I have attempted to set out in fairly plain terms in Appendix 1 why this is so.  I have also tried to show that whether FED packages are preferred in practice depends crucially on how well they are designed -- and in particular, on the level at which the deductible is set.

My impression is that the level of the deductibles seems too low to make them attractive to most people.  Indeed, the funds seem to have largely "followed the leader" (Medibank Private, presumably), since their offerings are very similar.  It would seem also that they are unenthusiastic about FED insurance.  As of late 1985, only 14 of 33 member funds of the Voluntary Health Insurance Association of Australia offered it to their contributors, and amongst those 14, there has been little effort to promote this new product.  One large fund's advertising brochure contains a single line of text advising of the availability of a FED package, but the reader is asked to telephone or go to a branch of the fund to obtain details.

This stance may be contrasted with the funds' eagerness to sell supplementary insurance for a range of health expenditures (on minor optical and dental work, podiatry, dietetics, pharmaceuticals, and so on) which are individually quite trivial, and for which, therefore, insurance is fairly pointless.  It may be that the funds have correctly discerned what their customers want, but an alternative explanation is that they are more interested in maximising the volume of contributions, since this, more than anything else, justifies better offices and computers and larger numbers of staff -- things more important to the managers of the funds than to their contributors.  FED insurance tends to work in quite the opposite direction, since it insures only against atypical, large, health outlays, not run-of-the-mill small ones of no more financial significance than a family night out at a pizza house.

The measures announced by the Minister for Health in September 1985 included a provision to require "commercial insurers" (as distinct from the health funds) and other organisations wanting to sell health insurance, to seek registration.  This is likely to bring about the situation where profit-oriented, limited liability companies compete with the existing non-profit funds on equal terms.

It will be interesting to see how many such firms seek registration, what their strategy in relation to FED packages will be, and more generally, what quality of management they bring to this sector by comparison with that currently existing.  It will also be interesting to see whether the existing funds will be encouraged to seek incorporation as profit-oriented companies themselves.

The second reason why private insurance may be failing to serve the community as well as it might can be found in the various kinds of government intervention in the health insurance field.  One such is the monopolisation of medical insurance and insurance for public hospital treatment (of non-private patients) by Medicare.

Another relates to an aspect of the Medicare system already discussed -- namely, the fact that we have to pay twice over if we want more adequate cover, since as taxpayers we cannot escape from paying for Medicare.  To put it differently, the fact that Medicare makes public hospital treatment free to the patient creates a strong disincentive for people to choose private hospital care, and thus to take out private insurance.

Third, much of the funds' freedom to design insurance packages in accordance with what they perceive to be their contributors' preferences is denied them by the Government.  They must offer a basic table of benefits, the components of which are prescribed by the Minister.  Tables offering higher benefits are built up from this basic table.  In other words, to add insult to the injury of having to be insured by Medicare, even when we turn to purchase other insurance we are compelled to acquire the kind of package which our political masters deem appropriate!

Not so long ago, the basic table contained only two levels of benefit per day of hospitalisation -- $28 and $16 for surgical and non-surgical patients, respectively. ⁽¹²⁾  Then it was changed, in recognition of the fact that different hospitals had case-loads of differing complexity, and therefore, cost per bed day of treatment.  The private hospitals were put into one 0f three different categories, with a different day rate applicable to each one.  At present, a good deal of energy is being expended to try to come up with yet another variation, possibly based on a system of classifying the patient (according to the type of care he is receiving) rather than the hospital.

Another system which could be considered is that of basing benefits paid on patient diagnosis, wherein there is a specified benefit for hospitalisation in relation to appendectomies, hysterectomies and so on -- one important advantage of which is that it gives the hospital the strongest incentive to get the patient treated and back to good health as quickly as possible.  Yet another has already been mentioned -- namely, insurance per medium of HMOs.

I have no wish to become involved in an argument as to which of these -- or any other -- systems might be best.  What seems important is that the insurers and the hospitals jointly be given maximum freedom to do their own research and planning work, along whichever lines they think most promising.  It seems most unlikely that any team of bureaucrats, any working party, any committee that we can think of, will come up with the "right" answer, nor one which will be favoured by all the competing interest groups involved.

The search for better solutions in the provision of health insurance will be in large part a matter of trial and error.  This being the case, conditions need to be such as to give the various entities scope to develop and test their own ideas, both in competition and in cooperation with each other.  Regrettably, the current rules of the game are such that everyone is obliged to undertake the same experiment at the same time.  The progress of mankind would have been slow indeed if this had been true of all other fields of human endeavour!

A fourth aspect of government intervention in the provision of private health insurance is the insistence that the funds follow the practice of community rating (CR) in determining their premium levels.  What this means, roughly speaking, is that all contributors pay the same rates -- regardless of how risky they might be to insure.  This may be contrasted with the usual commercial practice of actuarial rating (AR) or risk rating, under which the premium paid is commensurate with the insurer's assessment of the risk involved for the individual in question.  In other words, low-risk groups pay less for insurance than high-risk groups under AR;  everyone pays the same rate under CR.

The purpose of requiring the funds to operate in this manner is to redistribute income from low-risk to high-risk groups.  Whether this is a sensible objective (from the social welfare or equity point of view) will be discussed below.  For the present, I want to concentrate on the impact of community rating on economic (as distinct from managerial) efficiency in the insurance market.

Before proceeding, it is necessary to understand clearly the term "economic efficiency".  Broadly speaking, consumption of a unit of a good (or service) by an individual can be said to be economically efficient if the value he places on its consumption ($15, say) is greater than or equal to the cost of producing it (say, $10).

Suppose instead the cost of producing the good is $20, which exceeds its value to the individual -- in this case, by $5.  It follows that we can save $20 of costs by not producing the good, give him $15 -- and thus, by his own assessment of his well-being, leave him just as well off as he was previously.  The point of this is that there would then be a residual of $5 which could be given to him, or any other individual, to be spent on whatever goods are most strongly desired.

Thus we can bring about a net gain in well-being to society as a whole.  In other words, there would be an improvement in economic efficiency, since resources (of labour, capital and so on) would be reallocated to the production of goods and services which are more highly valued than previously -- without making anyone worse off.

When the price charged for goods is determined under competitive conditions in a free market, there is a clear tendency towards efficiency in this sense.  On the consumption side, individuals can increase their well-being by purchasing the product (assuming it has at least some value to them).  But the more they buy, the lower the value they put on additional (marginal) purchases.  Thus we might be prepared to pay $5 for one pizza, but probably very little for the fifth, if we have already consumed four!

It follows that the amount we purchase of anything is determined by its price, because we will limit our purchases to the level at which the value to us of the marginal unit falls below the price we have to pay.  Furthermore, on the production side, competition ensures that the price is competed down to a level just sufficient to cover the cost of production.  At this point we have the optimal level of consumption by the individual.  His marginal valuation of the good is just equal to the cost of production, so there is no further scope for efficiency gains of the kind just described.  Higher or lower levels of consumption would lead to efficiency losses.

It can now be seen that the imposition of community rating on the health funds results in inefficiency in the consumption of health insurance by individuals.  The reason is that CR causes the price of insurance to low-risk groups to be higher than the cost of providing it to them, and conversely for high-risk groups.  It follows that the former group will consume too little insurance, while the latter will consume too much.

To put this another way, the low-risk groups are effectively being taxed and the high-risk groups thus subsidised.  The former group will tend to purchase less insurance, so as to avoid at least some of the tax, while the latter will tend to purchase more, in order to obtain an even bigger subsidy.

That this occurs in practice is well known to health policy-makers:  it was precisely this self-protecting behaviour by low-risk individuals which led to concern that quite significant proportions of the population were opting not to purchase insurance.  In turn, this formed a basis for advocacy of compulsory insurance, which was eventually brought to fruition as Medibank Mark I in July 1975.  One could find no better illustration of the maxim that regulation begets even more drastic regulation.

As has been argued already, the current version (Medicare) is deficient in a great many respects:  what a pity the practitioners of regulation have never thought to treat the cause of the illness -- namely, the CR principle -- rather than its symptoms!  Interestingly enough, if the analysis presented in Appendix 1 is valid, the same problem can be expected to manifest itself again, now that the funds have been permitted to offer FED packages.

I predict that it will not take long for the funds to realise that offering such packages affords a profitable means of making insurance more attractive to the low-risk groups.  They will be able to use them to effectively split up the market into different target groups -- and when this happens, the chances of maintaining effective CR will be small.  The health bureaucrats will then have three basic options:  to try to resuscitate it, by becoming even more deeply involved in monitoring and controlling what the funds do;  to revert to prohibition of that most sensible variant of health insurance -- front-end deducibility;  or, to finally abandon CR, and set about creating the conditions under which a more rational and effective insurance system can evolve.

There is one type of health insurance contract that is commercially broadly compatible with the CR principle -- namely, lifetime insurance, taken out prior to the birth (or even conception) of the insured.  With this kind of contract, the insurer would have to base the premiums on likely health expenditure for the "average person" (except that differential health profiles might be able to be related to parental characteristics).  Thoughtful parents-to-be could take out lifetime insurance for their children of the future, safe in the knowledge that the insurer could not later refuse to continue to provide cover at the same rate for all contributors, regardless of their health experience.  Australian insurers have never been encouraged to provide such contracts, however.  Indeed, the conformity imposed on them presumably means that they would not have even countenanced the possibility.

I turn now to the equity aspects of community rating.  It seems to me that the principle of transfers of income from the well-off to the poor is reasonably widely accepted (provided the magnitude of the transfers does not become too great).  The principle that there should also be transfers of income from people who are low health risks to those who are high risks, however, is something quite different.

Who are the low-risk and high-risk groups in the community?  Broadly speaking, the low-risk group comprises children and teenagers, and men from about 20 to 45 years old.  In the high-risk group are included pre-school children, women from their late teens through to old age, and men from the mid-forties through to old age.  Risk also varies with several discretionary aspects of lifestyle, such as drinking, diet, exercise, leisure activities and occupation.  The much higher health costs of women in their late teens through their thirties are presumably related to child-bearing;  this can also be regarded as discretionary.

With these points in mind, deficiencies of community rating from the point of view of fairness become apparent.  Many of the low-risk people are also poor, while many of the high-risk people are also wealthy.  For these cases, community rating perversely transfers income from the poor to the well-off.  At the same time, those who avoid excessive alcohol consumption, do not smoke or use other drugs, have a healthy diet, get plenty of exercise, opt for relatively safe sporting activities and occupations, and remain childless, are forced through community-rated insurance to subsidies those who make the opposite choices in these respects.

It may be argued that since there is a clear tendency for health risk to increase over time for any given individual, the taxes and subsidies inherent in community rating tend to be offsetting over one's lifetime.  This is not altogether correct, however.  When a community rating scheme is non-compulsory (as is the case with private health insurance in Australia), the inherent taxes and subsidies bias participation towards high-risk individuals, as noted above.  One implication of this is that some people will wait till they are older to join, rather than doing so when they are young.  This means that the premium will not accurately reflect the average person's health costs over his lifetime, since the contributor base will not be "average".  In other words, the premium will be biased upwards because of the over-representation of high-risk individuals.

It follows that non-compulsory community rating is inequitable in its treatment of the more risk-adverse members of the community.  These people will be more likely to purchase insurance when they are young and therefore in the low-risk category.  Roughly speaking, such people will pay more in contributions than they will take out in claims over their lifetimes.  Those who are less risk-adverse will stay out of the schemes until they get older, and can then be expected to take out more in claims than they contribute in premiums.

The lifetime experience argument also ignores the point made above about discretionary aspects of health risk.  Moreover, my guess is that most young people would prefer to choose for themselves the proportion of their incomes they spend while still young, and the most desirable form of saving, rather than effectively being forced to save for old age through the insurance system -- especially when the nature of the system is observed to change significantly and frequently, making future benefits highly uncertain.

2) GOVERNMENTS AS HEALTH CARE PROVIDERS

Total health expenditure in Australia is about 7.5% of GDP -- something of the order of $13 billion a year.  Hospitals account for more than half of this, and the public sector -- largely funded by taxes -- for around 80% of total beds. ⁽¹³⁾  Governments are also involved in many other sectors of the health industry, including, for example, Community Health Centres.  Inefficiency in the public sector of the industry, therefore, has very important consequences for us as taxpayers.

In the preceding section the importance of the incentives facing managers of organisations was discussed, as was the efficacy of the free market in allocating productive resources amongst competing uses.  Both points are also relevant here.

The incentives to manage well tend to be even weaker in public sector institutions than in the non-profit health funds.  As people in the public sector often complain, important decisions are often taken out of their hands, and there is anyway little incentive to find more efficient ways to do things if the reaction of the overseeing arm of government is merely to cut back on budgets.  The following paragraphs indicate to some extent how the public sector of the industry fails to match the performance of the private sector.

A health industry professional to whom I spoke recalled that he had once worked in the public hospital sector, where he observed that people responsible for allocating financial resources for recurrent spending worked virtually independently of those responsible for capital spending, since each had a separate budget.  A capital outlay on some item of equipment might be capable of more than paying for itself in terms of reduced staff (recurrent) costs, but it would be unlikely to go ahead because of this compartmentalised way of doing things.

The same kind of management shortcoming can also adversely affect the quality of health care.  Consider, for example, knee cartilage injuries -- the scourge of football players, amongst others.  Until fairly recently, even to discover how serious the injury was required the knee to be opened up -- which necessitated about seven days of hospitalisation, not to mention a further period of convalescence.  A new, but quite expensive, piece of equipment is now available which is basically a tiny video camera mounted on a very thin probe which can be inserted into the knee, without the need for an incision, and moved around so as to allow the doctor to observe the damage on a TV monitor.  If the cartilage is torn, it is possible to repair it in such a way that only one day in hospital is required, and the person can be back at work within a week.  In an environment in which capital expenditure decisions are based mainly on political considerations, and in particular, on the current state of the budget, it is not surprising that equipment such as this is not always acquired, despite its clear benefits for the patient.

Some appreciation of the significance of this point can be gained from a perusal of capital expenditure on health by all levels of government in Australia during 1970-71 through 1981-82. ⁽¹⁴⁾  From around $300 million per annum (in 1979-80 prices) through the first four of these years, the level had virtually doubled just two years later!  Within the next six years, however, spending fell steadily and rapidly, to just over $200 million per annum.  Since our health needs do not fluctuate so wildly, it would seem clear that capital spending decisions have not been based on rational investment criteria.

Many of the large capital expenditure decisions in the public sector are quite clearly politically based, bearing little relationship to economic rationality and "social welfare" concerns.  For example, a third government-funded hospital was built not so long ago in Canberra, presumably to win votes from people nearby.  Several of its wards have never been used, since the other two (and one private hospital) have been able to cope with demand. ⁽¹⁵⁾  I understand also that a whole hospital has remained empty since its construction several years ago in the Melbourne suburb of Essendon.  These days, the other two public hospitals also have beds and wards closed because of the "nurse shortage", despite long waiting lists for admission.  For the bureaucrats to be able to preside over both excess supply and excess demand simultaneously is quite a feat!

I do not suggest that private sector managers never make mistakes when they estimate potential demand in their investment decisions.  I do suggest that the incentives for them to make good investment decisions are much stronger.  Even if mistakes are made, there are right and wrong ways to proceed from there.  These become fairly clear in a market setting where price can vary in response to altered circumstances.

In the context of unintended over-investment in hospitals, if patients were being charged for hospitalisation, rather than being treated for free under Medicare, there would at least be the option of lowering the price in order to stimulate demand.  More importantly, if there were a shortage of nurses, the obvious thing to do, other than giving attention to the kinds of industrial relations and personnel management issues discussed below, would be to increase the salaries offered.  So long as there are queues of patients waiting for admission, as there are at the time of writing for public hospitals, there is scope for raising the price to them to cover the extra salaries needed to get nurses to look after them.  When the price has become high enough to get rid of the queue, there is, by definition, no more nurse shortage.  This kind of flexibility does not exist in our public hospital system, because of the small role permitted the price mechanism.

Another hospital professional described for me how the principles of cost accounting are being used as a tool for improving efficiency in some of the private hospitals.  Briefly stated, if data are collected on a sufficiently disaggregated basis, they can be used to compare the costs of providing treatment of various kinds in different hospitals and under different doctors.  Such comparisons can highlight the more and less efficient ways of doing things, thus providing very useful information to management.  He doubted public hospitals were following this practice. ⁽¹⁶⁾

If they did so, it would be interesting to see the cost of treating patients who utilise casualty wards for many essentially minor problems which could easily be dealt with in a GP surgery.  These wards are not intended to be used in this way, but to provide emergency care for victims of heart attacks, car accidents, fires and so on, for which they are very expensively equipped, and have full access to the hospitals' facilities.  One potential benefit of the 24-hour privately-operated clinics, which have recently been emerging in the larger cities, is that they can be expected to partially offset this waste of resources by lessening the burden on the public hospital casualty wards of treating non-emergency out-of-hours cases.

Industrial relations is another area where public hospital management seems to perform poorly by comparison with private.  Despite the so-called "critical shortage" of nurses, the private hospitals seem little affected:  none, to my knowledge, has closed down beds for lack of nurses, in stark contrast to the public hospitals.  Some private hospitals pay higher wages, but certainly not all.  By far the most important factor is that in various ways they present a more human face to their nursing staff.

It needs to be appreciated that a very high proportion of trained nurses are also wives and mothers.  Bearing in mind that nursing goes on 24 hours per day, every day, flexibility in relation to the number of hours and the times of the day nurses are asked to work is of the utmost importance.  The public hospitals have been slow to recognise this, so it is not surprising that many nurses have voted with their feet for the private sector of the industry -- even though they may expect to be worked even harder.

Nowadays, because of their inability to recruit and keep nurses, public hospitals have become heavily reliant on "agency nurses", that is, nurses who register themselves with agencies as being willing to do nursing work, but specify the times which are suitable for them.  There is a cost, in the form of the agency commission, but this is outweighed by the freedom to be able to work at convenient times.  It hardly needs to be said that the quality of nursing care is bound to be lower under these arrangements, since nurses are unlikely to develop close ties with hospitals when they are constantly on the move -- much less with the patients.  And there is little doubt that patients themselves are made uneasy by frequent changes of the nursing staff entrusted with their care.

The fact that public hospitals are basically financed by budgetary allocation rather than by patients also puts them in a weak position vis-a-vis union militancy, whether amongst the ranks of nurses, paramedicals, technicians, cleaners or administrative workers.  Under competitive conditions in the free market, successful wage and salary claims, and claims for improved conditions of work and employment, are for the most part passed on to the buyer of the product in question.  They are not, as many union leaders and their supporters would have us believe, paid out of the pocket of the employer.

Buyers of the product, of course, have the option of cutting back on their purchases in response to the price rise, which in turn reduces demand for labour in the industry concerned.  In this way, union militancy is constrained by the adverse effect which salary increases and the like have on the demand for labour.

In the context of organisations funded by taxpayers rather than users, however, this constraint is severely weakened.  There is no buyer resistance to a price rise, since there is no price.  If there is to be any resistance, it will have to be from the taxpayer, working through the very indirect electoral process so as to put pressure on his elected representatives to, in turn, put pressure on the Sir Humphreys of the health bureaucracy so that they will do more to oppose union claims.  It should be emphasised that the costs involved are by no means trivial in aggregate:  salaries account for roughly two-thirds of recurrent costs, even in a very efficient hospital.

The extra cost will be widely spread, and buried in the list of thousands of other things on which governments spend our money, so it is unlikely that the taxpayer -- lacking the buyer's freedom to cut back on the amount he pays -- will have much influence on the outcome.  For these reasons, we should not be too surprised to see vigorous opposition to privatisation of public hospitals from the unions involved.

A final example of the relative merits of allocation of resources by the market, on the one hand, and by the bureaucratic decision-making process (the basis for which remains largely a mystery), on the other, can be found in the application of the practice of market research -- the epitome of profit-oriented behaviour -- by the private sector.

The approach can be very simple:  first, ascertain the kinds of surgical procedures for which there are very long waiting lists at the public hospital nearby, then acquire the necessary equipment and expertise to be able to undertake those procedures, and finally, ensure that doctors specialised in this field are made aware of the hospital's new capability.  It is hard to imagine a clearer illustration of the propensity for the market mechanism to effect the production of goods and services desired by the community, as distinct from what some far-away public servant thinks it should want.

There is a good deal of evidence, from many industries in many countries, supporting the view that provision of goods and services by the state is usually less efficient than provision by the private sector. ⁽¹⁷⁾  Appendix 2 contains a small contribution to the literature in this field -- namely, the results of a comparison of the efficiency of government-owned community health centres and private practices supplying GP services.

There is no reason to imagine the state could do better than the private sector:  after all, it has to hire the factors of production, including managers, from the same pools from which the private sector hires them, and it has access to the same accumulated scientific and technical knowledge.  The important difference which usually arises is in the incentives given to management, and in its degree of autonomy.

It follows -- and there is empirical support for this too -- that if the incentives can be made similar to those of management in private firms, then performance can be brought up to similar levels.  In the present context, what this means is that a second-best alternative to privatising public hospitals would be simply to require them to operate along commercial lines, in competition with each other and the private hospitals.

This would entail:  extension of Medicare cover to private hospital patients;  charging all patients the full cost of treatment (in other words, taking the financing of hospitals out of government budgets);  freedom for management to carry out whatever investments are expected to provide a profitable return;  freedom to hire and fire employees of all kinds;  freedom from political interference in all operational and planning decisions;  appointment of top class executives (not necessarily drawn from the hospital industry) to the highest levels of management;  and holding these executives accountable for results achieved.

Lest it be objected that this would lead to huge increases in the cost of hospital care, it should be clearly understood that quite the reverse is true.  There is, I believe, no doubt that profit-orientation as distinct from direction by public servant bureaucrats would lead to markedly increased efficiency -- costs would be lower, not higher.  The price to the patient (or to his insurer) would of course be higher, but there would be a more-than-offsetting reduction in taxes needed to pay for the system.  Whether our political masters would pass on these savings as tax cuts, or spend them on things such as new Houses of Parliament, is of course a moot point, but it should not be allowed to obscure the logic of the argument.

3) RESTRAINT ON COMPETITION

It should be clear from preceding sections that the process of competition in free markets brings very significant benefits to the community.  The quest for profit leads producers to find out what people want, and competition forces them to produce it as efficiently as possible.  The price mechanism has the effect of ensuring that goods and services are produced in optimal quantities -- neither too great nor too little.  Since there is a tendency for many to argue that markets do not work "perfectly", let me readily concede that point:  indeed, nothing in this world, it would seem, is perfect.

But of course, this is not sufficient justification for governments to intervene in markets.  It is just as obvious, I think, that governments are also imperfect in many ways, so the relevant question must always be:  whnih is the least imperfect?  Or, more precisely, in which circumstanees can governments be expected to outperform markets in catering to the material desires of the community?

A great deal has been written on this topic, which need not be repeated here.  I wish merely to discuss a simple extension of the analysis of the workings of competition -- namely, the implication that factors which inhibit competition necessarily detract from its potential benefits.

Some important constraints on competition in the health industry have already been noted, such as the Medicare monopoly of medical insurance and insurance of public hospital patients.  Licensing requirements have also restrained the level of competition in the hospital industry.

Another aspect worthy of attention is the prohibition on the practice of medicine by people other than registered doctors.  To become registered, a person has to be a graduate of a medical faculty at an Australian, British, Irish or New Zealand university (with a few exceptions in practice), to have spent a year as an intern, and to be "of good character".

To the layman, these kinds of restrictions may seem a quite reasonable way of protecting the public from "quack" practitioners likely to do more harm than good.  Indeed, it was just this sort of argument which was used in the process of lobbying parliamentarians to pass the relevant legislation around the turn of this century. ⁽¹⁸⁾  But the matter is by no means as simple as that.

Protestations that anti-competition restrictions are needed for the benefit of the consumer, or the general public, need to be treated with considerable scepticism.  It is no coincidence that, typically, the people and organisations who most loudly advocate such restrictions are precisely the ones who stand to gain the most from them, in the form of higher incomes made possible by the existence of less competition.  Doctors would appear to fit this pattern.

In the latter part of the nineteenth century, of course, medical science had not progressed anywhere near as far as it has today -- the practical implication of which was that for many health problems, one had as good a chance of being cured (or harmed) by a so-called "quack" as by a doctor -- regardless of whether the latter may have had a more extensive formal education.  Under such circumstances, it would have seemed reasonable that practitioners of what is known today as "alternative medicine" should have been permitted to compete on equal terms.

Further evidence in support of the view that the doctors' concern in having the legislation passed included their own interests, is found in the special provision which allowed established "orthodox" practitioners to become registered (thereby maintaining their political support), even if their qualifications did not match the new requirements, and in the extremely difficult barriers placed in the way of qualified doctors who might have wished to migrate from countries other than those mentioned above.

There are difficulties in establishing the standard of university degrees from other countries, of course, but other professions like engineering, architecture, science and economics manage to surmount the problem.  A medical profession truly concerned with bringing better health to the people, in the form of an expanded supply of doctors, could presumably do likewise.  But in the present environment in which there is downward pressure on doctors' incomes, there are already some doctors at least arguing in favour of a reduced intake of medical students at Australian universities -- despite the fact that Australia appears to have somewhat fewer doctors per head of population than the US, Canada, and almost all Western European nations ⁽¹⁹⁾ (I hasten to add that, in this, they may be quite unrepresentative of the profession as a whole:  there are, no doubt, many doctors for whom the satisfaction of healing the sick is more important than the amount they earn from doing so).

Regardless of motivation, the reduced supply of medical practitioners due to entry barriers (that is, below the supply which would have been observed in the absence of those barriers) has raised the cost of medical care.  This effect has been re-inforced more recently by the artificial stimulus to demand for doctors' services resulting from the reduction in the patient price of those services to, or near to, zero by Medicare.  The obverse of this, as has already been pointed out, is higher incomes for doctors.  This is not the same, however, as saying that doctors are better off:  paradoxical as it may seem, it appears they are not. ⁽²⁰⁾

By increasing the incomes of doctors, entry barriers have the result of making medicine relatively more attractive than it would be otherwise.  If universities responded by expanding their facilities in order to train all these additional aspiring doctors, much of the initial increase in doctors' incomes would be competed away again.  Such a development would please neither the Ministers for Education (who would have to finance the universities' expansion), nor those doctors already practising.

Perhaps for these reasons, then, we find that it has been made very difficult to obtain a medical degree.  There are such stringent entry quotas that only young people with extremely high grades in their last year of high school can gain entry.  The course itself is longer (six years) than for most other professional disciplines, and is said to be relatively arduous and tough in its assessment procedures.  Nor should the requirement to work for a year as an intern be forgotten.

It can therefore be argued that doctors are no better off -- even though their incomes are higher than they would be in the absence of entry barriers.  We cannot look only at their incomes.  It is essential to take into account the necessary initial "investment" by aspiring doctors, in terms of having to study harder at school and at university, and of the incomes they must forego during their lengthy period of training.  Their education is heavily subsidised, of course, but this is true of other disciplines as well.  Evidence suggests the pecuniary costs alone of foregone earnings are sufficient to offset the relatively higher income stream in later years. ⁽²¹⁾  This did not apply, of course, to the doctors who were already practising when the registration legislation was enacted.

Of course, it may be true that only the cream of our high schools' output has the intellectual capacity required of a general practitioner in his day-to-day activities, and that this must be complemented with a very lengthy and intense period of study.  But I doubt the intellectual demands of medicine are any greater than, say, those of engineering.  To put it another way, it would seem that the system tends to draw people who are excessively capable into medicine, bearing in mind the cost involved -- in that they could otherwise make valuable contributions to other disciplines.  Moreover, it tends to train them too intensively and for too long.  This point warrants further discussion.

An important feature of medical practice is that doctors concern themselves with an extremely wide range of health problems -- all the way from sore throats and ingrown toenails to cancer research, heart transplants and the like.  Moreover, medical practitioners specialise.  This makes sense, since it would be an absurd waste of, say, a heart surgeon's abilities to have him spend his time doing pap smears and inoculations.

It also makes good sense that general practitioners act as referral agents for doctors who are more specialised.  This enables health problems to be screened along the way:  the GP will likely have a better understanding of the problem than the patient, but if he is unable to handle it, he will be able to refer the patient to a suitable specialist.  This system could be made considerably more effective -- less costly to the community -- if people could practice "simple" medicine without needing to fulfill the very onerous conditions currently required for medical registration.

Suppose, for example, that a degree in elementary medicine (M.E.D.?) could be obtained from 3 years' university study plus a year of internship at a general hospital or even at a GP practice.  Such a course could be more selective in its coverage, emphasising diagnosis and treatment of a range of less complex cases presently handled in GP practice.  Holders of such degrees could then set up their own practices, and operate in just the same way that GPs now operate -- that is, treating patients themselves when possible, and referring them to more highly trained or more experienced practitioners if not.

Alternatively, they could become employees or junior partners in group medical practices in which they would handle uncomplicated work -- in much the same way that junior motor mechanics change sets of spark plugs, while their more experienced colleagues overhaul engines.  We could expect people with such qualifications would undertake further training, refresher courses and the like to keep abreast on new developments.  The usual legal safeguards against fraudulent misrepresentation of the nature of one's qualifications would of course apply.

The benefit of this would be that a much wider range of young people could enter the medical profession, rather than the few who belong to the academic elite and whose families are sufficiently well-off to be able to support them through at least six years of studies.  Moreover, the use of educational resources would be rationalised, by reducing the educational input to those individuals undesirous or incapable of undertaking the more rigorous levels of training appropriate to more demanding fields of medicine.

The cost to the community of treating medical problems of all levels of complexity would fall, since the entry of "M.E.D." practitioners would force those more highly qualified to specialise in the treatment of more complex cases, thus effectively increasing the supply of specialists at all levels.  These benefits are largely foregone, however, given our present medical registration rules.

Interestingly enough, a variant of this approach to health care, in which less highly trained manpower is used to deal with less complex medical work ("simple medicine", as I referred to it above), is already being practised in the public sector -- specifically, in government-owned community health centres (CHCs).  Practices vary between centres, but in those I have visited it is normal for nurses to carry out many of the less complex tasks which GPs normally undertake in private practice.  Sometimes this is in conjunction with a doctor, but often it is not.  Many patients go directly to a nurse, and only if she feels unable to do what is necessary does she then refer them to a doctor.  I believe the same is true of HMOs in the US.

Whether or not this activity is legal is an interesting question:  this would depend on interpretation of the phrase "to practise medicine".  The term "medicine" is not defined in the Medical Registration Acts, and I am not aware that the courts have ever been asked to make such an interpretation.  Fear of prosecution, however, is probably sufficient to dissuade trained nurses and partly-trained doctors from setting up their own "simple medicine" practices.  This point aside, the private sector is unlikely to follow the good example set by the CHCs, because Medicare would not reimburse the patient for consultation with someone who is not a registered doctor.

This point has even wider ramifications, in that there are paramedical practitioners who may be better able than doctors to deal with some problems.  Physiotherapists, for example, are more appropriately trained for the treatment of some sporting injuries.  Clinical psychologists are specialists in things like finding the causes of patients' depression.

A well-designed health system would encourage patients with such problems to go straight to the relevant paramedical, rather than a GP.  But the ability of the paramedicals to compete with doctors is seriously constrained by the fact that Medicare provides no benefits for paramedical services.  If the price a patient pays for a GP consultation is only about $2.50 or less, compared with, say, $15-$20 for a consultation with a paramedical, it is not surprising that many patients choose the former -- however inappropriate this may be.

Clinical psychologists (and other paramedical groups as well) operate a system of accreditation in order to help assure their patients of the quality of service to expect from accredited practitioners.  Accreditation can only be gained if the person concerned can convince his peers of his competence to practise.  The important difference between this and the process of registration of doctors is that there are no legal sanctions involved (other than in respect of misrepresentation), so that the patient is free to go to someone who is not accredited if he so desires.  This precludes the possibility of unjustifiably tough accreditation standards being maintained (as seems to be true of doctors), since this would bring with it such a high price that the patient would seek instead the services of practitioners who are not accredited or those accredited by other organisations.

Another aspect of restriction on competition between doctors is the prohibition on advertising.  A purpose of advertising, like that of accreditation, is to provide useful information to customers.  If it were permitted in medical practice, it would presumably emphasise such matters as the existence, location and opening hours of surgeries;  the willingness of doctors to make home visits;  their qualifications;  whether they were accredited, and if so by whom (if accreditation rather than registration were the norm);  their knowledge of foreign languages;  their sex;  their number of years of experience;  their specialisation, if any;  and, of course, the fees that they charged.

But the provision of this kind of information by doctors, notwithstanding its obvious usefulness to patients, is considered "unethical".  To do something unethical would be to risk being found to be not "of good character", which in turn would be grounds for deregistration.

Before leaving this topic, it is worth mentioning that in recent years a process similar to the semi-closure of entry to the medical profession has been going on in nursing.  Once the most quiescent of workers, nurses have, in the last few years, discovered what can be gained from militant unionism.  Some have been much distressed by the trend:  it is one factor cited in explaining the drift of nurses to private hospitals, where better personnel management weakens the attractions of unionism.

One of the demands of nursing organisations has been that nurse training be taken out of the hospitals and into academic institutions, so that nurses of the future will also be holders of college and institute of technology degrees.  Health industry policy-makers have already agreed to this, and the transition is well under way.

There have also been demands for salary increases, and the nurses' case is in the process of arbitration at the time of writing.  Given the "nurse shortage", increases in the nursing awards would seem sensible, although it is regrettable that ways of getting more trained nurses back into the workforce simply by making work arrangements more flexible have not been explored with greater enthusiasm.

People in the industry seem to think it inevitable that very large salary increases will be awarded.  The more sensible procedure of successive moderate upward adjustments to levels just sufficient to attract the necessary additional numbers of nurses into the workforce is apparently not an option.  It will be interesting to see whether the various health ministers react by passing on the extra burden to the taxpayer, or by closing more wards, or even by redesigning the Medicare system yet again so as to provide an offsetting cost reduction.  It should be noted here that nurse salaries are by far the biggest single item in the costs of providing hospital care.  Much has been made of the attempt by the NSW Government, for example, to keep hospital payments to doctors down, but the potential savings would appear to be small by comparison with the costs which would be occasioned by even a moderate increase in nursing salaries.  There are roughly twice as many full-time nurses as there are doctors, not to mention half as many again working part-time or as casuals, and a similar number of nursing aides. ⁽²²⁾

Whatever the ramifications for the hospital system, higher salaries will increase the number of persons wanting to train as nurses and the number of nurse immigrants, but it seems unlikely that nurse intake to the hospitals will be able to expand commensurately under current budgetary conditions.  This in turn will make it easier to impose quotas (based, perhaps, on prior academic attainment) on the student nurse intake and to dampen the enthusiasm of many aspiring nurses by making courses more demanding.

Some doctors have expressed their concern to me that the new degree-wielding nurses will not feel inclined to carry out the more menial tasks that have traditionally been their responsibility.  If this were the case, it would then be necessary to have less highly-trained nurse subordinates available to do these tasks, and this matching of function to training would be entirely appropriate.  What may happen, however, is that these more highly-trained and much more expensive nurses will continue to do the same things nurses have done before, just like the highly trained GP spending much of his time on medical trivia.

It seems unfortunate, therefore, that the previous nursing system was not modified, rather than being so drastically transformed.  A major complaint of nurses has been the lack of a career structure in their profession.  There were only five levels to which trained nurses could be promoted;  further advancement required them to turn to either teaching or administration -- unattractive options for those who wanted to work directly with patients.  It is a pity that there was no experimentation with creating further levels of advancement -- based on demonstrated ability and further training in the more esoteric aspects of nursing -- rather than moving to a new, more academically-oriented system.

CHAPTER 3
THE UNDERLYING CAUSES OF HEALTH INDUSTRY PROBLEMS

In the preceding section I discussed the proximate causes of a range of problems in the health industry.  Broadly speaking, all of these causes related to some form or other of governmental intervention in the workings of the health markets.  Since we are encouraged to believe that the purpose of government is to advance, rather than harm, the interests of the community, it might be helpful to step back and ask what has led governments so far astray.

There are three aspects worthy of attention:  first, the tendency to use the health system as an instrument to achieve quite a wide range of policy objectives;  second, inadequate consideration of alternative means of achieving those objectives (and even failure to contemplate the possibility that some of the objectives might be incapable of achievement);  and finally, the ever-present tendency of governments to be influenced more by concentrated special interest groups than by concern for the interests of the community at large.

The last of these points has already been touched on in the discussion of entry barriers to the medical profession.  Pressure exerted by health industry unions in opposition to privatisation of hospitals is a further example.  The following discussion will concentrate, therefore, on the first two aspects.  It may be helpful to proceed by working through a list of objectives which seem to have been in mind when current policies were determined.

1) FREE HEALTH CARE FOR EVERYONE

This is not literally possible.  Health care uses resources of both labour and capital, so it cannot be free.  It is possible for it to be made free at the point of service -- that is, for someone receiving health care not to have to pay for it -- but of course it still must be paid for somehow, by someone.  Some people seem to believe that there is something intrinsically desirable in not requiring the patient to pay (or requiring him to pay only a fraction of the actual cost), but if it is borne in mind that someone must pay, it becomes obvious that we cannot really say anything much about the desirability or otherwise of such an arrangement without knowing what the financing system is.

2) REDISTRIBUTION OF INCOME FROM THE WELL-OFF TO THE POOR

One way of doing this would be to have the entire health system within the public sector, delivering health services to the public at no charge, and collecting a special tax or levy pro rata with peoples' incomes to finance the system through the budget.  This, of course, is not so far removed from our present system.

The fact is, however, that we already have a progressive income tax system in place (not to mention many other policies) for redistributing income to the poor.  In other words, redistribution of income can be, and is being, achieved anyway, without any need to involve the health industry.  If society were to decide that more redistribution is called for, this could be done by altering the tax scales.  The important points here are that by using the health sector for this purpose, it becomes necessary to build up a complete new system of administration, and by changing the way in which health care is supplied and paid for, it is inevitable that serious inefficiencies will be introduced to that system.

3) CONTROL OF PRACTITIONERS' FEES

"Fee" is of course just another word for price, and we have seen already that the price mechanism performs the important function of allocating productive resources in roughly optimal quantities to various different uses.  Fee control, of course, works counter to this.  It is amusing that governments keep doctors' fees above what they would be in competitive conditions, both by the restrictions on entry to medicine they have legislated, and through the poorly-designed system of Medicare benefits outlined earlier.

Fee control also results in undesirable changes in the quality of service provided by practitioners.  To the extent a government succeeds in holding the fee for, say, a GP consultation lower than what the doctor would prefer to charge (given the demand for his services), he can be expected to react by economising on costs, thus causing quality to deteriorate.  For example, the air-conditioner and the heater might go from the waiting rooms, and the carpet wear thin;  the surgery might move to a place less convenient for patients.  Most importantly, the average duration of the consultation might be shortened, resulting in less accurate diagnoses and even inappropriate treatment.

In short, it may be possible to control the price, but not the quality, of the product.  All that will be achieved will be to reduce the doctor's ability to find the price/quality combination which is preferred by his patients.

The motivation for controlling doctors' fees, presumably, is the underlying desire to control their incomes.  But this raises some important issues:  why shouid medical people alone have their incomes controlled?  Why not all professional people -- dentists, lawyers, engineers, accountants, architects, economists and so on?  Why not other groups of relatively high income earners:  successful actors, writers, sportsmen, shopkeepers, entrepreneurs of all kinds? ⁽²³⁾

Lest it be argued that "ordinary working people" already have their incomes controlled by the various industrial tribunals, it should be pointed out here that, for the most part, award wages are set above levels which could be earned in the absence of those awards.  Controls on doctors' fees are intended to hold their incomes down, not up.

This gets back to the point about the (admittedly imperfect) progressive income tax system, which taxes people more heavily if they are better off, regardless of their occupation.  One would think this the fairest and most sensible way of redistributing income.

4) CONTROL OF THE LOCATION OR GEOGRAPHIC DISTRIBUTION OF PRACTITIONERS

Some people find cause for concern in the fact that the geographic distribution of doctors differs from that of the population in general. ⁽²⁴⁾  That is, if we calculate the ratio of doctors per head of population, we find that -- as is likely with every occupation -- it differs between localities.  These people would like health policy to relocate doctors to areas where they are presently "under-represented".

One way governments have sought to achieve this objective has been to create community health centres, staffed by salaried doctors and fully financed by the taxpayer, in areas least well served by local doctors.  The NSW Government has also gone to some lengths to relocate hospital facilities to Sydney's western suburbs in order to increase the number of doctors there.

The first point to note is that these are really means, once again, to redistribute income in favour of the poor -- in this case, by provision of a subsidy in kind, in the form of cheaper and more convenient access to medical facilities.  However, this is a very crude redistribution mechanism.

To begin with, not all the people living in the areas in question are poor, nor do all the poor people live in these areas.  It follows that the subsidy is somewhat indiscriminately spread, with some better-off people receiving it, and many poor people being made worse off because of the implied depletion of doctors in other areas where they happen to live.

Moreover, increasing the availability of doctors in a given area makes it more desirable to all (so far as amenity is concerned), and thus pushes up land values.  The increased cost of living in such a location (whether in terms of rent or the opportunity cost of not selling one's house and moving to a cheaper area) will not be sufficient to offset the benefit of having closer doctors (otherwise, market forces would have led doctors to locate in the area at the outset).  In the longer run, therefore, some people will move away, and others like them who would have moved there will not now do so.  In short, the attempt to bring the doctor to the people may only succeed in driving the latter away!

Besides, where does such a policy end?  Do we really have to impose uniformity on our way of life?  Will we calculate the same sorts of ratios foi dentists, veterinary surgeons, lawyers, social workers, supermarkets, swimming pools, park lands, libraries and so on?  Why should we attempt to deprive people of the right to make their own choices as to the preferred combination of housing cost and local amenity?  I suggest no answers to these questions, since they involve making value judgements -- an activity in which the economist has no special expertise.  I do think it important, however, to draw attention to them.

5) CONTROL OF THE MIX OF PRIVATE AND PUBLIC PROVISION OF HEALTH CARE

There are many in the community who seek to downgrade private sector provision of health care because they find the notion that people would want to prosper from other peoples' suffering offensive.  It appears to me that this way of thinking is singularly unhelpful, because it rests on the unrealistic view (or wish) that health care providers should be a good deal more saintly than the rest of us.  Why should we expect surgeons, nurses and pharmacists to be differently motivated from motor mechanics, room maids and grocers?

If I break my leg, or contract a disease, or need to have my haemorrhoids removed, all I ask is that the person who treats me does a good job and keeps the cost as low as is consistent with this;  his motivation is of no concern whatever.  If he wishes to treat disadvantaged people for free he will gain my respect, but I do not expect him to do so, any more than I expect supermarket proprietors to give away food to such people.

It seems more to the point to ask if the system provides health practitioners and administrators with the incentives to use available medical technology competently and efficiently, and it seems to me that self-interest (the profit motive), combined with plenty of competition between health care producers, does precisely that.

Those who wish to increase or even maintain the extent to which health care is produced in the public sector would seem to be under some obligation to explain their reasoning, since there is no evidence to suggest that this will reduce the resource cost of health care.  On the contrary:  there are very few examples of public sector production of either goods or services being more efficient than private sector production, although there are countless examples where the reverse is true.

CHAPTER 4
REDESIGNING THE HEALTH SYSTEM

I contend, therefore, that incautious selection of objectives in the health field, all needing government intervention of some kind in the various markets for health and health insurance, is responsible for the unhappy state of our health industry.  I have no doubt that total non-intervention would leave us with a far more efficient system.

It seems improbable, however, that there would be much support in the community for such a policy.  Clearly there would be significant numbers of individuals and families -- particularly amongst the poor -- who would choose to gamble (in some cases, unsuccessfully) against ill-health by not purchasing insurance;  there are those also who already have serious medical problems needing very costly treatment, which of course cannot be insured under ordinary commercial conditions.

It is not the place of economists to pass judgement on the moral issue of whether society should assume the burden of looking after such people.  On the assumption that a substantial majority of the community favour doing so, however, economists have a role to play in the design of suitable policy instruments.

The discussion below assumes that the following single, clear objective is adopted as the basis for health policy:

That no Australian resident suffering sickness or injury should ever be forced to forego treatment and care by the lack of the means to pay for it.

I have chosen these words very carefully.  They do not pre-judge the issue of how the objective is to be met.  Moreover, they are intended to avoid the likelihood of "overkill" in the design of the health system -- that is, the situation in which more than the necessary amount of resources is devoted to it.

I propose just a single objective, rather than a multiplicity of them.  Some objectives to which others subscribe do not appeal to my sense of values (for example, that of controlling doctors' incomes), nor would they, I suspect, to those of most Australians, once their full implications are laid bare.  Some other objectives are simply ill-conceived, as I have tried to show above.

Moreover, it is important to note that the design of the system is made vastly simpler if the number of objectives is kept small.  When there are multiple objectives, there are almost certain to be conflicts between them.  This makes it necessary to make difficult decisions, with more complete fulfillment of one objective being traded off against less complete fulfillment of another.

I have been at pains to argue that, from an efficiency point of view, the provision of health care and health insurance, by profit-oriented private sector entities under conditions of free market competition, will be superior to policy regimes in which governments intervene in those markets.  Nevertheless, as has already been noted, such a system would not achieve the social objective set out above.  In other words, some government intervention is necessary if we are to achieve that objective:  the problem is to design a form of intervention whose deleterious effects on efficiency -- including the cost of the bureaucracy needed to implement it -- are minimised.

The scheme 1 propose is, by comparison with the present absurdly complex set of interventions on the part of both Commonwealth and State Governments, very simple.  Its components are as follows:

1. A minimum level of health insurance cover is required of all Australian residents;
2. All health care and health insurance is provided by the private sector, operating free of government constraints (other than the laws applicable to business activity in general);
3. There are no restrictions on competition amongst health care providers and health insurance organisations;
4. A subsidy for the purchase of insurance may be paid to individuals or family units if their income is very low and/or if the cost of insurance to them is very high;
5. In determining the parameters for payment of the subsidy (as described below), the target number of recipients is 15% of the population (a figure chosen fairly arbitrarily for purposes of argument);
6. Evidence that insurance is being maintained to at least the level prescribed, and of the cost of that insurance in cases where a subsidy is claimed, is included with each individual's annual income tax return;
7. In calculating the subsidy entitlement, "income" means income as calculated for income tax purposes;
8. Individuals judged to have very great difficulty in arranging insurance for themselves (through mental or physical incapacity) can elect to have the Commonwealth Government's Department of Social Security arrange it on their behalf, allowing free competition between health insurance organisations for the supply of such insurance.

DETAILS OF THE PROPOSED SCHEME

The thinking behind each of these points is elaborated below.

1. Compulsory insurance

   Although this may not meet with the approval of libertarians (that is, those who oppose the curtailment of individuals' freedom), there is in fact no way that an objective such as the one outlined above can be met without some loss of our freedom.  It is for the community to decide the extent to which the one is traded off against the other.

2. Private sector provision

   The rationale for this should be plain from earlier parts of the discussion.  The essential point to be noted is that, once insurance has been made compulsory, our social objective has been met.  All that remains is to ensure that this is done efficiently.

3. No restrictions on competition

   Again, the rationale should already be clear.  Restrictions on who can compete raise the cost to the buyer.  Restrictions on how they can compete reduce their ability to discover, by careful research and by trial and error, the combination of product price, characteristics and quality most preferred by different buyers.  One specific point of interest is that the principle of community rating of health insurance will no longer be enforced.  Presumably, CR will be replaced by actuarial rating;  the former is of course no longer appropriate, since redistribution of income to those in the high-health-risk groups who are also poor is now effected directly, by way of the subsidy.  I assume that the community has no particular desire to subsidise high-risk people who are sufficiently well-off to be able to afford insurance, expensive as it may be.

4. Insurance subsidy

   The subsidy is targetted at those individuals and families who are both poor and subject to high health risk (that is, who will face insurance premiums higher than they can readily afford).  A suggested set of formulae for calculating the maximum subsidy entitlement (S_m) is as indicated in the Table below.  The numerical values used are for illustration only;  considerable effort would need to be given to determining their appropriate values before implementing a system such as that proposed here.

   Premium Range ($)	Maximum Subsidy ($)
   0 to   500	0
   500 to 3000	0.5P   -250
   3000 to 6000	0.9P -1450
   over 6000	P -2050

   where P is the insurance premium (actuarially calculated by the insurer) for the individual or family in question.

   
   

   The actual subsidy paid (S_a) depends on income after adjustment for insurance outlay, as indicated by the following relationship:

   
   S_a = S_m - 0.25 (Y - Y_m - P),

   
   

   where Y is taxable income, and Y_m is an income cut-off level determined as described in 5) above, S_a may not exceed S_m.

   The arrangements are illustrated in the diagram.  The maximum subsidy (S_m) depends of the premium (P).  For the case chosen, the premium is $4,500, and the maximum subsidy $2,600.  To find the actual subsidy (S_a), we draw an additional vertical axis at the value of the premium, and then draw a line with a slope of -0.25 through its intersection with the maximum subsidy curve.  For any given level of income (Y) in excess of the cut-off level (Y_m), the actual subsidy is then read off the S_a schedule.  In this example, excess income (Y - Y_m) is $10,000, and the actual subsidy is $1,225.

   This somewhat complicated arrangement has the following characteristics.  First, people with net incomes, after payment for insurance, at or below the cut-off level (Y_m) will receive the full insurance subsidy (S_m) from the consolidated revenues of the Commonwealth Government.

   Second, for a certain range of net incomes above Y_m, the subsidy will be partial, declining by $1 for each extra $4 of income.  If income is so low that there is no tax liability, this is equivalent to a fairly low marginal tax rate of 25%, thus avoiding the possibility that the subsidy would provide a strong disincentive to earn a higher income.  If the income cut-off turns out to be at a level already subject to income taxation, the overall marginal rate would be higner, so some reworking of the tax scale may be called for.

   Third, the subsidy falls to zero when insurance cost is low ($500 or less) or when net (after-premium) income is sufficiently high -- specifically, when

   Y - P = 4S_m + Y_m

   (or greater).

   Fourth, to a very considerable degree, there remains a strong incentive to the individual to search for the lowest cost insurance cover, subject to the compulsory requirements as to the extent of that cover.  It is virtually impossible, however, to maintain this incentive for very low income earners subject to very high insurance premiums.  This is an inherent difficulty which arises from our assumed basic objective -- which insists that everyone be insured (one way or another), even though not everyone will be able to afford insurance.

   The subsidy proposed above is such that, amongst the 15% (or other arbitrarily-chosen proportion) of the population whom it is intended to subsidise, all have to pay the first $500 p.a. of insurance costs, then 50% of costs above this level and up to $3,000, plus 10% of the additional costs up to $6,000.  Only for insurance costs in excess of $6,000 does the Government's share rise to 100%.  For these cases, it may be worthwhile for the Government to act as a purchasing agent, bearing in mind that it will be involved in this activity anyway (Item 8).  An indication of the amounts of insurance subsidy people might receive is shown in the following Table, which assumes an income cut-off level of $10,000.

   Health Insurance Subsidies

   Taxable Income ($)		0	10000	15000	20000	25000
   Premium	Maximum Subsidy	Actual Subsidy
   0	0	0	0	0	0	0
   500	0	0	0	0	0	0
   1000	250	250	250	0	0	0
   2000	750	750	750	0	0	0
   3000	1250	1250	1250	750	0	0
   4500	2600	2600	2600	2475	1225	0
   6000	3950	3950	3950	3950	2950	1700
   8000	5950	5950	5950	5950	5450	4200
   15000	12950	12950	12950	12950	12950	12950
   20000	17950	17950	17950	17950	17950	17950

   
   

5. Limitation on number of target recipients

   It is impossible to subsidise everybody.  This is a reminder of one of the great conceptual shortcomings of Medicare:  it encourages us to think that we are all benefitting from cheap health care.  Rich and poor alike are subsidised every time they visit a doctor, but of course to finance all these subsidy payments, they all have to pay higher taxes.  Even the poor contribute, by way of sales taxes and so on.

   It is hard to imagine an administratively more cumbersome way of effecting income transfers to the high-health-risk poor than one which requires transfers to everyone, backed by a correspondingly enlarged tax collection effort.  It seems wise, therefore, to make explicit the numbers of people who can expect to receive the subsidies.  Whether the ratio to total population is 15% or some other figure is, of course, a matter for the political processes to decide.

6. Reporting by individuals

   If insurance is to be compulsory, there will need to be monitoring.  The intention here is to keep the cost of monitoring as low as possible, by using the existing income taxation infrastructure.  It is natural, in any event, to work through the tax department, since the subsidies are directly related to taxable income.

7. Subsidy based on taxable income

   In principle, if we want to know how well off people are (for the purpose of either taxing or subsidising them), we should take account of their assets as well as their income.  This is especially the case here, since many of the people with the highest health insurance costs will be retired persons with little income but, in many cases, significant accumulated savings in one form or another.

   Rationalisation of the tax system along these lines, however, has proved more than any government has been prepared to take on, so for the time being at least it will probably be necessary to rely on the same measure of "well-offness" for purposes of administering the health insurance subsidy as is done for determining income tax liability.

8. Government as purchasing agent

   The intent of this is sufficiently obvious not to require elaboration.

CHAPTER 5
THE "UNINSURABLES"

People in the health insurance business occasionally refer to the "uninsurables" -- a group of people who, it is said, cannot be insured.  The existence of this group is cited as one reason why we could not rely on a free market (competitive) outcome in the health care/insurance industry, because this group would end up being uninsured and burdened with intolerably high health bills.  People with pre-existing ailments are supposedly uninsurable, as are the frail aged.

Consider the latter group.  These people are undoubtedly in the very high-risk group -- which is another way of saying that they have a very high probability of becoming ill, and requiring lengthy hospitalisation.  In the case of those with pre-existing ailments, the need for treatment is certain, and the cost of treatment is known reasonably accurately.

It is certainly true that the health insurer using community rating would not want to have such people as contributors.  Since premiums are based on claims expected from contributors as a whole, these people would draw more in benefits than they would pay in premiums.  To allow them membership would require an increase in the premiums paid by other groups to cover this higher rate of claims.  Insurance would become less attractive to all other (lower risk) groups as a result, and many would perhaps opt for no insurance.  In turn, this would require even higher premiums for those who remained.

Nevertheless, it is a matter of degree involved here, not one of principle.  The same problem arises whenever there is a mix of different risk groups being covered, if the charge for that cover is the same for all of them.  It is incorrect, literally, to say that the very high-risk group is "uninsurable".  It is true that the fund will make a loss on them, as a market segment.  But the same is true of all groups whose risk is higher than average.

Whether or not the problem is insurmountable depends on how numerous are the "uninsurables" relative to other groups, and how high are their expenditures.  It also depends on whether the low risk groups can opt out (this is avoided under a monopolised, compulsory insurance scheme), and on whether they are sufficiently risk-averse to choose to stay on regardless.

Things work out rather differently under actuarial rating.  Here, the fact that the insurer provides cover to the very high-risk groups is of no consequence to the others, since each identifiable risk group is separately rated in line with its own claims experience.  Whether insurance will then make any sense to the high-risk group is open to doubt.  In normal (free market) circumstances, it may not.

The premium must equal the probable cost of health expenditures to be incurred, plus a mark-up for the administrative costs of the insurer.  In the case of a pre-existing ailment, the probability becomes unity (certainty), so the consumer would necessarily lose an amount equal to the administrative mark-up by purchasing insurance.  For people such as the frail aged, the probability of incurring high health care costs is close to unity, so much the same is true.

In short, insurance is not a sensible option for very high-risk groups, but this does not mean that they are "uninsurable".  What it means is that the premium which would need to be charged to make the contract attractive to the insurer would have to be so high as to make it unattractive to the insured.  The distinction is important, as explained below.

Under the scheme of compulsory insurance outlined above, the high risk group will have to pay very high rates, which, in the absence of the subsidy, some people would not be able to afford.  But given the details of the proposed policy, the fact that the premium is very high is not of such great consequence to the buyer, since his purchase is being heavily subsidised.

Nor is it illogical to require insurance for the "uninsurables", paradoxical as this may seem in the light of the argument above.  The cost of caring for the "uninsurables" is a cost which will have to be met, one way or another, given our social objective.  Under the present system, a very large part of these costs is met through the budget, with the involvement of the Social Security Department, the various Health Departments, and Medicare.  What is at issue here, then, is whether the administration of all of this would be more efficiently handled by private insurance companies.

Whether or not insurance qua insurance for the "uninsurables" makes any sense is really not the point.  What matters is that in the scheme proposed here, the insurance companies would compete for this business -- since there is no reason why it should not be profitable for them -- and this competition would tend to keep the cost of health care for this group down.  The incentives for progress in this direction would certainly be much stronger than they are at present.

CONCLUSIONS

In this paper I have outlined what I believe to be the most significant problems existing in Australia's health industry, and I have tried to explain both the proximate causes and the underlying causes of these problems.  It is no easy matter to rank the proximate causes in order of importance, but my guess is that the poorly-designed insurance system has the most severe negative impact.

Governments' direct involvement in the provision of health care probably ranks second, but very significant improvement could be had simply by requiring all government health care institutions to operate along profit-oriented lines -- provided their managements are given the autonomy to make most of their own operating decisions.  Restraints on competition, especially those which prevent a greater supply of new entrants to the practice of "simple medicine", are also to blame for making health care more costly than it need be.

I have argued that the underlying cause of these problems has been the failure to specify exactly what we wish to attain through conscious policy intervention in health, and to investigate with sufficient care different means of achieving any given objective in this field.  This has led to the multiplication of difficulties and to an ever-widening set of regulations and other forms of government intervention -- to an extent which, I think, justifies the assertion that the Australian health system is a shambles.

My intention has been not only to explain the deficiencies of the system, but also to suggest a single new policy instrument, which would allow almost the entire health industry to be put in the hands of the private sector, and most forms of governmental controls to be removed, without compromising the community's desire to protect the indigent.  This policy involves making health insurance compulsory, while also subsidising its purchase for those in the higher-risk groups who are too poor to afford it.  The proposal is quite radical, but I hope this will not prevent it from being debated on its merits.

---

ENDNOTES

1.  The term "elective" should not be interpreted to mean "trivial".  It refers to non-emergency surgery, but it should be appreciated that such surgery may be necessary on medical grounds, and something that the patient himself regards as quite urgent.

2.  Relevant data can be found in Australian Health Expenditure, 1979-80 to 1981-82, published by the Australian Institute of Health, and in the World Development Report, 1985, by the World Bank.

3.  See, for example, "Medicare:  the cure we can no longer afford", by Paul Gross, Director of the Institute of Health Economics and Technology Assessment, in The Australian, March 15 1986.

4.  The point has been made before by M. Pauly, "The Economics of Moral Hazard:  Comment", American Economics Review, June 1968.

5.  Insurance against an act of nature (such as rain on the day of an outdoor concert), where the insured's behaviour cannot affect what happens, is an exception.

6.  The argument here has been developed by John Logan, "Australia's Health:  An Epidemic of Regulation", paper presented to the Montpelerin Society, Sydney August 1985, p.5.

7.  Health Insurance Commission, Annual Report for 1984-85, Tables 7 and 9.

8.  Mark Lawson, "A big leap forward in patient diagnosis", Financial Review, 21 February 1986.

9.  Commonwealth Department of Health, Annual Report for 1984-85, p.28.

10.  Cost to the hospital may appear to rise, if we look only at the sessional payment to doctors chosen by it.  This was at the heart of the doctors' dispute in NSW last year.  The effective payment to doctors was of course much higher than the sessional payment, since there was an implicit agreement that the hospitals would permit the doctors to also treat a number of their much more remunerative private patients, using the hospitals' facilities.

11.  See K.J. Arrow, "Uncertainty and the Welfare Economics of Medical Care", American Economic Review, December 1963.

12.  This is one factor which has contributed to the private hospitals' image as a group of second-rate institutions.  With a fixed per bed day rate of insurance cover, it is hardly surprising that the hospitals -- especially those not operated by non-profit organisations -- concentrated their attention on the least complex (cheapest) medical and surgical cases, and made little effort to keep down the patient's length of stay (since average cost per day tends to decline, the longer the patient remains in hospital).  The opportunity of making a profit from building high quality hospitals able to accommodate even the most difficult cases, and thus the incentive to do so, hardly existed.

13.  Data can be found in "Australian Health Expenditure", op cit.

14.  Loc cit, Table 4.03.

15.  Canberra's two older public hospitals have so far failed to obtain accredited status with the Australian Council on Hospital Standards.  The one mentioned here (Calvary, funded by the Government but operated by the Congregation of the Little Company of Mary) is accredited, as is the one private hospital.

16.  Rather, it seems to be the task of ad hoc official inquiries of one kind or another.  See, for example, the Auditor-General's Efficiency Audit of Canberra's public hospitals in 1983.

17.  See World Development Report, 1983, by the World Bank, Chapter 8.

18.  T.S. Pensabene, The Rise of the Medical Practitioner in Australia, ANU Press, Canberra, 1980, p. 123.

19.  See World Development Report, 1985, by The World Bank, Table 24.

20.  The following argument is developed in some detail in John Logan, "A Brief Exploration into the Anatomy of the Medical Profession:  the Market for GP Services", in R. Albon and G. Lindsay (eds), Occupational Regulation and the Public Interest, Centre for Independent Studies, Sydney, 1984.

21.  Loc cit, p.53,

22.  Relevant data, now rather out of date, can be found in the Commonwealth Department of Health's 1980 publication, Handbook on Health Manpower.

23.  The argument has been made that doctors are indeed special, because they are supposedly able to have a strong influence on the demand for their own services.  (See J. Richardson, "A Model of Doctor Practice:  An Empirical Analysis Using Sydney Survey Data", in the Proceedings of the Second Australian Conference of Health Economists, ANU Press:  Canberra 1981.)  This view has met with little acceptance amongst other economists.  Recent medical graduates, struggling to earn anything like the amounts they had expected at the time they decided to study medicine, would presumably question this line of thought also.

24.  See, for example, J.S. Deeble, "Unscrambling the Omelet:  Public and Private Health Care Financing in Australia", in G. McLachlan (ed.), The Public-Private Mix in Health:  the Myth and the Reality (1982), pp.433-434.

---

APPENDIX 1

FRONT-END DEDUCTIBLE INSURANCE

Insurance does little if anything to reduce the total cost of whatever is being insured against, but merely ensures that the costs are spread around the pool of those insured, so that no single person need be concerned about incurring some intolerable expense.  All actuarially-rated insurance is an unfair bet -- in the sense that the premium paid must exceed the expected value of costs which may be incurred, because of the need to build in a charge to cover the additional costs of administration involved in providing this insurance -- that is, the costs of labour, capital and other inputs used by the insuring organisations.

People are prepared to accept these unfair bets if the possible costs which might be incurred are large.  In other words, they are prepared to pay, in order to remove the risk of incurring an intolerably high expense.  But one would imagine they would not be prepared to do so if the possible outlays were small.  That is, it seems irrational to insure against the possibility of small losses.  According to this argument, front-end deductible (FED) insurance -- insurance against large outlays rather than against all outlays -- should be preferable to full cover insurance.

It is of interest, then, to look closely at the comparison between insurance packages offering full cover, and those offering only cover for expenses over a certain level.  The example chosen here for illustration is that of Medibank Private's Basic Private packages.

For full cover for a family, the rate at the time of writing is $374.40 per annum.  With a front-end deductible of $300, the rate drops to $222.56 -- a saving of $151.84 per annum.

In the first case, the maximum outlay is the premium ($374.40).  In the second, it is the premium ($222.56) plus the worst-case expense ($300) -- a total of $522.56.  That is, by purchasing the cheaper package, the family runs the risk of having to spend as much as $522.56, compared with only $374,40 (a difference of $148.16) if it buys the more expensive package.  But recall that it saves $151.84 in premiums.

The amount at risk is really quite small for most people (by comparison with their incomes), so it might be expected that many families would prefer the front-end deductible package.

In practice, it seems that relatively few do so.  One possible explanation for this is that the advantages of FED packages have not been clearly explained by the health fund.  A second is that the probability of actually incurring expenses in excess of $300 is relatively high, making the gamble unattractive.  This point is elaborated later.

A third is that the full cover package can be interpreted as containing a forced saving element, which may be seen as desirable by many families.  In the FED case, there is always the possibility that there will occur an outlay of up to $300 p.a. (over and above the premium).  Families could of course build up this amount over time in some kind of savings account:  indeed, they could use the saving on premiums for this purpose.  But it seems that many people save more effectively if there is some element of compulsion or commitment -- as can be seen in the popularity of life insurance and superannuation schemes.  Thus the full cover premium can be seen as containing contributions which will almost certainly be "withdrawn" at a later date.  In other words, this allows the insured to "save up for a rainy (unhealthy) day", at the same time that insurance is purchased.

The main point of interest here, however, is the question of the trade-off between saving on the FED premium by comparison with that for full cover, and the risk of having to face a higher total payout.  The relevant factors here are:

• the size of the premium discount;
• the additional payout that may have to be faced (presumably, relative to income);
• the probability or likelihood of incurring this higher payout.

To assist in explaining these points, Table 1 should be helpful.  It is based on a hypothetical frequency distribution of health costs (Column 1) incurred over a year by all the individuals in a total population of 1 million.  This distribution (Column 2) is based on past experience with similar populations, and insurance premiums are calculated on the assumption that the distribution is unlikely to change much from year to year.

Premiums (Column 3) are calculated on the basis of the expected level of insured health costs, plus an assumed mark-up of 7% to cover the costs of administering the insurance scheme.  (The "expected" level is the average of all possible costs, weighted by their respective probabilities of occurrence.)  The premium in the first row ($1,694) is that charged for full insurance of all health costs.  Premiums in the following rows are those charged for FED insurance, where the insured agrees to meet the cost of the first $x of costs -- $x being the amount shown in the first column.  For example, for a FED of $1,000, the premium falls to $1,436.  The premium discount ($258) is shown in Column 4.

Column 5 shows the maximum possible amount the individual would have to pay -- for health care and insurance combined -- depending on the level of FED chosen.  For example, with a FED of $1,000, the premium is $1,436, so the maximum exposure is $2,436.  By comparison with the outlay for full insurance, there is now the possibility of an increase in total outlay of $742 (the difference between the maximum possible outlay and the full insurance premium).  These figures are shown in Column 6.

We now know the possible savings from purchasing FED insurance, and the risk to which this gives rise.  At this point, personal preference must be brought to bear on the problem, for it is the individual's attitude to risk which causes him to choose a high premium with little or no risk, or a lower premium with a correspondingly higher risk.  For example, some individuals will prefer to pay the full premium of $1,694, so that they can be sure they will not need to incur any further health costs.  At the other extreme, some people will be prepared to gamble against ill health, by choosing to purchase no insurance.

Expected levels of uninsured health costs for different FED levels (i.e. the average of possible costs -- up to the FED level -- weighted by their respective probabilities of occurrence) are shown in Column 7.  The probable net saving consequent upon the purchase of FED insurance (Column 8) is the difference between the FED premium discount and this expected level of uninsured costs.

As a matter of interest, people who do not purchase insurance can expect to save about $111 -- which is the insurance administration overhead (7%) based on the average level of expenditure for people in the group as a whole ($1,583).  Likewise, the expected savings from purchasing any given FED package is always equal to 7% of the expected level of uninsured health costs.

One way of summarising the premium discount vs risk trade-off is to focus on the probability of incurring the maximum possible outlay.  It should be stressed that it is not necessary to become extremely ill in order to face the maximum outlay.  A degree of ill health sufficient to incur costs equal to, or in excess of, the front-end deduction is all that is required.

Cumulative probabilities (Column 9) are derived from the frequency distribution in Column 2.  By summing the numbers of people in all the rows up to and including the row of the FED level in question, we can easily calculate the proportion of individuals who are likely to incur costs less than that level.  The figures in Column 9 show this in reverse:  that is, the proportion of individuals incurring costs this high or higher.  For example, the probability of incurring costs of $1,000 or more is 62%.

To summarise the content of Row 6:  based on past experience, the number of people with $1,000 of health costs in a year is 87,583 (from a total of 1 million).  The premium for insurance against health costs in excess of this amount is $1,436, representing a discount of $258 from the full cover premium of $1,694.  People choosing this level of cover would have to pay a total of $2,436 in the worst possible case, which exceeds the full insurance premium by $742.  The probability of this occurring is 62%.  The probable (or expected) level of uninsured costs is $241, so the probable net saving from choosing the FED package rather than full insurance is $17.

We can make some predictions about the kinds of people who are more likely to opt for lower levels of cover.  Most obviously, these will include people who are not particularly risk-averse.  As well, some people -- particularly those with low incomes -- will tend to be concerned about the level of the premium relative to income.  Supposing, for example, that they work to the rule that the premium should not account for more than 10% of their income, then such people will effectively constrain themselves to choosing lower levels of cover.

Nevertheless, such generalisations need to be made with great caution.  A counter-argument is that people with relatively high incomes -- who feel they could afford to meet the moderate health costs which may be incurred if they are not fully insured -- may well decide to choose a FED package.

Another important factor in the individual's decision whether to opt for FE deductibility is his perception of his own health risk.  In general, individuals do not have access to the kinds of data bases built up by the health funds, showing the frequency distribution of health costs incurred by people in the same risk category.  They are therefore effectively forced to make their own, highly subjective, guesses as to what these distributions are like (although few would think of the problem in these terms).  Presumably they do so on the basis of knowledge of the kinds of medical problems suffered from time to time by people of their acquaintance, publicity given to victims of accidents and so on, as well as to their own recent health experience.

To illustrate the importance of peoples' own self-estimated probability distributions of health costs, we can substitute some hypothetical distributions in place of the more objective one shown in Column 2 of Table 1, while retaining the schedule of premiums already calculated.  The resulting changes will appear in Columns 7 and 8.

Thus, Table 2 uses a subjective probability distribution of someone who regards himself as a low health risk.  By comparison with the objective distribution, he thinks himself much more likely to incur relatively low health costs (up to roughly $1,400), and much less likely to incur high health costs.  Thus in Column 7, the expected level of uninsured health costs is high in the low range of costs, relative to the group as a whole.  For example, Row 6 shows that the individual expects to incur health costs up to the $1,000 FED level of $384, compared with only $241 for the whole group.  The converse is true in the high range of health costs.

In turn, this means that the expected net saving from the purchase of FED insurance actually becomes negative if the deduction is in the low range of health costs, while it is positive and considerably higher than Table 1 would indicate for FEDs of $2,400 and over.  It can be appreciated, therefore, that since it is not usual for health funds to offer a whole range of FED levels (as shown in these Tables), the level of FED actually offered is of crucial importance in determining the popularity of FED insurance.  In particular, if there is only one FED option (as is the case with the Medibank Private example and with all of the other health funds for which information is to hand), and if the FED level is relatively low, then it can be expected that few low-risk individuals will choose this option.

We can easily undertake a similar kind of simulation for people who think of themselves as high health risks.  Table 3 has been constructed in the same way as Table 2, using a high-risk frequency distribution for health costs in Column 2.  As can be guessed from the preceding discussion, the relatively low FED levels are the ones which are more attractive for this type of individual.

Note, however, that the expected net savings in this case are quite small -- the maximum being only $127 for a FED of $1,600.  Furthermore, there is a 53% chance of having to face the maximum possible outlay of $2,697.  By contrast, the low-risk individual could expect (for example) to save about three times this much with a FED of $3,200 -- with only a 1 % chance of incurring the maximum outlay of $3,483.

These comments reflect the fact, of course, that community-rated (CR) insurance taxes low-risk people and subsidises those who are high risks.  Those who perceive themselves to be low risks seek to avoid at least some of the tax by purchasing less insurance.  Those who are high risks, on the other hand, would lose part of their subsidy if they chose this course of action.

It may be concluded, then, that if the FE deduction offered by the funds is relatively low, the introduction of FED insurance will not be particularly attractive to low-risk individuals.  It will be somewhat more attractive to the high-risk groups in terms of their expected net savings, but these savings are really quite modest when weighed up against the risk of extra outlays to which the relevant individuals would be exposed.  It may be the case, then, that the rather cool response of the public to the introduction of FED insurance reflects an implicit complaint that the FE deduction level typically offered is too low.

FRONT-END DEDUCTIBLES AND COMMUNITY-RATED INSURANCE

In September 1985, new regulations were introduced by the Minister for Health which allowed the funds to offer FED packages.  (The Medibank Private example used above is that organisation's response to this new freedom.)  The Government's continuing commitment to the CR requirement was also re-emphasised.  The analysis above suggests some important implications for these new arrangements.

Already, a proportion of fund contributors have chosen to switch from "first dollar" cover to the FED packages now on offer.  As well, presumably, some people who previously chose to risk not being insured (other than by Medicare) rather than pay the high cost of full insurance, have opted for partial insurance now that it is available.  It would appear however that, as yet, the number of people involved is small.

Nevertheless, it is interesting to speculate on future developments.  The funds are required to base their premium for the FED option on actuarial experience with their contributing membership as a whole -- that is, presumably, to calculate their premiums as has been done here in Table 1.  They are free to choose the level of the deductible they offer, and presumably have done so on the basis of some kind of market research.  Within their total membership, there are those who regard themselves as relatively low risks, but who are sufficiently risk-averse to purchase insurance anyway, knowing that they are subjecting themselves to the tax just mentioned.

As was shown in the discussion relating to Tables 2 and 3, such people are much more likely to opt for the FED package, provided the FE deduction is high enough.  And, as noted also, at least some previously uninsured low-risk people are likely to purchase newly available FED insurance.  In other words, there is likely to be a preponderance of low-risk people purchasing this kind of cover.  In turn, what this means is that FED insurance will be highly profitable for the funds (so long as the old rates continue in force).

Since the funds compete with each other for market share and profits, it can be expected that there will be a tendency for premiums for FED insurance to be reduced (see further discussion on this point below) and/or for marketing efforts to be directed to the low-risk groups.  The effect of this will be to draw even more low-risk individuals into FED insurance -- leaving the full insurance category populated largely by the high-risk group.

The funds will therefore be induced to raise rates for full insurance, because of the increased dominance of high-risk individuals amongst those buying it.  There will be some tendency for people in these groups to switch to FED insurance because of the change in relative price of the two kinds of package, but the risk mix of contributors to the two types of package will retain at least some of the bias which emerges initially.  It follows that the degree to which community rating will be able to effect a transfer from low-risk to high-risk individuals and families will be reduced -- perhaps to a significant extent.

Changes in the price of full insurance relative to FED insurance would be contrary to the Government's desire to maintain the CR principle.  Whether it would be able to prevent this from happening, however, is a moot point.  In practice, CR is easy to enforce when there is only one insurance package on offer, but when there are more than this, there is scope for argument regarding the "correct" relativity of premiums.

The Government would have to monitor claims frequency distributions in order to be able to satisfy itself that the FED premium discounts offered by the funds were indeed based on the claims of their total membership, and not particular risk groups within it.  This in itself would be no easy task:  there would be no claims from many contributors to FED packages, since the level of costs they incur will be below the level of the deductible.  Interesting questions would arise in the event that the distributions proved to differ across different funds -- as would almost certainly be the case.

Funds most successful in marketing FED insurance to low-risk groups would be able, presumably, to justify more realistic discounts for these groups, so they would rapidly gain market share at the expense of those slower to market their FED packages.  A process such as this could see the market being divided up into funds specialising in FED insurance for low-risk groups and those specialising in full insurance for high-risk groups.  The CR principle could not be applied effectively in such circumstances.

In short, this whole issue could create a whole new area of data collection, analysis and negotiation which would no doubt provide employment to yet another large group of health industry bureaucrats, both in and out of government!

A final point to note is that Medicare is effectively a compulsory CR insurance scheme, in which the premium is related to income rather than being the same for all contributors.  As long as it continues to exist, subsidisation of high-risk by low-risk groups will continue;  it amounts perhaps to overkill to insist that additional, optional insurance should still adhere to community rating.

TABLE 1

Health Insurance Simulations:  No Differentiation of Risks

1 Health Cost	2 Freq Dist	3 Prem	4 Disc for FED	5 Max Outlay	6 Incr in MO	7 Expect UHC	8 Expect Net Sav	9 Cum Prob%
0	19824	1694		1694		0	0	98
200	34366	1686	7	1886	193	7	0	95
400	69538	1657	37	2057	363	35	2	88
600	82620	1604	90	2204	510	84	6	79
800	86930	1529	165	2329	635	154	11	71
1000	87583	1436	258	2436	742	241	17	62
1200	83524	1328	366	2528	834	342	24	54
1400	75322	1215	478	2615	922	447	31	46
1600	69149	1097	597	2697	1003	558	39	39
1800	61920	978	716	2778	1084	669	47	33
2000	54560	861	833	2861	1167	778	54	27
2200	47996	748	946	2948	1254	884	62	23
2400	41609	641	1053	3041	1347	984	69	19
2600	35840	541	1152	3141	1448	1077	75	15
2800	30688	450	1244	3250	1556	1163	81	12
3000	26901	363	1331	3363	1669	1244	87	9
3200	23422	283	1411	3483	1789	1318	92	7
3400	19944	210	1483	3610	1917	1386	97	5
3600	16466	147	1547	3747	2053	1446	101	3
3800	12987	94	1600	3894	2200	1495	105	2
4000	9725	53	1641	4053	2359	1534	107	1
4500	4381	32	1662	4532	2838	1554	109	0
5000	2221	20	1674	5020	3326	1565	110	0
6000	1134	12	1681	6012	4319	1571	110	0
7500	792	6	1688	7506	5812	1577	110	0
10000	559	0	1694	10000	8306	1583	111	0
Total  1,000,000

TABLE 2

Health Insurance Simulations:  Using Subjectively Estimated Frequency Distribution for Low-Risk Group

1 Health Cost	2 Freq Dist	3 Prem	4 Disc for FED	5 Max Outlay	6 Incr in MO	7 Expect UHC	8 Expect Net Sav	9 Cum Prob%
0	26453	1694		1694		0	0	97
200	52907	1686	7	1886	193	11	-3	92
400	132267	1657	37	2057	363	63	-26	79
600	145494	1604	90	2204	510	151	-61	64
800	136676	1529	165	2329	635	260	-96	51
1000	123449	1436	258	2436	742	384	-125	38
1200	105814	1328	366	2528	834	511	-145	28
1400	79360	1215	478	2615	922	622	-143	20
1600	61725	1097	597	2697	1003	720	-124	14
1800	44089	978	716	2778	1084	800	-84	9
2000	29099	861	833	2861	1167	858	-25	6
2200	19399	748	946	2948	1254	901	45	4
2400	13227	641	1053	3041	1347	932	120	3
2600	8818	541	1152	3141	1448	955	197	2
2800	6172	450	1244	3250	1556	973	272	2
3000	4409	363	1331	3363	1669	986	345	1
3200	3527	283	1411	3483	1789	997	414	1
3400	2645	210	1483	3610	1917	1006	477	0
3600	1764	147	1547	3747	2053	1012	534	0
3800	882	94	1600	3894	2200	1016	584	0
4000	617	53	1641	4053	2359	1018	623	0
4500	441	32	1662	4532	2838	1020	642	0
5000	309	20	1674	5020	3326	1022	652	0
6000	220	12	1681	6012	4319	1023	658	0
7500	150	6	1688	7506	5812	1024	664	0
10000	88	0	1694	10000	8306	1025	669	0
Total  1,000,000

TABLE 3

Health Insurance Simulations:  Using Subjectively Estimated Frequency Distribution for High-Risk Group

1 Health Cost	2 Freq Dist	3 Prem	4 Disc for FED	5 Max Outlay	6 Incr in MO	7 Expect UHC	8 Expect Net Sav	9 Cum Prob%
0	16255	1694		1694		0	0	98
200	24382	1686	7	1886	193	5	2	96
400	35761	1657	37	2057	363	19	18	92
600	48765	1604	90	2204	510	48	42	87
800	60143	1529	165	2329	635	97	68	81
1000	68270	1436	258	2436	742	165	93	75
1200	71521	1328	366	2528	834	251	115	67
1400	73147	1215	478	2615	922	353	125	60
1600	73147	1097	597	2697	1003	470	127	53
1800	71521	978	716	2778	1084	599	117	46
2000	68270	861	833	2861	1167	735	97	39
2200	63394	748	946	2948	1254	875	71	33
2400	56892	641	1053	3041	1347	1011	41	27
2600	50390	541	1152	3141	1448	1142	10	22
2800	43888	450	1244	3250	1556	1265	-21	17
3000	39012	363	1331	3363	1669	1382	-52	14
3200	34135	283	1411	3483	1789	1492	-81	10
3400	29259	210	1483	3610	1917	1591	-108	7
3600	24382	147	1547	3747	2053	1679	-132	5
3800	19506	94	1600	3894	2200	1753	-153	3
4000	14629	53	1641	4053	2359	1811	-170	1
4500	6502	32	1662	4532	2838	1841	-178	1
5000	3251	20	1674	5020	3326	1857	-183	0
6000	1625	12	1681	6012	4319	1867	-185	0
7500	1138	6	1688	7506	5812	1875	-187	0
10000	813	0	1694	10000	8306	1883	-190	0
Total  1,000,000

---

APPENDIX 2

EFFICIENCY COMPARISON OF GOVERNMENT HEALTH CENTRES IN THE A.C.T. AND PRIVATE G.P. PRACTICE

These notes describe the results from a comparative study of the efficiency of two government health centres operating in the ACT and a sample of private GP practices drawn from all states.

The Capital Territory Health Commission (CTHC) operates twelve Community Health Centres in Canberra, the first of which opened in 1973.  The intention of these centres was to "decentralise social workers, district nurses and other associated health workers by placing them in community health centres, where they might be more closely associated with general practitioners" (Director General of Health, Annual Report for 1972-73, p.89).

Two of these centres were chosen for detailed study, one of which is quite small and is administered largely as an annex of the other.  In this report, the two centres are considered as one (that is, data from their operations are aggregated).

Quite a wide range of health professionals are involved, but GP practice dominates.  In the year selected for study (1983-84), there were based at the centre 4 full-time and 2 part-time GPs backed by 2 full-time and 1 part-time nurses, 1 dentist, 2 part-time physiotherapists, and 1 social worker.  A privately-owned pharmacy was also in operation (space being rented from the centre).

Various other health professionals employed by the CTHC visited the centre for a few hours each week, including nurses dealing with infant welfare and immunisation and a dietician.  There was also provision for visiting (private practice) specialists, but this seemed a fairly minor activity.

Of the total of all occasions of service to patients ("patient contacts"), doctor and/or nurse contacts (excluding infant welfare and immunisation) constituted the majority -- about 67%.  The nurse contacts referred to here, moreover, are concerned primarily with the kind of consultations or treatments which, in the absence of nurses, would typically be dealt with by GPs.  It is justifiable, therefore, to regard the major activity of the centre as the provision of the same kind of services as would be provided at private sector GP practices.

In turn, this raises the possibility of making a fairly direct comparison of the operating efficiency of general practice in the government and private sectors, provided that the costs of running the government health centre can be sufficiently disaggregated to be able to allocate them to either GP or other health services.  Fortunately, this has been possible.  Information concerning the government health centre has generously been supplied by the CTHC.  Data on the operations of private practice have also kindly been made available by the Royal Australian College of General Practitioners (RACGP), which undertook a survey of a randomly chosen sample of 155 practices throughout Australia in October 1984 for the year 1983-84.

The RACGP survey classified practices as to their size -- solo practice, small groups (with from 2 to less than 5 full-time-equivalent doctors), and large groups -- and their location -- city, suburban, provincial city, country town, and rural.  For the purpose of the present study, it was decided to exclude data from solo practices, and to use observations from only suburban and provincial city locations (that is, to try to include only practices of a similar size and location type to the government centre chosen for comparison).

The number of practices which could be included in the sample had to be reduced further, however, since not all respondents supplied information on the number of patient contacts for the year.  As noted below, these data were essential to the analysis.  Additional observations had to be discarded because preliminary analysis suggested that some of the more important data were inaccurate (resulting in unrealistically low or high numbers of patient contacts per doctor, doctor incomes, and so on), leaving a total of 31 usable observations.

The approach followed has been simply to compare efficiency in terms of cost of inputs per unit of output.  In the case of self-employed doctors (that is, the owners of the practices), their "cost" is taken to be their revenue net of all practice expenses.  It can be noted in passing that average partner income from the RACGP survey observations used here is about 12% higher than the average total payment to doctors at the government centre.  This cost advantage for the government centre is much more than offset, however, by the fact that the private sector doctors see about 50% more patients on average.  "Output" is taken to be the patient contact.

Of course, it may be objected that this is not output at all:  what the patient desires is output in the form of recovery from illness or injury, and for a given cost outlay, one practice may well be more effective than another in terms of curing patients of what ails them.

However, to push this line of argument to its logical extreme, for each occasion of service we would need to know "how sick" was the patient, and how great was the "degree of recovery" following treatment -- and perhaps even the extent to which recovery was due to treatment as distinct from natural recuperative capacity.  Such things are not readily susceptible to measurement, so this does not seem a very fruitful basis for empirical research.

More to the point, there is no reason to believe that the effectiveness of treatment received by patients at government health centres is systematically different from that received at private practices.  If there were such a difference, it could be expected that there would be a steady drift of patients towards the one or the other as the result of word-of-mouth "advertising".  No such drift is evident -- which indicates that, from the point of view of the health service consumer, output quality is much the same.

Another possible objection concerning comparability of the two sets of data is that the "case-mix" may differ systematically as between the health centre chosen for study and the "typical" GP practice represented by the aggregated data from the RACGP survey.  I have no reason to believe that the government centre deals with health problems that are on average more or less complex than those encountered elsewhere.

The results of the study are contained in the attached Table, which presents estimates of costs per patient contact.  These costs are subdivided into various components -- the most important of which are the various categories of labour (doctors, nurses and other medical staff, and non-medical staff).  Other cost components need to be treated with caution, since they do not always mean the same thing for different practices.

Comparison of the first and second columns indicates quite a large cost difference in the provision of GP services in favour of the private sector -- $14.68 per patient contact (PC), compared with $19.34.  This amounts to an excess of about 32%.  All categories of labour are substantially less costly in private practice.  Other costs are slightly higher, but this is largely explained by the inclusion of various costs which are not included in the government health centre costs, such as insurance, depreciation, interest, bank taxes, bad debts written off, and debt collection charges.  The difference in non-labour costs is more than offset by these items.

These comments do not apply only in respect of the average figures based on the private sector survey.  Even if we look at the range of observed values across the sub-samples of GP practices from the various states, we still find the government centre operating at a cost disadvantage.

Insofar as the provision of straightforward GP services is concerned, therefore, the government health centre appears to be significantly less efficient in the use of resources than private GP practices.  By contrast with many private practices, however, the health centre employs nurses who are able not only to assist doctors, but also to deal with many of the simpler patient problems themselves (that is, independently of the doctors).

In the absence or nurses, these cases would probably be handled by the GP, and so it can be argued that patient contacts with nurses alone should also be treated in this analysis as GP patient contacts.  To put it another way, excluding nurse-only patient contacts could be expected to introduce bias regarding the case-mix facing GPs at the government health centre.  When this is done, the results are as shown in the third column of the Table.

Note that the data for nurse PCs also record whether they were in conjunction with a doctor or not.  The figures in the second column take into account only the former figure, while those in the third include both.  Accordingly, total nurse salaries are included in Column 3, but only a proportion of this total -- based on the share of nurse PCs in conjunction with doctors as a share of total nurse PCs -- is included in Column 2.

It can be seen that the employment of nurses to handle some of the cases which would otherwise be presented to GPs results in an improvement in the cost figures for the government health centre.  Labour cost per PC falls by more than $1, and total cost by $1.16.  Nurse salaries are less than half those for doctors.  Nevertheless, the private practices still have a substantial cost advantage, amounting to $3.50 per PC, or 24% (without taking into account the few aspects of non-comparability mentioned above).

Comparison of Costs per Patient Contact ($ per PC)

	Private Practice	Government Health Centre:
		Nurse-only patient contacts
		Excluded	Included
Doctors Medical Staff Non-medical Staff     Total Labour	7.67 0.70 2.51 10.88	11.05 1.36 3.45 15.86	9.11 2.24 3.46 14.81
Consumables Occupancy Administration Vehicles & Transport Other     Total Non-labour	0.36 1.05 0.64 0.61 1.13 3.81	0.40 1.61 0.52 0.94 0.03 3.49	0.33 1.71 0.52 0.79 0.03 3.37
Total Cost	14.68	19.34	18.18

There are advantages of community health centres over the typical private GP practice, aside from the efficiency gain resulting from having clinic nurses take over some of the doctors' workloads.  These stem from the fact that, in the former, several different kinds of health professionals are brought together under the one roof.  The composition of the team varies across health centres.  With a significant proportion of cases there is interaction between doctors and the others -- in particular, with physiotherapists, clinical psychologists, social workers and community nurses.  Again, the degree to which this goes on depends on the working atmosphere built up in each individual centre.  There would seem little doubt that the "team approach" to patient problems often results in more appropriate treatment.

Unfortunately, the private sector has very little incentive to operate in similar fashion to this, because Medicare benefits are not available for these paramedical services.  This means that a doctor could only refer his patients across to hypothetical paramedical colleagues, knowing that the cost of the additional consultations would have to be fully borne by the patient.  At the government health centres, these services are not charged to the patient.