Risk-Sharing Fellowships

A New Model for Health Care Insurance

Harvey S. Frey MD PhD JD

There is nothing more difficult to carry out, nor more doubtful of success, nor more dangerous to handle, than to initiate a new order of things.
For the reformer has enemies in all those who profit by the old order, and only lukewarm defenders in all those who would profit by the new...
Niccolo Macchiavelli, The Prince

Few will deny that the U.S. health care system is in serious trouble, but few will agree on the form of an optimum replacement system. There are three major models currently before policy-makers: Indemnity insurance with fee-for-service, managed care, often with capitation, and government-run, "single payer", incorporating either of the first two models. All of these have serious failings, and I here present a fourth model, "risk-sharing fellowships".

This model should be attractive to business, as it relieves them of the responsibility of providing health insurance for their employees, though they may have to increase salaries to compensate for the loss of coverage. It should also be attractive to employees, as it restores their rights under state laws, now preempted by ERISA.

The problem of non-deductibility of individual payments for health insurance premiums should be resolved by the mechanism suggested here, whereby the premiums are converted to deductible state tax payments.


Without specifying the particulars in which each of the current models falls short, I shall start by outlining the desiderata of an optimal system. It is probably impossible to simultaneously completely satisfy all of them. The design of a system consists of giving reasonable weights to these factors, and combining them into a workable organic synthesis. Most political disagreements over health care result from political groups assigning a weight of one to one of the desiderata, and zero to the rest. I present them in the order of my own subjective weights.

1. High Quality care
Medical decisions should be made by doctors and patients on the basis of individualized assessments, with minimal interference by third parties with agendas other than optimal care. There should be no financial incentives for doctors to under-utilize available and desirable procedures. Providers should be able to compete on the basis of quality of care, not merely cost.

2. Humane care
Sick patients should not have to endure the added anxiety and stress of having to fight the system in order to obtain needed care. Benefits and exclusions should be clearly and unambiguously known before a particular plan is joined, and rigidly adhered to. The milieu of medical care, such as waiting times, access to information, and convenience, should be patient centered and patient friendly. Deviations from such patient-friendliness may well be enforced by patients' free choice of plans and doctors.

3. Low cost
Competition is a far more efficient means to achieve reasonable costs than is rigid regulation and intrusive enforcement. Multiple plans, with equivalent benefits, may compete on the basis of cost. Individual doctors may set their own fees, constrained by those of their competitors and what plans will pay. Draining money from medical care, by paying it out to investors or highly paid executives, will increase costs.
There must be incentives against over-utilization. Patients must benefit personally from cost-control, so as to protect against the "moral hazard" of exerting pressure for unnecessary procedures.
Protection against the costs imposed by adverse selection can be achieved by universal coverage.

4. Choice
In order for competition to occur, patients, physicians and plans must have the ability to choose between various options, and enjoy or suffer the consequences of their choices. Thus there need to be many plans, with different costs and benefits, and free choice by patients among them, e.g., with no constraints imposed by circumstances of employment.
Doctors should be able to set the professional and financial parameters of their practice, and patients should be free to knowledgably choose those that suit them best. Competition will be quick to weed out the doctors who choose to be too expensive or uncaring.

5. Universal Coverage
Health care is primarily a personal and familial responsibility. While there is no individual "right" to health care, even the most primitive societies recognize the moral obligation to care for those who can't take care of themselves. Certainly a powerful, wealthy, civilized society such as ours should do no less. Health care, for those unable to assume personal responsibility for it, should therefore be a public responsibility, and society must assure the universal availability of at least some level of care.
A child or accident victim will be treated, in a civilized society, in spite of the absence of insurance. It is unfair for people to take advantage of the good-will of their neighbors by consciously failing to buy health insurance. Those who can afford to pay for themselves and their family should not absorb government or charitable funds meant for the indigent.

6. Self-regulation
An optimal system will try to align the duties of the various parties involved with their self-interest, and will minimize one-sided self-serving behavior by introducing competing incentives for every party. This will encourage them to work with the system, and minimize "gaming" and the need for expensive and intrusive policing.


I believe that the above constraints can best be satisfied by a competitive network of a new kind of entity: fully democratic, cooperatively owned and managed Risk-Sharing Fellowships (RSFs), paying fee-for-service to independent competing health care providers, with government collecting and distributing the premiums, supplementing the premiums of the poor, and setting standards for the RSFs.

A Risk-Sharing Fellowship is a private cooperative insurance company, entirely owned and controlled by its policyholders, who vote directly for all its officers, their salaries, and the terms of insurance policies issued, without the interposition of a Board of Directors. It is not a Republic, but a true Democracy, and its officers and employees must also be policyholders, and subject to the rules they implement and enforce. RSFs may be set up by any group, such as fraternal organizations, churches, towns, or businesses, but will be open to all, and will typically contain several thousand members.

Since the officers will not generally be insurance professionals, the RSF may, indeed must, contract with outside actuarial firms to assure that the policies adopted and the premiums charged comply with government mandated rules of minimal coverage, solvency, and capitalization.

In addition, each RSF must carry two kinds of reinsurance, purchased from competing conventional insurance companies: a very high-deductible catastrophic coverage for each of its members, and an overall reinsurance to cover claims in case of insolvency. The premiums charged for this reinsurance will obviously depend critically on the policies adopted by the RSF and its reserves, and will thus serve as another discipline moderating well-intentioned but risky decision-making. With such reinsurance, there will be no danger in making shares in the RSF non-assessable.

If the RSF makes enough profit, some will be retained to build up reserves, to decrease reinsurance costs. These reserves can be invested in government-approved investment vehicles, with that income further reducing premium requirements for members The amount that a for-profit insurer would ordinarily pay out as dividends, thereby removing it from health care use, is instead paid into each member's medical savings account (MSA), keeping it in the system, and recapturing the 25%-30% otherwise drained from health care.

These MSAs may be used for any medical purpose: co-payments, deductibles, even to decrease subsequent premiums, or for otherwise non- covered medical procedures. They follow a member when he changes from one RSF to another, and could be donated or bequeathed to someone else's MSA. They are immune from judgments, except for medical expenses. They can be invested by the member in approved investments, and of course would not be taxable. Since these MSAs will be perceived as having real value, they will provide an incentive for the members of the RSF to limit their desire to maximize benefits, and instead give some thought to efficient management.
Unlike the managers of conventional insurers and managed care businesses, the owner/insureds of the RSF have an incentive to balance profit versus benefits, rather than to prefer one at the expense of the other, and instead of value being drained from health care by payments to private investors, value is brought in, by investment of the MSA funds and reserves. The retention of this value in the system may be almost enough to provide coverage for the currently uninsured, both figures being 25-30%.

These RSFs compete against one another for members, on the basis of both benefits and profitability. As successful RSFs grow larger, they become more efficient, as the statistical law of large numbers reduces their risk, and therefore their reinsurance expenses. This competition prevents the arbitrary acts possible to a single payer.

Members may select RSFs annually, based on each plans' public explicit and unambiguous disclosure of coverages and exclusions, binding for the year.

Everyone must belong to an RSF of his choice, which he will indicate by a check-off on his annual federal income tax return. The IRS will add the amount of that plan's premium to his tax bill, and forward it to the plan. Since the return will demonstrate the filer's income and dependents, the IRS will easily be able to identify those whose income falls below a prescribed level, and will be able to supplement their premiums. In this way, everyone who files an income tax form will be covered. No doubt, those who receive such government supplementation will have some lower level of choice of plans than those who pay the entire premium. This system also takes care of the problem of the well-off person who nevertheless fails to buy health insurance, and then presents for health care for himself or his family at government expense. If a filer fails to select a plan, the IRS will select the closest cheapest plan for him. If the federal government fails to become involved, the plan could be implemented at an individual state level, with the advantage that if the premiums are added to the state tax bill, they might automatically be deductible for federal tax purposes.

Each plan pays fee-for-service to health providers who charge what they like, controlled only by competition. Patients are free to select any provider, paying any excess over what their plan allows, or bargaining for lower fees.

Every plan must require mandatory proportional (not flat-rate) co- payments. The patient is the one best able to control health care costs. Once a flat copayment is paid, there is no further incentive for that patient to be concerned about increased costs. A proportional co-payment maintains patient oversight for expensive procedures. These co-payments are capped at a level determined by the patient
Plans may nevertheless choose to waive co-payments for procedures which may ultimately decrease health care costs, such as immunizations, cancer-screening, smoking cessation, and weight loss.

If a provider offers to waive a co-payment, patient cost-saving incentive is subverted, so this practice should be made impossible. This is done by having the RSF pay the full allowed amount to the provider, and collecting the co-payment from the patient or his MSA. This system avoids the necessity for policing provider-patient transactions. Any excess payment agreed to by the patient will be paid directly to the provider, possibly by a check drawn on his MSA.

Needless to say, all premiums and co-payments must be fully deductible to insureds. It is irrational to allow medical deductions to employers but not to insureds, especially since these private expenses relieve the government of the necessity of paying for the care itself.

The only way competition can be effective is if the buyers are effectively informed about their choices before making a decision. To this end, each RSF should maintain a data base, accessible to its members, listing its experience with each local doctor and hospital, including costs and patient evaluations. Though patients may not be the best judges of doctor competence, such lists will provide a useful complement to government published lists of hospital process and outcome statistics. In addition, they will provide the incentive for doctors to keep prices down and patient satisfaction up.

As in any human endeavor, RSFs may be affected by incompetence and fraud, so government regulation is a necessity. Government sets minimum levels of solvency and good management. It sets actuarial and risk standards. Since medical records must be confidential, RSF members can't be allowed to review the books and records. Therefore strict audit and bonding requirements must be established for all RSFs.


1. High Quality Care
Government sets minimum mandatory level of coverage. Care is determined by physician and patient, without interference by outside third parties. There are no "gatekeepers", and no second-guessing of doctors. There is no capitation, and no unethical incentives to withhold needed treatment. The risk assumed by the RSF is controlled by government setting minimal membership, and solvency levels and requiring bonding of officers, purchase of reinsurance, and use of actuarial consultants and auditors.

2. Humane Care
The RSF is run by the people it covers. They will have to live with any unpleasant policies they put into place. No uncaring government bureaucrats or self-serving profiteers are involved in medical decision making. The clerks who run RSFs will depend upon their customers for their jobs, so a high degree of courtesy is to be expected. Since there is a multiplicity of RSFs and providers, patients have the freedom to leave those with uncongenial attitudes or policies. Personal MSAs are transferable between RSFs. Since coverage is unrelated to employment, there is freedom to change jobs with no effect on insurance, and there is no ERISA preemption. Patients have full access to the protection of law.

3. Low Cost
No dividends are taken out of health system. Since members vote on executive salaries, they are kept reasonable. Money previously spent on administration is available as reserves. The investment income of the RSF decreases premium requirements. Government regulates permissible investments, for safety. With everyone insured, much more premium income comes into the system, so less needs to come from taxes.Mandatory proportional co-payments, cut demand. Competition among RSFs for members, lowers premiums. Competition among doctors lowers their fees. Member/owner insureds make money in their own private MSAs if claims experience is good, as well as paying lower premiums. It is to the benefit of the insureds to keep costs down.

4. Choice
There is free fully informed consumer choice of RSF and free informed consumer choice of provider. Providers have complete choice of fees and practice parameters, subject to competition, but with the knowledge that their choices will be made known to prospective patients. Consumers control policy terms & price of their own insurance, subject to government standards and market forces, and live with the consequences. This is in distinction to a government-run plan, where patients must accept what government offers, since it's the only game in town, or to a selection from an employer-chosen slate of HMOs, selected for their low cost to the employer.

5. Universal Coverage
Everyone who files a 1040 will have insurance. Those who can afford it pay for it, even if they'd prefer not to. Premiums of indigents are supplemented by the government. All this is done by the agency which has immediate access to the figures, and can most efficiently make the distinction. Coverage is unrelated to employment, so is available to all.

6. Alignment of Duties and Interests
Patients, instead of single-mindedly demanding that "everything be done" have a financial incentive to cut costs as well. Providers, subject to informed competition, have an incentive to lower fees and establish patient friendly procedures and environments. The RSF insurers have the incentive to keep costs and red tape down and benefits up, because they will have to live with the policies they put in place, and benefit from the money they save. Government is not in the position of an unregulated payer, but serves its proper neutral function of regulating private organizations. All this is very different from managed care programs where there is unilateral pressure for low cost, simple indemnity insurance where the pressure is for high cost, or government programs where the pressure is for political advantage.


There is one problem with the above system which is particularly difficult. Of course, no existing or proposed system has a solution which satisfies all the desiderata, either.

RSFs which have a more liberal policy toward a particular expensive disease (such as AIDS or Cancer) than most others, will tend to attract people who already have that disease at each annual RSF selection period. These new members will cost far more to treat than the premiums they pay, since the disease is pre-existing. There are several possible solutions, but none which satisfies all the desiderata, to wit:

1. Exclusion of pre-existing diseases
Once a patient develops a disease not covered by his RSF, he will essentially be locked out of the system. He may join another plan, but his pre-existing disease will not be covered, or may not be for a significant exclusion period. How is he to get needed treatment?

2. Government mandated uniformity of coverage
This destroys the choice and control of insureds over their own plan, thereby destroying competition between plans over benefits and price, which is the essence of the system. Government must specify certain minimal levels of coverage, but not demand uniformity.

3. Assigned risk for pre-existing diseases
This has the disadvantage of rewarding RSFs which choose to give low levels of benefits. The members know that they will be taken care of anyway, thereby putting the risk onto other plans. A plan can't rationally calculate its risks if it may be involuntarily assigned expensive patients at any time.

4. Pools to treat certain diseases
This is of the nature of re-insurance, and if the costs to join are related to each RSF’s treatment policies may be able to be constructed fairly.

5. Make patients live (or die) with their initial choice of benefits.
This has a certain severe fairness to it, but will certainly be unacceptable when it is the children of the decision maker who are affected. While "personal responsibility" is desirable, a death sentence is an unjust punishment for negligence, or even greed.

6. The solution which seems the fairest is to allow a patient with a condition which arose when he was a member of an RSF which didn't cover it, to switch to an RSF which does, upon the payment of (say) 10 years difference in the annual premiums of the two RSFs, and to pay a heightened copayment for (say) one year. This should be enough of an incentive to prevent some RSFs from taking advantage of the largesse of others more generous with benefits, yet doesn't lock individuals out of needed treatment.


Policy is determined by the self-interest of the most powerful players, and this plan, unfortunately, gores the oxen of most of them.
It cuts out direct government involvement, and the many bureaucrats who would be out of a job will lobby vigorously against it.
It cuts out most direct insurance, so the private insurance and managed care industries will surely use their massive wealth to defeat it.
Health care providers may not welcome the advent of informed patients and real competition to their practices.
Many doctors are rabidly opposed to government involvement of any sort, subscribing to a radical libertarianism, except when the government offers to pay part of their bills. The inclusion of the IRS, and government subsidization of insurance premiums for the poor will surely spark their violent opposition.
Control freaks will complain about the power given to each plan to set its own policies.
The wealthy currently covered by Medicare will complain about the means-testing of this plan.
Business will be for it, as it gets them "out from under" the responsibility for providing health insurance for their workers.
And Patients will surely be for it, since it gives them the best possible solution to the problem.

Posted: 7 Oct. 1999. Last revision: 4 Sept. 2006