Health insurance and the tragedy of the common risk pool (Doc Holliday)

In my earlier posting (8 October 2009) I discussed the adverse effects of community rating on the efficiency of health insurance markets, and put forward a case for a modified voucher system for health care (which in the Australian context, I will call Medicare Savings Accounts or MSAs). In this posting I examine the tensions between health insurers and medical care, and propose some policy solutions.

A heated point of discussion in the health insurance debate is how to reimburse (that is, pay) doctors and hospitals. The traditional method of payment is the ‘fee for service’ – in which a doctor and a hospital is paid for the services they render. To an economist, this model appears to be fairly obvious – as a fee for service is basically the price of providing a service. Yet in the context of health care it has created all sorts of strange notions and policy prescriptions which are ultimately ill-conceived and self defeating.

The case against fee for service is that, put simply, doctors have an incentive to over-service their patients. The economics of this incentive is rarely put forward with any clarity by its proponents. They generally appeal to notions of ‘information asymmetries’ and ‘market power’. But none of these are adequate explanations. For example, market power assumes that doctors reduce their supply of services - but this is the opposite of what has been observed. Similarly, information asymmetries lead to adverse selection (Akerlof’s lemons problem), something which the medical colleges seem to adequately addressed.

The only explanation that makes sense to me is that over servicing is part of the wider problem of contractual enforcement - doctors have an incentive to charge and serve ‘that little bit more’ because insurers are unable to closely monitor doctors’ actions. Another way of looking at it is to think of pooled insurance being a 'tragedy of the commons'. The insurance pool is a common resource which doctors do not have any incentive in preserving. Rather, they have an incentive to encourage consultations and prescriptions believing that they are acting in the interest of their patient.

This problem is not unique to health insurance and examples can be found in other insurance markets. Consider the analogy of an automotive panel beater. I have lost count of how many times I’ve approached a panel beater and asked him to give me two quotes for the same job – a cash job and an insurance job. The quote for the insurance job is always more costly and extensive (in terms of work undertaken). The same logic applies in medicine. This phenomenon is analogous to moral hazard, but whereas moral hazard is normally associated as the patient's response to imperfect contractual arrrangements, this squarely puts it in the lap of suppliers.

One obvious implication is that free-riding on the insurance pool will give each patient additional care, but it is not clear whether the additional services necessarily lead to valuye-for-money. Another implication is that the logic of the tragedy of the commons applies to both private health insurance and to single-payer government fund insurance schemes.

My observations are not in any way new. Commentators have long observed the expansion in supply, and it has been dubbed as 'supplier-induced demand'. However, this explanation seems to have the best ring of truth to it.

But if fee for service in a pooled insurance setting leads to free-riding / supplier-induced demand (these terms should be used interchangeably), what have been the policy responses? In Europe, the United Kingdom and the United States, supplier-induced demand has led to the changes in the reimbursement of doctors and an increased monitoring of doctors' practices. In the United States where fee for service arrangements are commonplace, health management organisations (HMOs) have emerged. HMOs enter into contractual relationships with patients and doctors alike, and set limits on the practices of doctors. They can, for example, refuse to fund a course of treatment if it believes is not fiscally responsible. In the UK, the National Health Service has long salaried doctors in hospitals and has recently been changing the reimbursement of GPs. Now, doctors are reimbursed in terms of the number of patients on their books not how frequently they see their patients. In Europe, mutual sick funds have long operated their own hospitals and salaried doctors, and have closely monitored the prescribing behaviour of doctors.

There are similar pressures in Australia. Public hospitals in Australia have traditionally been funded on a capped budget basis, and doctors in public hospitals are salaried. Though some jurisdictions have increasingly adopted casemix funding of hospitals (such as Victoria) (which is essentially a fee for service arrangement) to improve productivity, in NSW at least, a public hospital is now funded according to how many people there are in its catchment area.

However, these policy prescriptions have two obvious deleterious effects. First, the close monitoring of the practices of doctors is in effect equivalent to regulating doctor behaviour. While it might be argued that it is welfare enhancing (relative to free-riding) it is not a first-best policy solution. Telling a doctor he can not give medicine x to a patient with condition y is a blunt (inefficient) instrument and misses the point that patients are highly individual and what might be an appropriate course of treatment for one patient may not be effective for another.

Second, paying doctors and hospitals fixed amounts reduces their productivity. As a community, you want doctors and hospitals to be as productive as possible (technically efficient) as they are very scarce and valuable resources. Yet paying a doctor on an hourly-basis will yield lower levels of productivity than if you were to pay them on a piece-work basis. Consider for example asking two house painters to give you a quote to paint the house: the one that gives you a fixed price quote will get the job done quicker than the one that quotes you an hourly rate. If doctors in public hospitals work long hours, it may not be because of the excessive workload they face, but because of the internal inefficiencies of the hospital caused by the financial incentives the hospital and doctors face.

This is not to say that regulating doctor behaviour and paying doctors on a salaried basis (or funding hospitals with fixed budgets) has no role in the community. But these are at best second-best solutions.

What then is the first-best policy solution? If it is the case that supplier-induced demand is simply free riding by another name, the answer is to ensure that the financial implications (market failure) of doctor behaviour are fully internalised. There are two complementary solutions to this. The first is to empower patients to resist over servicing. In my previous posting I put forward the case of governments creating Medicare Savings Accounts, abolishing community rating and making governments pay the actuarial risk premia to patients into MSAs. Under these reforms, patients will directly directly face the costs of their own health insurance, and will have a strong incentive in protecting their portion of the insurance pool. Doctors will not over service if the patient believes that the additional activity will not yield any additional benefits.

The second approach is not to employ and salary doctors but for them to own the health insurance business. Once doctors own the health insurance arm, they in effect internalise the costs of their own tendency to over service. Indeed, not only do patients have an incentive to look after their own health, but doctors will change their own behaviour away from simply disease treatment to disease prevention.

There is nothing in legislation or regulation that prevents doctors from starting their insurance fund, so in principle, it is possible. In Australia, however, there are a number of financial disincentives that inhibit such an institutional arrangement from arising. Two come to mind. First, there is little incentive for new businesses to enter into the health insurance industry. The presence of Medicare alone has severely curtailed the profitability (and degree of competition) among health insurers. When was the last time you read in the news of the start up of a new health insurer in Australia?

Second, since Medicare arrangements provide doctors with a nice little earner there is no incentive for them to champion a new world order of competition. You can bet your pretty penny that the Australian Medical Association will resist any attempts to reform Medicare. Replacing Medicare with MSAs implies greater competition for health insurance, which of course, doctors will not necessarily support.

This is not to say that doctor-owned insurance businesses do not exist. One standout example is the famous Kaiser Permanente of California. But a case study of Kaiser Permanente is for another posting.

regards

Doc Holliday