An Approach to National Health C

Universal, Individual Health Funding Accounts (HFAs): Federally Supplemented from Birth to 21

Guaranteed-Renewable, Incentive-Compatible, Private Health Insurance Policies Encouraged

SUMMARY
This plan is bipartisan since, as Democrats require, it provides for universal coverage. Each person has a Health Funding Account (HFA) which is supplemented by Federal vouchers. (HFAs are similar to Republican-fostered Health Savings Accounts (HSAs) but differ in critical ways.) Purchasing from such Accounts, together with guaranteed-renewable private health insurance policies, the premiums of which may be paid from the HFAs, will foster a free market in health care at the procedure level and hence will lower cost and increase accessibility. All citizens will be able to pay with their own funds for the high quality care needed for their important individual needs. If treatment for the same illnesses continue to have very different costs at different locations in the country, individuals paying from their HFAs will demand that the high-cost providers improve their practices.

When the plan is fully implemented, each person will have an HFA from birth. Their HFA would be supplemented each year, from birth to at least age 21, by a taxable federal voucher of $2000. (HSA funds are not taxable.) During a transition period, all older individuals would also receive vouchers which allowed for pre-existing conditions. The small number of pre-existing conditions at birth could be handled by a high-risk pool. Thereafter, guaranteed renewal should make it generally unnecessary for underwriting for pre-existing conditions to be necessary. Recent innovations in "incentive compatible" insurance should also make it possible to switch insurance companies without need for new underwriting for pre-existing conditions. See below for references and further discussion.

No health insurance, including employer-funded insurance, would be tax-advantaged.

Wage earners would be required to fund their HFA, using after-tax dollars, at 5% of wages up to $5000/year until their HFA balance reached the maximum total of $125,000. Additional money may be placed in the HFA at any time. HFAs should be invested only very conservatively.

All earnings in the HFA are fully taxable as income; this also is different from the treatment of HSAs in current law.

The HFA savings, together with health insurance for those under age 21, are the only mandates in the plan.

The procedures covered by the mandated policies for those under 21 are to be at least those found in 95% of all privately purchased policies for adults. Dental, aural, and vision care for children can be covered by their HFA funds not used for premiums.

HFA funds may be used to pay for private health insurance premiums. This is a change from the current law, under which premiums cannot be paid using funds from one's HSA. Another change from present law is that HFA funds may no longer be used for non-medical purposes. Under current law HSA funds may be used for non-medical purposes if a 10% fine is paid.

The health insurance policy purchased could be of any type, including high deductible and "first dollar". Persons over 21 could choose to purchase no insurance at all.

At start-up, if a citizen applies for a private health insurance policy, Federal vouchers are to be added to each person's HFA each year to cover the 50th percentile marginal cost of any pre-existing conditions revealed by an initial underwriting. This would make more possible the purchase of a high-deductible, guaranteed-renewable plan that includes those pre-existing conditions. The full values of these vouchers would be available only to those individuals that apply at startup. In subsequent years, the pre-existing conditions vouchers would be reduced until being discontinued after the transition period of about 21 years.

Except for those under age 21, and for those in the basic minimum policy (BMP) safety net, there are to be no mandated procedures, although the Federal government should issue guidelines that purchasers may consult.

The guaranteed-renewable policies will also be "incentive-compatible" so that policy holders need not worry about having their premiums raised due to insurance company competition. It is likely that incentive-compatible policies will make it possible for policy holders to transfer to another insurance company regardless of their health experience. Recent research bearing on these relatively new insurance policy concepts is discussed further below. However, here we remark that guaranteed-renewable policies are a good buy for the healthy young. This removes the disinclination of the young to buy community-rated policies which charge much higher rates due to the requirement to subsidize the policies of their elders.

After the transition period, insurance companies no longer will be required to cover pre-existing conditions for new applicants for insurance. But, compensating for this, after the transition period, in the long run, almost all citizens will acquire their life-time policy at birth when pre-existing conditions will be at a minimum.

Genetic testing may make it possible to predict risk of various diseases at birth. However, this might well not lead to different policy premiums because this knowledge itself can lead to relatively inexpensive preventive measures, and it is further likely that, before these illnesses manifest themselves, new treatments may emerge which will be unknown, and hence random, at time of birth. Also, since everyone will die of some illness, the net expense to an insurance company, over a lifetime ... which is relevent for guarantee-renewable policies ... is likely to not vary significantly for a very broad range of genetic profiles.

Lifetime health insurance policies solve a problem which most other health systems experience: young workers are required to pay taxes to fund the high expenses of older retirees. Young workers hence tend to not purchase insurance when they are young and healthy, leading to the classic "free rider" problem.

Such policies also solve the problem of pre-existing conditions, which most current plans solve by an individual mandate coupled with guaranteed issue and community rating. These conditions degrade market mechanisms and also do not eliminate the free-rider problem since individuals, if they become ill, may change from an inexpensive policy to an expensive policy at the annual free enrollment period which most plans afford.

Lifetime policies will obviously also require portability across state lines, so many current restrictions on interstate health insurance sales will have to be eased.

Premiums for high-deductible policies might desirably be additionally subsidized by the Federal government in order to encourage the development of a market for high-cost procedures, independent of insurance companies or government agencies.

It is important from a policy point of view that individuals have insurance coverage so that, due to a serious illness, a minimal number of individuals fall into the Basic Minimum Policy (BMP) safety net, described below.

Employers could withhold a portion of an employee's salary and contribute it to their HFA. They also could pay premiums for a health insurance policy for the employee, and could also supplement employees' childrens' HFAs and pay their insurance premiums. Anyone could contribute to any individual's HFA. For example, parents could contribute to their children's HFAs.

Employers could also, of course, continue to offer smoking cessation programs, exercise facilities, etc. Such programs may minimize sick-leave and lead to a more productive workforce.

Non-wage earners (e.g. the self-employed) could make contributions to their HFA accounts if they so desired, but would not be mandated to do so. This would enable them to devote all possible personal capital to a personal venture such a non profit charity or to a new business.

More than sufficient federal funds are available for these vouchers, and for other parts of the plan, in part because the tax exemption for employer-provided insurance will cease. Also, taxes and fees currently levied for Medicaid, Medicare, including Medicare B and D, and SCHIP will be available, since this plan replaces all of these programs. The total of these federal and state funds is approximately 1 trillion dollars per year. This comes to approximately $3000/person/year, including children. (Another trillion of health care funding currently comes from employers and out-of-pocket.)

It is anticipated that substantially less government funding than one-trillion dollars will be required for equal results when the plan is fully operational, because there will be a universal, fully functioning, well-regulated free market at the procedure level.

Replacing the Medicare and Medicaid programs, which have been achievements of the Democratic party, may well be viewed negatively by Democrats. However, these benefits will actually still be available to all citizens through this new approach, since the plan is universal, supplying medical funding to every citizen.

Because of the need to replace Medicaid and Medicare, there would have to be a multi-year transition.

In brief, during a 21-year transition period, all citizens would receive a Medicaid-Medicare transition voucher of $2000/year. (Those children born in the first transition year would comprise a small portion of this total expense. It is possible that some or all over-21 citizens will also receive a similar voucher after the transition period, at the discretion of Congress.)

In addition, those citizens who become 65 during the first transition year will also receive a yearly Medicare transition HFA voucher sufficient to purchase a high-deductible incentive-compatible, guaranteed renewable policy with benefits actuarially equivalent to the median policy available on the private market. The voucher's value would be adjusted for pre-existing conditions.

Presumably, most of the recipients of these Medicare-transition vouchers will use them to purchase a suitable insurance policy, but they need not do so.

During the following 21 years, each wave of citizens who become 65 will receive a similar voucher for life, except that, for each wave, the voucher's value will be reduced approximately linearly so that it's value will be zero for the 21st wave. For later waves there will be no more Medicare-transition vouchers. These individuals be in the steady state of the new system and will be able to depend on the accumulated funds in their HFA, together with the accumulated value in their guaranteed-renewable, incentive-compatible health insurance policy.

During the transition wage earners would have to contribute to their HFA from their wages, as discussed above. Medicare taxes would be continued for 21 years or possibly replaced by some other tax.

The loss of tax-deductibility for employer contributions to health care premiums, clearly makes the HFA system more financially progressive than the current system

The HFA may be spent only on medical care, and procedure payments paid directly from the HFA could be only at or below the 50th percentile in the region of the provider. Charges higher than the 50th percentile would have to be paid out of pocket or from a health insurance policy. After 21 years in the plan, at death, remaining funds in the HFA are to be transferred to the owner's estate. The proportion transferred to the estate would be prorated for years in plan less than 21.

Preventive care would be encouraged by an additional yearly use-or-lose Federal voucher for all individuals of $200.

For those who exhaust their HFA and all other resources, there is to be a Federal Safety Net, which will be comprised of a Federal Basic Minimum Policy (BMP) which pays only at the 10th percentile in the region of the provider. A BMP will cover only those procedures which are covered by 95% of the privately sold health policies in the US. The maximum lifetime expenditure of a BMP will be equal to the 5th percentile of the privately sold health policies in the US. The overall system discussed here is designed so that very few people should require this BMP safety net.

Since insurance policies are generally written in terms of what is not covered, a better way of characterizing what is covered in the BMP would be to not cover anything is explicitly not covered in 5% or more of health policies. For example, if psychotherapy is explicitly not covered in 20% of US health policies purchased in the market, it would not be covered in the BMP.

Some citizens may not want to, or be able to, keep track of premiums and pay them on a regular basis, and they may be incapable of keeping track of their HFA or of a debit card. Even these individuals, however, will be enabled to access care by, for example, simply walking into a clinic. There the staff will be able to find their Social Security Number, and hence their HFA and be able to show them the dignity of enabling them to pay for their care. Even at that point, the citizen can decide if the prospective treatment is worth the cost of the deduction from their HFA.

It is hoped that at least a substantial segment of the citizenry would use their HFA and out-of-pocket payments to pay directly for expenses below a high deductible. This should build a large free market for most procedures and lead to lower cost. This would be in preference to purchase of low-deductible insurance, resulting in never perceiving the cost of medical care.

One benefit of this free market is that it should draw doctors to the countryside and to inner cities to gain a market share of the individual HFAs of patients in those locations who will be enabled by their HFAs to, for the first time, be able to pay what the market requires to induce doctors in general, and general practitioners in particular, to move to the countryside or inner city. The increase in the market should also lead more individuals to choose to study medicine, further leading to a decline in prices, again, so long as the patients are conscious of cost.

Very high deductible insurance, for example at $50,000, should be encouraged to be widely marketed for the large number, in absolute terms, of individuals who have high balances in their HFAs. Policies with a high life-time deductible, such as $200,000, could also be usefully marketed.

Providers will compete for the high-cost procedure business of such wealthy individuals. This competition should determine competitive high-cost procedure prices which are currently determined in the relatively non-competitive market of insurance company buyers.

Another market-stimulating feature for high-cost procedures that could be implemented by insurance companies would be to issue rebates to patients who selected less expensive hospitals for elective or possible future emergency surgery, such as a coronary artery bypass This idea was first proposed by James Pendleton in "Market Driven Insurance and Health Savings Accounts", and could be applied in either high or low-deductible policies.

Such rebates can be substantial, reflecting large savings for the insurance company if a patient chooses a low-cost hospital. Of course these savings will likely also be reflected in lower premiums and should lead to rationalization of the currently observed large differences in hospital charges for identical procedures such as heart-bypass operations, even in the same metropolitan area. The rebates could be added to the patient's HFA, or taken as cash if the HFA were already at the upper $125,000 limit.

In addition, insurance companies should consider offering high-deductible policies which pay only to the 50th percentile. Such policies could be quite low-cost. Patients could choose to augment these payments from their HFAs or out of pocket.

Such policies would be analogous to indemnity insurance (a form of insurance which pays a fixed amount when a condition is diagnosed) and should also constrain costs for high-value procedures.

This HFA-based plan has substantial resistance to fraud. Although the HFA supplemental vouchers are funded by a single payer, the single payer does not control costs. There is very little rationing, all procedures currently accepted as deductible by the IRS, including long-term care, could be purchased directly using the HFA. Providers may charge any amount. Each person can have costs up to the 50th percentile in the neighborhood of the provider paid with their HFA, and the induced strong competition should reduce costs.

If patients pay with HFA, or cash, or their lifetime insurance policy they will not have to switch their primary care doctor or specialists as they often must do now when required to switch PPOs or HMOs because they change jobs.

Parents are required to purchase guaranteed renewable insurance for their children at birth, using, if necessary, their child's voucher. High deductible policies should be encouraged. Such a policy should, in general, be inexpensive, and will ensure that the child's HFA retains value past adolescence.

In contrast to high-deductible insurance, since comprehensive, low-deductible insurance undercuts the healthcare market, no government subsidy is suggested.

The program is designed to discourage purchase of comprehensive insurance. If comprehensive insurance is largely unregulated, and realistically priced, the initial tendency will likely be for the healthy not to purchase it. The resulting adverse selection will drive up the cost of comprehensive insurance leading in turn to fewer purchasers. This need not lead to more ill health for the unhealthy because the HFA should enable patients to obtain care up to the high-deductible limit of their high-deductible policy.

Persons may pay providers directly out of pocket to avoid spending down their HFA.

Most bills will be paid via a third-party, strictly financial transaction, probably commonly using a debit card, with a Medical Agent (MA) who holds the HFA account. At the point of purchase a preliminary estimate could be immediately provided as to whether the purchase would count as a medical expense. A record would be automatically kept of likely allowable expenditures below the yearly high-deductible limit and, so long as the deductible was not apparently exceeded there would be no need to definitively determine allowability.

When the MA is notified that a medicine or procedure had been purchased, it would send to the patient Federally-approved information on cost and indications for the medicine or procedure. This information for all procedures would also be available on the web.

The information sent would include the 10 and 50 percentile prices-paid information for the neighborhood of each provider of the procedures or medicines. This information should make it possible for patients to shop for a satisfactory price, a task which is made very difficult today, even for holders of HSAs, by the imposition of PPO network prices by the insurance companies who provide the high-deductible policy required to be purchased as an adjunct to the patients' HSA.

From a political point of view, the HFA vouchers could be regarded as a "Democratic" addition to level the “Republican” HFA/HSA playing field for the poor. The vouchers extend benefits to the young and to the poor; having an HFA for all helps ensure that a robust market mechanism exists.

Medicare, Medicaid and SCHIP would be replaced by the combination of the universal HFAs, high-deductible insurance, and the BMP safety net.

It is expected that in the steady state, after 21 years, that less than the current percent of GDP spent on health care would suffice for this plan because of simpler bookkeeping, because of cost reductions due to competition for the HFA holders' dollar, and because of reductions in demand for low-value expensive procedures due to individuals being more thrifty with their "own" money.

The virtues of this plan are that costs would be contained through competition, because, with the risk abatement inherent in the supplemented HFAs many persons would not choose comprehensive insurance and instead would choose only high-deductible insurance. Thus they would mostly obtain medical care under open Fee for Service so that a market could develop at the procedure level. Good health care would be available to the poorest, preventive care would be encouraged, particularly by the annual $200 use-or-lose voucher, and there would be few waiting lines.

To a first approximation, the administrative costs of the HFA system should be less than that of a single-payer system since for most procedures, for high-deductible policies, there would never be the need to determine if the procedure were medically necessary. This would be required only if the high-deductible limit were exceeded, so there should be considerable savings over most other systems. I expect that costs of keeping track of a patient's voucher, sending consumer information about treatments, estimating percentiles of procedure costs, etc., will be comparatively small and trade off against the diminished need to check for fraud and to set prices. (top)

RISK AVOIDANCE CONSIDERATIONS Some risk-averse individuals will continue to buy comprehensive insurance even though it is not the best value from a financial point of view, and even though they run very little risk of being without care without it.

On the other hand, most people will realize that they are almost fully protected from any health change by the combination of their HFA and their high-deductible insurance; and will feel that they are doing better than those with high-deductible insurance because in almost every year they will be out of pocket less than those who have comprehensive insurance. I expect that after the transition there will be far fewer persons with comprehensive polices than there are currently. (top)

Each person would have an account with the Federal Government, State Government, or Private Institution for their HFA. These paying institutions we will call “Medical Accounting” (MA) institutions. A large list of procedures and prescription and over-the-counter drugs would be authorized to be charged to these accounts. (Perhaps not facelifts, or extensive psychotherapy, but including dental care, glasses, hearing aids, preventive care and long-term care, and perhaps health clubs.)

The patient could use any licensed doctor or hospital in the US. Perhaps consideration should be made for licensed overseas providers.

If the patient wanted their bill to be covered by their HFA (they are free to pay the bill directly, out-of-pocket), the provider would send a copy of the patient's bill, perhaps simply using the patients HFA debit card, to the MA processing office which would pay from the voucher up to the 50th percentile for the same procedure charged by other providers in the provider's neighborhood. (Note that the provider's neighborhood is not necessarily the same as the patient's neighborhood.) The MA would charge the expense against the HFA (or "use or lose" voucher, see below).

Note that the 50th percentile is not the same as 50% of the typical payment, it is, of course, more likely to be almost equal to the typical payment.

Suppose that the price of an office visit in a neighborhood which contains 1000 doctors is as in the table below:

Number of Doctors	Price of Office Visit $
50	20-30
150	31-40
300	41-60
300	61-70
200	71-100

In this table, assuming that some doctor did charge exactly $60, the 50th percentile, aka the median, would be $60 since 500 doctors charged $60 or less and 500 doctors charged more than $60. The 10th percentile would be between $31 and $40.

To calculate the average price, we would have to know the actual price each doctor charged, but if the prices were uniform in each interval, calculation shows that the average would be ~$58, fairly close to the 50th percentile in this case.

Simply stated, the 50th percentile is a dollar figure at or below which ~50% of providers are charging; similarly, the 10th percentile is a dollar figure at or below what ~10% of providers are charging. The difference between the 50th and 10th percentiles may be quite large, or may be negligible, depending on the distribution of prices among the providers.

To obtain the data required to determine the 50th percentile, once a year each provider, would be required to report to an MA the median price the provider received for each procedure they actually delivered. Perhaps the list of procedures for which data was provided could be restricted to those 50 most frequently provided by the provider.

Such "payment received" data is currently generally unavailable because HMO and PPO organizations regard the data as proprietary. It seems plausible that government intervention will be required to produce such consumer-useful statistical data. Similarly, of course, drug stores would have to report median prices received for drugs (perhaps only prescription drugs), opticians for glasses, etc.

As a second approach, instead of being required to provide the median data directly to the government, the government could simply require that each provider post the price they will accept for any procedure they offer. It is thought that posting of such prices would rapidly lead to a narrowing of the price range.

Then the government, and possibly private consumer organizations could gather these data and make them usefully available to consumers.

For example, imagine a web page on which the patient enters their zip code, the procedure desired, a price threshold, and a distance threshold and a map would result showing all providers meeting the criteria.

Note that a patient might choose to travel to a less expensive part of the country, or find an especially low-overhead, high-volume provider for an expensive procedure, in order to find a provider who would require a lower deduction from their HFA. Note also that it is in the patient's interest to use providers who charge below the 50th percentile because that will result in a smaller deduction from their HFA. There is no strong incentive to seek a provider who will charge exactly the maximum (50th percentile) payment available from the HFA.

Instead of having their HFA reimburse the provider, the patient could pay the bill directly and send the receipt to the MA office, which would reimburse the patient from their HFA account.

Or, a third alternative, it would be possible to have the patient simply withdraw money from the MA, perhaps via "HFA checks" as they needed to pay cash to the provider. Then, accumulating receipts, the patient would be prepared, if necessary, to justify to the Internal Revenue Service that all withdrawals were spent on medically necessary items.

However, paying with withdrawn cash, may be an invitation for individuals who are not wealthy to use the medical funds for food, clothing, entertainment, etc. The IRS is unlikely to scrutinize returns for items below the deductible, i.e. those that will not increase tax revenues. In addition, even well-off people would rather use health funds for non-health items when they are well; for example, consider how many employees use sick leave allotments for regular days off.

Therefore, it seems desirable that most medical expenses paid with the HFA (not out of pocket payments) pass through an MA for payment. With a debit card type system, this should be minimally restrictive.

Perhaps a suitable compromise would be for the IRS to not require documentation of expenses which passed through an MA; but in an IRS audit, documentation would be required for expenses which were funded by cash withdrawn from the HFA. This policy should encourage people to use their debit cards, and reports from the MA of high levels of cash withdrawals should provide a flag indicating fraud to the IRS.

From a positive point of view, the option to use cash withdrawn from the HFA for health purchases will enable patients to obtain products which are allowable by the IRS but may be denied by the MA because, due to some bureaucratic bungle, it was not on the list of approved medical products. The cash option would also make it possible to obtain privacy, to the extent that documentation would only go to the IRS, and even then, only in case of an audit.

The HFAs could be regulated by the states. However, of course, the lifetime, guaranteed-renewable insurance policies themselves must be portable from state to state.

Once a month the MAs would send statements to the patient showing the costs incurred and the amounts left in their yearly voucher and HFA accounts.

In addition to competing on price, a provider might compete on convenience, perhaps offering weekend and evening hours or perhaps videophone or email consultations, possibly for extra expense; or possibly not, in order to gain market share.

In addition to competing on the basis of cost and convenience, as discussed above, providers could also compete on the basis of quality. For example they could cite board certifications, number of procedures provided per year, and quality of outcomes. For superior quality they could ask a higher price.

This sort of competition would represent an improvement over the status quo where competition at the procedure level is mostly on cost as negotiated by the insurance company, and in which providers are denied access to patients unless they promise insurers a low cost.

In the status quo situation, providers can only make profits by increasing the numbers of procedures executed, and minimizing, below the contracted cost, the actual cost per procedure. This minimization is likely to lower quality; and there is no immediate penalty for such lowering of quality; although in the long term, of course, the patient's health may suffer.

In our suggested plan providers can charge each patient any amount. For example they might charge poor patients less and rich patients more.

Of course, it would presumably be illegal to, for example, charge different racial groups different prices.

The "neighborhood" (the provider's neighborhood) over which the percentile is calculated will vary according to the procedure. For example, for general practice office visits the neighborhood would be local to the provider, since such trips must be made frequently.

For difficult operations, the metropolitan area of the provider could be suitable, or even the entire nation. For long-term care, a state or multistate region might be suitable.

A result of this might be that those few poor city patients in the Basic Minimum Policy (BMP) safety net which pays only to the 10th percentile, (see below) would tend use the less expensive long-term care facilities in the countryside.

One can imagine patients scanning the data for a metropolitan area for the general practitioner with the lowest weekend 50th percentile cost for a physical examination and traveling well outside their neighborhood for that purpose.

Having the MA pay from the HFA only up to the 50th percentile reduces the possibility that some patients would rapidly spend their voucher mostly for the prestige of being attended to by famous doctors or by expensive "society" doctors. In addition, this mitigates the possibility that some doctors might purposely raise their rates to attract such poor customers and thus, indirectly, raise the overall costs of healthcare.

Since at the initiation of the voucher system there would be no persons in the BMP safety net, a fairly free market in procedures should spring up and a full distribution of costs for a fixed procedure should develop. This will provide the statistical foundation for estimating the 50th and 10th percentile points for individual procedures.

It is sometimes suggested that doctors initially charging below the 50th percentile will raise their rates to the 50th percentile since the HFA will pay only up to that amount. This seems unlikely to occur since some patients will want to pay less than the 50th percentile to conserve their HFA and to avoid dropping into the BMP safety net, and will seek out doctors who do charge less. Furthermore, I would expect that groups of doctors would form new organizations to deliver care at prices lower than the 50th percentile in order to gain market share and make greater profits through greater volume and reduced costs. This also would ensure that costs would not stick at the 50th percentile.

Some anti-monopoly regulation by the government might be necessary to ensure this. Economists should study the expected market development using techniques similar to any which would have predicted the course of prices and changes in business practices under telephone and airline deregulation. (top)

When people use up their annual use-or-lose voucher, HFA, most assets, and are without a health insurance policy, or have exhausted their health care policy; they would receive a Basic Minimum Policy (BMP) as a safety-net system. (They would, of course, continue to receive the annual $200 use-or-lose preventive care voucher.) The basic idea is that the BMP would be awarded for those who are so poor that they would have qualified for welfare and/or Medicaid in the past.

The BMP will have characteristics much like the policies that all other citizens have purchased, but will not be so desirable that citizens will squander their HFAs feeling that they will nonetheless have as good care when they have a BMP. If a citizen has had to cease paying premiums on an existing policy because of poverty, and begins drawing on a BMP, then the government should guarantee that they be re-offered their original policy on the original terms when they emerge from poverty.

The BMPs' features are to be determined by a survey of what is purchased on the open market and not by a political process. Thus their features will not be subject to lobbying by special interests.

The procedures covered by a BMP would be restricted to those offered in 95% of all private health policies in the US. In this sense the procedures offered would be a consensus of those procedures determined by the market to have sufficient value for money. The formulary would probably include only generic drugs, with, perhaps a very few exceptions for those cases for which there is no plausible generic counterpart.

The allowed BMP payment would be at the 10th percentile for the region of the provider. (This is in contrast to payment at the 50th percentile which is the threshold for payments from an HFA.) In addition, although only payments at the 10th percentile would be available from the BMP; that level of payment would guarantee that providers who will provide the service at the 10th percentile do exist near the patient. The patient could supplement this payment with other money, if possible, from charity, friends or family, to obtain the services of a more expensive provider.

Since only a few individuals should be in the BMP safety net, there should be sufficient providers to meet the need. At the 10th percentile, by definition, 1/10 of all doctors in the neighborhood provide services at that price.

The annual deductible would have to be set at a low value, since this is a safety net policy. I suggest a deductible of $1000 accompanied by a Federal supplemental deposit of $1000 into the patient's HFA. (Presumably the HFA is near-empty, otherwise there would be no need for a medical safety net.) This should encourage the patient to shop carefully even after the award of the BMP. The patient could use this HFA money to pay for visits to providers that charge above the 10th percentile.

The lifetime maximum benefit would be the value at the 5th percentile of all private policy lifetime maxima in the US. In this sense this maximum benefit would represent a minimum acceptable lifetime maximum benefit.

Should the lifetime maximum be exceeded for someone with a BMP, additional expenses would have to be provided "free". That is, they would be cost-shifted by the providers, as is often the case at present. The total amount of such cost shifting should, however, be far less than at present because far fewer individuals will find themselves in this situation.

A person would try to avoid falling into the safety net because they would likely have to change from their usual doctor (unless that doctor offered to reduce their fees for, e.g. needy long-term patients). It would also be undesirable because patients would likely have to travel longer distances because, of course, only 10% of the doctors perform procedures at the 10th percentile of cost.

The fact that the consensus list of procedures for which the 10th percentile payment is authorized under the BMP will be substantially shorter than that which could be purchased with HFAs or cash, ensures that many of the cost control features of the system will be retained and there will be incentives for persons to buy insurance to avoid falling into the safety net in the event that they contract a serious illness.

It is likely that this list of safety net procedures will contain many fewer procedures than is available today under Medicaid.

Although it is true that the reduced benefits available in the BMP are crucial to the success of the overall plan; and although it is true that this reduction is vulnerable to political attack; every health plan is vulnerable in this way wherever in the plan the reduction in benefits aka rationing is, in fact, embodied.

There is always political pressure to increase benefits. In the HFA system, the weak points are clearly isolated in the value of the yearly HFA voucher increments awarded from birth to age 21, and in the procedures available in the BMP safety net.

The BMP procedure restrictions are, unfortunately, easy to attack because they are so visible. However, because very few people will fall into the safety net, and, because of the clear rationale for the restriction and method of selection of procedures, they may also be relatively easy to defend. It will be worth pointing out that 5% or more of policies sold on the open market also do not offer some procedure not available in the BMP.

Licensing of providers should ensure that the 10th percentile providers meet minimum standards and thus have low costs principally because of such reasons as low overhead, efficiencies, and large volume.

It is to be expected that many moderate quality low cost providers would emerge in order to take advantage of high volume, much as Walmart replaced many moderate quality small retailers. There is no reason to think that the 10th percentile providers would necessarily be degrading to patronize; licensing should maintain overall quality. The offices may be spare and simply furnished and in low-rent areas. The doctors may have set up procedures whereby less completely trained personnel perform preliminary screening; or they may make extensive use of computer diagnosis, etc. (top)

The MA should also include with the report of the bill to the patient the median local, state, and national charges for the procedures used by the patient that month, broken out as to whether the provider had various quality characteristics such as board certification. This information should help the patient decide if their doctor's charges are reasonable. To further assist in this decision there should also be included indications, details, and perhaps some outcomes results for the procedure.

This consumer information should also be available generally on the web, in libraries, etc. to enable consumers to select doctors, hospitals, and procedures in advance of any illness.

Because patients would more often be spending their own HFA and out-of-pocket money under this plan they will question their doctors more closely about the relative efficacy of inexpensive and expensive procedures; for example comparing physical therapy to an initial MRI as treatment for back pain. When the doctors confess, as they must in many cases, that it is not clear which is best, the patient will likely choose the less expensive procedure, and may in many cases, exert pressure for further studies of comparative effectiveness. It is generally agreed that such studies are relatively underfunded.

The consumer information sent to the paying patient could also be generally available to consumers upon request so that they could evaluate the doctor's approach before treatment. It might also be required that insurance plans provide their patients with this information after every procedure so that they could evaluate at the end of each year whether they had obtained their money's worth from the insurance.

The collection of MAs need not be a large bureaucracy; it should be similar in scale to the Medicare or Canadian bureaucracies which are often said to be small and efficient. In fact, it should be smaller than those organizations, since it has no mandate to control costs, and almost no mandate to restrict funding of procedures. (top)

A political virtue of this funding approach is that if Congress wanted to steadily increase or decrease the share of government tax resources devoted to medical care, they could do so by increasing or decreasing the value of the yearly voucher for those under 21, without facing the politically more difficult task of choosing between procedures or beneficiaries.

For those over age 21 it should also be possible for Congress to target aid to those in specific need, for example the unemployed, by simply adding funds to their HFAs. Or, it may become apparent that there is a need for all citizens to receive an additional HFA voucher of some size each year.

The expectation is that a person would look for bargains and would not squander their HFA because they would want to conserve it for some later need, and because, at death, after age 21 the HFA would pass to their estate. Early (2008) experience with MSA and HFA employer group insurance indicates that the desire to conserve the voucher also encourages preventive care and good health habits. The search for bargains would further encourage competition and thus control costs.

It is often argued that patients cannot look for bargains when they are extremely sick, when most of the money is spent, and thus regular market forces do not apply to medical need. However, the patient may survey advertised prices of doctors and hospitals, weigh those against quality and tell their relatives and friends in advance where they should be sent in the event of an emergency, if possible.

In addition, any insurance bought on the open market may be less expensive if the insurer can specify cost-efficient doctors and hospitals to be used if possible.

Insurance policies could also include rebates to patients for selecting less expensive hospitals for elective or emergency surgery, such as a coronary artery bypass, as discussed by James Pendleton. Such rebates can be substantial, reflecting very large savings for the insurance company. Of course these savings can also be reflected in lower premiums and will doubtless lead to reduction of existing large differences in hospital charges. Depending on law, the rebates could be taken as cash or be added to the patient's HFA. (top)

Also, each person would be given a "use or lose" voucher allotment, say $200, each year (stagger using birthdays to avoid an end-of-the year rush.) This also would encourage preventive medicine. This would ideally be spent on dental and eye checkups, and on such other procedures as blood pressure measurements and pap smears.

Many useful medications are non-prescription, such as aspirin and Prilosec, and these too should be covered by the use-or-lose voucher (as well as by HFAs). It would also be desirable if the IRS could modify its rules so that exercise club fees were covered; it's likely that these would have to be capped.

Currently, many preventive procedures are covered free by high-deductible policies which accompany HSAs. This "routine" care would ideally instead be covered out-of-pocket so that prices are controlled by the market. If these preventive procedures are covered by insurance, providers are likely to keep their costs high. The "use or lose" voucher strongly encourages preventive care, but keeps it in the market. Providers will include cost among their inducements to potential patients. top)

The HFA system should change the current motivational structure in which doctors resist para-professional personnel for fear that they will displace the doctors. Instead, some doctors will likely head organizations trying to lower costs in order to gain market share. It has been suggest that it may be necessary to modify anti-trust laws to make it possible to create such organizations.

Since doctors may charge any amount, the best doctors income may rise. Due to competition, the salary of less able doctors will decline.

It may be that in rural or inner city areas where there are only a few providers the 10th percentile may be the same as the 90th percentile. In this case, the only restraint on cost would be the apprehension of doctors that, if they raise their prices, doctors will move to their neighborhood or that patients will travel away for care.

Looking at this from another point of view, there should be sufficient HFA wealth available in rural or inner-urban areas so that prices can rise to a level such that primary and other doctors will be interested in practicing in rural areas. Thus, various artificial government mandates for selected recent medical school graduates to practice in rural or inner-city areas should no longer be necessary. Also, the increase in average primary care doctor's salary, due to demand, should result in a substantial increase in the number of new doctors specializing in primary care.

Existing primary care doctors may choose to offer a lower price for patients who agree to an initial screening by physician assistants or registered nurses. Or a new doctor may offer low prices to build his or her practice. Such system changes in response to a new ability to pay should quickly increase the overall level of primary care available.

Under the HFA approach, there would be no more persons in Medicaid. Thus this transition is as "instantaneous" as the bureaucracy can make it.

However, as HFAs are exhausted for some individuals, and perhaps as they allow their insurance policy to lapse due to carelessness or to inability to pay their premiums, the equivalent of Medicaid will re-appear as the BMP safety net program; however, as discussed above, the BMP safety net would be less generous in terms of procedures covered, than Medicaid is today.

However, in the old tradition of medicine, doctors could donate services for uncovered procedures or for uncovered costs above the 10th percentile to those in the safety net. This is a tradition not often seen today where the government is expected to pay all bills if private insurance does not. (top)

Private comprehensive (aka low-deductible, low-coinsurance, low-copay) insurance could be purchased with HFA funds. (Under current law HFA funds cannot be used to purchase insurance.)

It is clear that many persons do prefer comprehensive insurance, as evidenced by purchase of Medigap insurance to supplement Medicare, even when the patients are aware that a large portion of their premiums will be used for administrative expenses.

However, this devotion to comprehensive insurance will likely be reduced in the HFA system. For example, for those currently covered by Medicare, since the HFA should be sufficient to meet expenses not covered by high-deductible insurance, the patient need not worry about coverage for the deductible, as he has currently to worry about not being covered by the gaps in Medicare. Thus, one would expect a lesser expenditure by individuals on comprehensive insurance, and greater expenditures on high-deductible insurance.

Private insurance could offer comprehensive policies which would pay to a level above the 50th percentile. However, so long as some patients are searching for low prices, comprehensive policies should not make it impossible to find low prices.

The administrative costs of private high-deductible insurance should not be as great as for medigap insurance since the high-deductible insurance payment system is seldom used. The administrative costs should be more similar to auto liability or homeowners insurance. As discussed previously, an expense account for most medical expenses can be automatically maintained by consistent use of a special HFA debit card. The account need never be consulted unless the deductible is exceeded. (top)

The costs of private insurance purchased with the HFA would not be controlled. The regulator would only want to ensure that the insurance company could meet its obligations with the scheduled premiums.

Some insurance is required to be purchased in anticipation of the birth of a child, and must be maintained to age 21.

The parent might pay for this insurance in order to conserve the child's HFA. The child's voucher itself could be used to purchase insurance and would become available for insurance three months before the expected birth.

Obviously insurance companies would give lower rates for insurance to children whose mother had prenatal care and who did not smoke or drink. For guaranteed-renewable, high-deductible insurance (see below) the rate schedule would, in effect, be set for the child's entire life.

While it might be expected that parents who had poor health habits would pass these traits along to their children, and hence that premiums might be higher for these children, there are other possibilities. Since premiums will almost certainly be paid while the child is young and is very seldom sick; the insurance company will be ahead while the individual is young.

Individuals with poor health habits may die while quite young and after a short illness or as a result of an accident. So it is not at all clear that insurance companies would charge such individuals high rates at birth.

Genetic illnesses which were not predictable would obviously be covered as classic insurance. Genetic defects which could be predicted or detected might raise insurance rates and use more of a child's voucher. It would be up to the parents to decide whether or not to have a child which it was known would be born with a genetic disease; they would have to keep in mind that the child might eventually fall into the safety net and that the parents might feel impelled to spend their own resources on their children's medical care.

The fact that the voucher/HFA would become available three months before birth will be controversial. However, allowing this is desirable from a health policy point of view because it maximizes the size of the high-deductible policy pool by making underwriting possible before some pre-existing conditions are apparent. Also, the earlier underwriting begins, the earlier the cost implications will become apparent to parents and hence the earlier they will adopt healthy life styles.

It might be objected that insurance companies may require higher premiums for the children of poor parents, anticipating that their medical expenses may be higher. But it is important to remember that the guaranteed-renewable policies will likely be valid for life. The children of wealthy parents will still very possibly have expensive illnesses late in life; and may also be more vigorous in seeking out expensive treatments. Also, the children of poor parents may die relatively early of non-chronic, relatively inexpensive, causes. Hence it is not clear to me that it would be worthwhile for insurance companies to underwrite on the basis of parent wealth.

The HFA may be used to pay the premium for only one year at a time, even for insurance policies which are guaranteed renewable. This restriction is in order to ensure that the government is not eventually required to bail out health insurance companies who have offered, for large up-front voucher payments, expensive long-term plans on which they can not deliver. (top)

The government should encourage the private insurance industry to offer incentive-compatible, guaranteed renewable, high-deductible insurance policies. Such policies have the characteristic that the policies cannot be cancelled, and the premiums cannot be raised in response to the the health experience of the individual.

Insurance companies would be allowed to reduce rates based on good health experience in order to prevent other companies from skimming off those customers. These premiums could be subsequently returned up to, but not above, the initial values based on the customer's subsequent poor health experience.

Herring and Pauly (2004) have shown that such guaranteed-renewable health policies are viable, (Incentive-Compatible Guaranteed Renewable Health Insurance, NBER Working Paper W9888, ). Similar results have been developed by John H. Cochrane of the University of Chicago. Regulators should require that such policies be made available to consumers.

Herring and Pauly (2006) Incentive-compatible, guaranteed renewable health insurance, J Health Econ, 25, 3, 395-417 have also demonstrated the substantial prevalence in today's free market of such renewable health policies in states that have neither guaranteed issue nor community rating. The policies have the characteristic that premiums are higher at the start than needed to pay average benefits, but higher than required toward the end of the policies.

The policies were studied only in the age range 18-64 and were found to be viable. Further studies are needed to determine viability in the age range from birth to death. Conceivably government subsidies will be needed late in life; although I hope that the many years of good health from age 0-21 when the government $2000/year HFA vouchers are available to pay premiums can set up funds for old age.

These policies should also be regulated so that, should the insured fail to pay premiums because of lack of resources, and hence fall into the BMP safety net; that he or she should be allowed to resume the original policy on the original terms when they must leave the BMP because of renewed financial health. Probably the same policy should obtain for those who have been unable to pay premiums on low-deductible plans. Of course care must be taken to ensure that this renewal policy does not allow individuals to game the system, paying premiums only when they anticipate that they may become sick.

Obviously, if a financially healthy individual simply fails to pay premiums, then insurance companies may impose underwriting tests before covering them again. A sign that such an individual was financially healthy would have been that they were not presently covered by a BMP. To have qualified for a BMP one would have had to have demonstrated that one could not afford premiums for a private policy.

A problem with lifetime policies is how to handle new, expensive beneficial procedures. I suggest that, as is normal, the insurance policies be initially written to support a particular basket of procedures.

Should a new procedure be very expensive so that the price of the new basket of procedures which includes it is excessive, the insured party could be given a one-time option of increasing their premium to cover that procedure without regard as to their health status. (The option cannot remain indefinitely open, since then the insured party would not choose the option until their risk of needing the procedure increased. There could however be a subsequent one-time option to drop a procedure and hence reduce options.)

In deciding after the initial offering whether to insure or not, and what rates to charge, for the new expensive, beneficial procedures the company could impose most life-style and pre-existing conditions restrictions. (Note that premiums for pre-existing conditions are subsidized by the government at program start-up at the beginning of the transition, see HFA INITIATION below.)

Another approach to handle new, expensive beneficial procedures would be simply to determine a premium path which allows for some increase in medical costs above inflation. Those who wanted to minimize the risk of not being able to afford future desirable procedures could select to pay for higher rates of increase of medical costs.

Note that the fact that the HFA vouchers are the same for all persons, and that insurance companies may use lifestyle in setting new-customer premiums and in deciding whether or not to reduce or to subsequently raise premiums to the original levels for existing customers, removes the necessity of having governments monitor the lifestyles of citizens in order to prevent tax dollars from being "wasted" on those with "sinful" or unhealthy behavior. The only reason that the government might engage in such monitoring would be to reduce the likelihood of individuals dropping into the BMP safety net. It is planned that the number of such individuals will be small, and hence that the government need not monitor the behavior of the typical citizen.

Insurance companies could increase rates for most life-style changes, and for normal changes in life. For example if a person began to smoke or to fly a private plane, rates could be raised. Of course rates do increase with age on guaranteed-renewable policies.

Companies could offer high-deductible coverage whose rates would increase more slowly so long as participants participated in a wellness program. Depending on politics a few life-style changes, for example entering a risky public service job, might or might not be disallowed in law as justification for setting rates.

Premiums might be reduced if patients undergo various screening examinations. These might not have proved profitable for insurance companies in the past, since patients often had moved on to another company by the time treatment of a discovered illness had paid off in premiums from a longer-lived customer. However, in the present situation, where people may keep a single policy for much of their life, screening examinations may pay off for the insurance company: the patient may pay premiums for a longer period of healthful life, instead of dying after only one medical incident which was not caught early enough by screening.

Due to Federal supplementation at startup, all should succeed in obtaining a high-deductible policy at startup. But, should they fail to keep it in force; and should they subsequently be unable to find high-deductible insurance below a threshold price, determined by the insurance industry, then they may fall into a regulated high-risk insurance pool, similar to the high-risk pools presently operating in several states. Should they be unable to afford these high-risk premiums, then they would fall into the BMP safety net.

The high-deductible yearly out-of-pocket limit should not be allowed to grow too large, perhaps not over 1/4 of the funds in the HFA; a maximum of $125,000/4=~$30,000. Should the limit become larger, then those persons who develop a chronic illness might exhaust their funds in a few consecutive years of paying the limit. In this case they may have to use the safety net, thus drawing on public tax funding. As noted elsewhere, it is important that the number of citizens in the safety net be minimized.

Of course it is desirable that the high-deductible limit be this high for at least some individuals in order that even expensive procedures feel the market pressure of customers who are spending their own money.

To protect against falling into the safety net due to chronic conditions, high-deductible insurance companies might offer policies that completely cover a chronic condition after several successive years in which the patient had exceeded a high deductible limit to which that chronic condition had significantly contributed.

It might be desirable to encourage insurance companies to offer plans which only pay up to the 50th percentile, or to offer other forms of indemnity insurance, so as to minimize upward pressure of insurance on expensive procedures.

From the market point of view, it is desirable that the maximum number of persons buy health care one procedure at a time, that is, without comprehensive insurance, so that a true market at the procedure level is created. Thus, the government should not encourage or subsidize comprehensive insurance.

Note that this approach to universal coverage differs from many reform proposals which restrict market competition to only that between companies selling comprehensive insurance policies.

States could retain the right to make the laws concerning insurance. The states would have to regulate to ensure that the policies offered, in particular the initial premiums, were actuarially sound if renewed so that the government did not have to pick up the tab for companies which over-promised. And of course, states must allow individuals who move to another state to take their lifetime guaranteed renewable policy with them.

It of course would be desirable if companies market their policies across state lines. top)

HMOs could continue to exist, but like other forms of insurance and fee for service, their premiums would no longer be tax exempt. Therefore, HMO's and similar organizations may lose market share.

On the other hand, there would be nothing to prevent an association of doctors from advertising that they will give discounts to patients who are referred within the group or who spend a prescribed amount within the group. This sort of creative marketing is impossible in our present system.

Such a group of doctors could also ensure that other doctors in the group whose patients regularly had bad outcomes did not continue to practice within the group.

These arrangements could result in many of the same favorable outcomes for patients as in historical HMOs. Since HMOs were initiated voluntarily before cost became the overriding consideration it is today, it is plausible that HMOs will continue to be offered. However, it is likely that, when everyone has their own HFA, HMOs will not be the dominant mode of health care delivery as was once predicted. The concepts of government specified capitation and risk adjustment would become largely irrelevant. (top)

Since rich people will more easily be able to pay above the 50th percentile and are less likely to run out of HFA and out-of-pocket money, they and their children will still usually have better medical care if they are seriously ill. But this HFA program will ensure that most prudent persons will have nearly equal access to excellent health care for most of their lives. Although rich people will likely always have an advantage over the poor; at least in this system, the vouchers from birth to age 21 ensures that there is a greatly reduced health care advantage resulting from being rich early in life as compared to the present system. (top)

The best doctor's incomes would probably increase since they would be free to raise their rates over those currently paid by insurance companies, Medicare, and Medicaid. Patients would presumably seek these doctors out because of their reputation. The income of the average doctor may decline due to competition. (top)

A patient could always pay the doctor out-of-pocket and not draw down their HFA if they wanted to, perhaps for privacy or to save the voucher for their old age.

Since any patient could choose to pay directly, and many would, this approach to medical care, as contrasted to a plan in which care was obtainable only through payment by the government, as in Canada, could not significantly aid in detecting illegal immigrants. On the other hand, illegal immigrants would have no access to HFA supplements or to vouchers and would, presumably, have free care only in an emergency as at present.

Resident aliens would initially receive vouchers equal to 1/4 that of a citizen of the same age and be required to build their HFA, if employed, as are citizens. After 10 years residence the voucher would increase to that of a citizen. (top)

The universal HFA supplements have the advantage over refundable tax credits for low wage earners in that they in no way comprise a disincentive to earn more money. They also do not increase the numbers of individuals who pay no taxes and hence feel little personal investment in fiscally sound government.

Wage earners will be required to place about 5% of wages (but not of other income) resulting in up to $5000 per year in a HFA account. That is, the HFA is mandatory. However, the degree of government coercion seems to me to be less than in the mandatory Medicare payments together with the mandatory taxes needed to support Medicaid. Taxes of this latter sort could be substantially replaced under the HFA system since they are no longer explicitly linked to the major government assistance, vouchers from birth to age 21.

The contributions from wages would be mandatory until the total contributions plus return on investment in the account reached $125,000 (inflation adjusted, of course). Once this amount is achieved, the account would no longer need to be supplemented from income, although, if desired by the patient, it must be later supplemented if negative investment returns or withdrawals result in the account dropping below $125,000. Should investment returns result in the HFA account being valued over $125,000, the excess is distributed to the owner.

An HFA account can be kept directly with the government or in a private institution. Money can be spent only on health-related expenses. An exception might be made for items such as rent after age 75. However, such possibilities should be closely examined since, should they exhaust the HFA, they could result in the general tax-supported government BMP safety net being required to provide for medical or long-term care.

A principal purpose of the HFA, and a reason for making HFAs mandatory for wage earners, is to minimize the number of persons who fall into the safety net, thus minimizing the safety net's additional burden on the taxpayers. Before safety net benefits are granted, as is currently the case for Medicaid, the patient will have to certify that their HFA account and other personal resources are exhausted.

For this reason, HFA funds may not be transferred to another person, except that it enters the owner's estate at death.

Also, the HFA is mandatory because it is desired that the voucher and HFA, together with incentive-compatible, guaranteed-renewable, high-deductible private insurance, as much as possible replace comprehensive insurance, whether private, Medicare or Medicaid with a market-based system at the procedure level.

If payments from wages to the individual's HFA is mandatory, as recommended here, employers will be naturally inclined to award any health benefit to their employees in the form of direct contributions to the HFA. If the HFA were voluntary, then many employers might instead continue to offer comprehensive insurance because that is what most employees and employers are accustomed to and because, at present, it is the paradigm for high quality employer-provided health care.

Of course employers may certainly continue to offer comprehensive insurance even if the HFA is mandatory. However, the employee and employer would then have to coordinate the voucher, HFA, and the comprehensive insurance. The complications, especially including those involved in changing jobs, would lead most employees to favor having the employer contribute directly to the employee’s HFA. Since most comprehensive insurance programs also cover dependents, one would expect that if the employer shifted to funding HFAs that they would also fund dependents' HFAs.

Employers may choose not to explicitly make a contribution to the HFA, instead asking their employees to choose how much of their salary to contribute to their HFA. In either case, however, employers will still have to pay enough salary to entice employees to work for them.

It is important that comprehensive insurance not dominate employees’ health plans. If it did, then the health market would be suppressed at the procedure level, because third party payments would continue to distort the market.

If many employers, especially employers of the middle and upper-middle quintiles, offered comprehensive insurance; then that paradigm would be seen to be standard and desirable. So advocates for the poor would continue to argue that the poor also should have comprehensive insurance. If, on the other hand, the middle-class receives HFA-support from their employers; then the Federal government will be urged to give the same benefit to the poor. Perhaps the Federal government would then go beyond the income transfers so far discussed and directly place money in the HFAs of the poor over age 21.

Also, the middle-class are the individuals who have the time to seek out the lowest prices and thus make the market. They also will make the best use of the information returned from the HFA MAs to those who use the HFA to pay bills, thus enhancing market efficiency.

Because it is desirable from a system point of view that persons not fall into the safety net, it should be required that the HFA be invested conservatively. Perhaps only in CDs, or a well-balanced account of CDs and bonds.

It would not be mandatory that the self-employed over age 21 contribute to HFAs from their non-wage income. This is in order that they may use their limited funds, possibly for a new business, as they see fit. Most of these individuals are unlikely to end up in the safety net, and, in addition, there are relatively few of them compared to wage earners so the risk that large numbers of individuals end up in the safety net is small.

The fact that all wage-earners would be required to contribute to an HFA will reduce the "free-rider" problem even without requiring that health insurance be purchased.

Other individuals who do not draw wages, e.g. non-working spouses, would be permitted to contribute up to $5000/year to an HFA account. They would not be required to do so. Funds to enable these contributions might constitute the "family health benefit" awarded by employers to employees. (top)

The 50th and 10th percentiles need a small "buffer" so that unreasonable decisions are not made. For example, if the 50th percentile were interpreted as "50th percentile plus $5.00" then small purchases of aspirin would not generate complicated paperwork because the purchase was made at a convenience store where prices were a bit high.

Similarly, why require a patient to pay $4.75 out-of-pocket when the hospital was striving to meet the 50th percentile on a $50,000 procedure, but failed due to unanticipated causes. (top)

To fund the vouchers I would favor canceling all Medicare and other taxes now going to medical care, and substituting a progressive Federal consumption tax in the form of a tax on income not invested. (the "USA" or "Unlimited Savings Allowance" tax, supported in 1996 by Nunn and Domenici) See also Maya MacGuineas article "Radical Tax Reform" in the January-February 2004 issue of Atlantic. However, this topic is clearly out of the scope of health care policy. (top)

Possibly, there could be squandering of voucher enhanced HFAs if patients believed that the government would later improve the BMP safety net. Hence it is important that the BMP not be too generous and that not too many people fall into the safety net.

Such patients might decide to get as much care, or engage in fraud in collusion with doctors in order to skim off payments; such behavior is currently observed in Medicaid, Medicare, and private insurance.

However, such behavior would seem to be less likely in the HFA plan as compared to Medicare, Medicaid, SCHIP, and comprehensive private insurance. In the HFA system patients are much more likely to be spending their own money, and hence much more unwilling to share any of it with criminals.

Children are born with a right to a $2000 yearly voucher supplement through age 21. This is a major advantage to a voucher approach: children get off go a good start in life with, e.g. well baby clinics, and good dental, aural, and vision care. Note that $2000 is much less than the $7000 spent yearly per capita in the US on health care in 2008, and also much less than the ~$8,000/yearly spent per capita for a child in the US on education.

Some employers may respond to the government's funding of childrens' HFAs by ceasing to provide extra salary, either directly or through dependent health care insurance, or for depositing in dependent's HFAs for those employees with children. However, economists often note that health care coverage by employers should be viewed simply as a higher salary. If this is true, then the loss of employer health coverage does not really affect the employee negatively. The extra money is available to be placed in family HFAs.

By the same token, if this system is in place, employers will be less inclined to be prejudiced against hiring individuals with children in hopes of saving the company money.

The childrens' parents would naturally administer the HFA; say until the children are 18. Parents would choose the health insurance policy which is required until age 21. Since children will not necessarily know when their HFA is being depleted they might be defrauded by their parents, so the investment and dispersal of their accounts should be closely watched by the MA, perhaps by requiring a second opinion for long-term drug prescriptions or any expensive procedure for children. However, it should be mentioned that, from the taxpayer's point of view, there is less likely to be a societal fraud problem under the HFA system than under Medicaid or private insurance coverage of the child.

Parents, of course, may choose to add additional funds to their child's HFA out of pocket. Should the child's HFA reach the cap, a future employer would have to choose how to make up for the fact that this employee will not be rewarded by a benefit which contributes to an HFA. The employee could simply be offered a higher salary, for example. This could be standard for employees who reach the HFA cap. These are all options for the parents and companies; the government should not regulate these options.

Parents would also be required to take care that any taxes on the child's HFA vouchers and HFA accounts are properly handled. In general, even though the vouchers are taxable, the total income of almost all children would remain below the level at which taxes would need to be paid. Tax filing then could consist only of certifying that this was the case and could be handled as an extra statement accompanying a parent's tax form.

Perhaps it would be satisfactory for parents to only buy only the lower-priced "health status insurance" elucidated by Prof. John Cochrane of the University of Chicago. The actual small expenses which most children need could be paid out-of-pocket or directly from the $200 use-or-lose voucher or from the childrens' HFAs. Should a high expense arise, the health status policy could be invoked and there would be insurance for following years. However, this would run the risk of ruinous expense in a single year; probably should not be allowed.

Because of the need to ease the immediate transition out of Medicaid and Medicare, there would have to be a multi-year transition.

In addition, those citizens who become 65 during the first transition year will also receive a Medicare transition HFA voucher sufficient to purchase a high-deductible incentive-compatible, guaranteed renewable policy with benefits actuarially equivalent to the median policy available on the private market. The voucher's value would be adjusted for pre-existing conditions. This voucher would be awarded each year for life.

Presumably, most of the recipients of these Medicare-transition vouchers will use them to purchase a suitable insurance policy, but they need not do so.

During the following 21 years, each wave of citizens who become 65 will receive the same voucher for life, except that, for each wave, the voucher's value will be reduced approximately linearly so that it's value will be zero for the 21st wave. For later waves there will be no more Medicare-transition vouchers.

Should the patient purchase an inexpensive high-deductible policy, or no policy at all, any left-over HFA money could be used to pay out-of-pocket; and any money left over at death would pass to the estate.

Once this one start-up year is past, insurance companies would have no obligation to sell new high-deductible policies covering pre-existing conditions although they would be required to renew existing ones. This would ensure that every person would have the opportunity to start out the new system covered for high-deductible expenses for every illness, even pre-existing ones. They will not have to exhaust their voucher immediately to care for a chronic pre-existing condition.

During the transition, as well as after the transition, wage earners would have to contribute to their HFA from their wages. Medicare taxes could be continued or replaced by some other tax.

After 21 years the system will be at the steady-state and individuals older than 21 will be receiving no direct aid from the Federal government, except for the annual $200 use-or-lose preventive health care voucher and the BMP safety net, should they fall into that safety net.

It may be that it will seem necessary, at that time, perhaps to forestall too many individuals from falling into the safety net, to enlarge the program by providing an additional yearly voucher; perhaps $1000-2000/year to each citizen. This would be an efficient way for Congress to pump additional funds into the health care system. (top)

In theory, the new system should not result in a great increase in total medical expenses since most known illness is currently covered by Medicare, Medicaid or private insurance.

In practice, the new system might result in a substantial cost increase since insurance companies would likely offer pre-enrollment diagnostic programs to discover pre-existing conditions so that they will be covered by the Federal government. When found, treatment for these conditions would begin, whereas before they would not have been treated.

However, if future severe illnesses can be delayed, so that working life is prolonged, the average yearly cost of health care will be reduced because there will be more premiums/illness. As a result, in the future, yearly premiums can be reduced.

There will also likely be a rush to get care right away because of the many procedures that many poor people will need and be able to pay for, and hence for the first time, feel entitled to.

Due to the increase in demand there will likely be too few doctors, so, again, prices may go up.

Thus, indeed, prices may go up during the first few years after initiation until several new classes of doctors have graduated. Healthy people, for the most part young or well-off people, will put off procedures that they do not seriously need, until prices decline again.

Sick people, especially including poor sick people, will use up more of their HFA to finally get the care that they need. It is better for the society that they pay more in the early rush in order to regain health, substantially reducing the funds in their HFA, than that they remain sick and unproductive.

Costs may not rise as much as might be expected since doctors may strive for efficiency in order to increase market share. Even poor people may try to conserve their vouchers by seeking low-cost providers. In addition, over the longer term, the supply of doctors should increase since the best of them will command top dollar and their financial success should be an inducement to others to enter the field. (top)

There have been Medicaid frauds with patients, in collusion with doctors, falsely claiming diseases which require unnecessary procedures which can be reimbursed by the government. This behavior would be less, not more, likely if everyone had an HFA because someone's HFA, probably the false patient's, must be reduced, and that person, when notified of the loss, will not want their HFA reduced.

In frauds, under the current Medicare, Medicaid and private insurance approaches, only the government or a faceless insurance company loses money. This will not be the case under the HFA approach.

Rewards of increased health care voucher limits could be given to persons whose information about unauthorized charges to their account resulted in convictions.

Despite the above remarks, a common reaction to the subsidized HFA approach is to believe that it will be vulnerable to fraud.

Logically however, the HFA approach is no more vulnerable than any current plan, including employer supplied health plans, not to mention Medicaid and Medicare. The current penalties of trial and jail terms should be sufficient to deter frauds, which are even less in the putative patient's interest than before.

As a result of Medicaid and Medicare fraud the government has instituted draconian criminal laws against fraud which have occasionally resulted in honest doctors going to jail for what were basically clerical errors. Guarding against such errors further increases the administrative costs of health care, and makes it less likely that a medical career will be attractive to individuals who are capable of being excellent doctors.

Hence the HFA approach, by undercutting most approaches to fraud, should further reduce costs, improve the skills of those becoming doctors, and prevent the criminalization of medicine.

There is the possibility that the patient will not use the money for medical uses. This might be especially likely for wealthy individuals who can self-insure and hence are happy to use the voucher and HFA as just another investment whose proceeds can be passed on in their estate at death. But, except for the fact that the money is not used for medical purposes, the individuals do not become a further drain on society; so that this "fraud" is not really a serious problem. Also, the HFA is limited to a maximum value of $125,000 and should be, by law, invested only in very conservative fashion.

Another possible, more serious fraud would be that a poor person might try to convert their HFA into money for food, rent or drugs by colluding with a corrupt doctor to share a kickback from billing a false procedure. This is serious because the patient would then cause extra expense to society if they did become sick and have to use the BMP safety net.

This would be a particularly unsettling fraud and, if detected, the legal penalty perhaps should be that in the future the HFA could only be spent for a health insurance policy. One would hope that food stamps and subsidized housing would forestall most such fraud.

This type of fraud is probably well guarded against by the present HFA policy of requiring only a 10% penalty for withdrawal for non-medical policies. However it seems that we have to give up this option if we want to keep people out of the safety net.

To reduce fraud expenditures must be easy for the IRS to check in an audit. For this reason, it is important that most HFA purchases be mediated through a debit card. These purchases will be easy to verify and hence, any not purchased with such a debit card will draw attention. (top)

There will be privacy problems; the MA could have a complete record of a person's medical history. However, this is already the case in principle for people on Medicare, and with respect to states, for Medicaid. Perhaps this problem could be alleviated by requiring records over, say, five years old to be discarded. The total fraction of the HFA remaining would, naturally, have to be retained in the MA files. A person could still assure privacy by paying personally for care. As has always been the case, the actual doctors who provided care would still have records, assuming that met with the patient's approval. (top)

Much of a person's HFA might still be used in the last year of their life, an unproductive use from some taxpayers' points of view. However, in many cases much of the HFA will have been used creating a productive life before the last years.

Since after age 75 having a comfortable place to live may reasonably be characterized as medically desirable, and may forestall having to enter a nursing home and more surely falling into the safety net due to high nursing home expenses, it seems reasonable that the HFA funds could be spent for rent or mortgage payments after age 75.

Visiting nurses, nursing home care and other long-term care should also be approved procedures under the HFA. This will encourage people not to squander their voucher while they are young. Long-term care available under the BMP safety net will be at the 10th percentile and, as discussed above, could result in the requirement to move to the country in order to find inexpensive enough care.

An portion of the HFA varying from 0 to 100% would go to the patient's estate if he died before age 21. The remaining part of the HFA donated by the state would revert to the state. This is, in part, to ensure that parents take good care of their children, and in part because persons under age 21 have not usually acquired dependents for whom it would be desirable that they leave an estate. (top)

Although I am not certain at this time whether or not the following is a good idea, it might be useful to minimize the number of heartbreaking medical cases which can lead to individuals falling into the BMP safety net and becoming dependent on the state, and to ensure prompt treatment in the event of epidemics to have the federal government provide high-deductible insurance providing payments up to the 50th percentile for epidemics, accidents and natural disasters, also known as "Acts of God" as normally defined by insurance companies.

This insurance would cover, for example, the next $1,000,000 after the first $10,000 for injuries sustained in fires, floods, tornadoes, earthquakes, etc. Auto accidents would not be covered since they are already well-handled by the insurance industry. The first $10,000 could, of course, be covered by the victim's HFA.

Also, victims of assaults would receive comprehensive insurance, which also would pay at the 50th percentile, because relatively full restitution to such victims seems appropriate. Police and firefighters could be covered in this way.

So long as the concept of accidents is not extended to genetic "accidents" or the "accident" of catching tuberculosis or AIDS, the total cost of such insurance should be relatively small.

Presumably, emergency life-saving procedures with a good prognosis would be included under the BMP safety net list of procedures, so that a criminal, injured during a commission of a crime and convicted of that crime, would be covered at the 10th percentile for those procedures after their HFA had been exhausted. The fact that the criminal might exhaust their HFA if injured in a crime and convicted of that crime (so that the injury could not be classified as an assault) should act somewhat as a further deterrent to crime.

Private high-deductible insurance or the BMP would obviously not have to cover the same epidemics, natural disaster, accident, and assault risks as covered by this possible Federal policy; at least not above $10,000. top)

If health care vouchers/HFAs are a good idea it does not necessarily mean that vouchers for schools is a good idea since there is no "community" question in health care. However, it is apparent that a very similar set of vouchers could be created for education. The fact that most of the education money would be spent while a person was young and less qualified to make their own decisions, would be another difference between education and health care.

The fact that at least a high-deductible insurance policy is required up to age 21 is analogous to the fact that education, public, private or home-schooled is required up to a late teen age. top)

For political reasons, some existing expensive procedures, such as dialysis, organ transplants, would probably have to be grandfathered in to the list of safety net procedures either for those alive today or for those with symptoms today, even if they are not in the BMP because they are not in 95% of all private insurance policies sold.

However, in the future, if persons could not pay with HFA or high-deductible insurance that they have purchased, or from other private funds, it seems plausible that procedures for these diseases should be like all others. They either are or are not in the BMP. (top))

This HFA approach has many of the universal features of the Canadian system, with the additional cost-containing feature that costs are set by competition and not by doctor/hospital/government boards who have various conflicts of interest; and that patients are not indifferent to cost. This should result in better cost containment than in Canada and thus in more medical care delivered.

However the exact cost savings likely cannot be accurately estimated because, as far as is known, there are no examples and no experiments in which a truly free market in health care has operated in a medical high-tech society. The RAND Study was in a market where there was no incentive for the providers to lower costs; and most of the market was controlled by contracts between large entities. Analysis of plausible cost savings is an object of the highest priority and should be attempted by professional economists. (top)

TRADEOFF

Unfortunately, by comparison to the Canadian system, the HFA approach does have the defect that a disastrously ill person, who can be cured by a very expensive procedure not covered by the BMP but which is covered in Canada, and who for some reason has not purchased high-deductible insurance will be worse off. This is the price society must pay for the HFA system's other benefits as compared to the Canadian system. From another point of view, the corresponding tax dollars go toward good health for a large number of other individuals.

It may be objected that this approach to National Health Care is not sensitive to the needs of those who become very ill. However, those who become very ill and exhaust their medical financial resources will in fact still receive more tax dollars than the average citizen does. This is because, after they exhaust their voucher, they continue to draw tax dollars from the BMP.

In addition, the citizen who enjoys good health until their death, or who buys insurance and therefore does not exhaust their voucher, never personally receives the full value of their HFA.

The HFA approach has the political defect that it exhibits the rationing inherent in every system up front and puts a specific dollar amount on the HFA voucher's value. This amount is not as mean-spirited as it seems, however since it is less than the total that society would expend on someone who exhausted their voucher and who then used the BMP safety net.

The problem is one of degree; the society should provide the seriously ill citizen more than average funding. But should society allow that citizen to seriously compromise the health care available to the poor or average citizen by placing a tremendously disproportionate drain on tax dollars taken from all? (top)

DISABLED INDIVIDUALS

This approach has the benefit that if the HFAs are used to purchase lifetime guaranteed-renewable insurance at birth, as all can do with their vouchers, then if a person becomes disabled they will be covered. Moreover, for those who are disabled at start-up, the extra-cost premiums to cover the pre-existing conditions are subsidized by the federal government. (top)

RESEARCH AND TEACHING HOSPITALS

Research and teaching hospitals would no longer be able to subsidize research and teaching by overcharging paying customers through their insurers. Direct subsidies of these institutions by government may be required. On the other hand, even the poor will have some money in their HFAs, and may have purchased an insurance policy with that HFA money.

It might be possible for these institutions to charge extra (perhaps above the 50th percentile) to their paying customers due to their high quality of care, thus minimizing the subsidies required. The institutions would, of course, be free to reduce their fees to those poor persons with illnesses the examination of which would be useful for teaching or research. top)

DOLLAR VALUE OF VOUCHER

The HFA voucher supplement of $2000/year through age 21 is thought by some to be too small. It has been tentatively selected for the following reasons. The cost for family insurance is about $12000/year. This would come to about $3000/year per person. Such insurance is for the people, children and working adults, who the voucher plan is designed to cover. For 40 years that comes to $120,000/person. Since costs should come down with this plan $2000x21=$42,000 should be more than sufficient to carry the typical person well into their working life where they begin to further build up their HFA via wages transferred to their HFA.

In principle, the individual will have had a healthy life to provide for old age through the HFA which will support maintaining their high-deductible insurance; otherwise, there is the BMA safety net.

$42,000 is on the low end of the taxes spent per capita for public education through high school; to protect that investment it seems reasonable to spend a comparable amount on health care. It should be enough to ensure that almost every person can be launched into a healthy working life.

If the HFA voucher is too high people will not conserve or perhaps will not even buy high-deductible insurance. Prices will be lower with competition, so current prices may not be a reliable guide; more will be bought with this voucher than would be possible currently. To the extent that we are protecting a public education investment, we want healthy working people, not necessarily healthy retired people; so we want at least enough money to carry people up to a healthy retirement but we may save as a society on funding health care after retirement. (top)

Cafeteria plans are often proposed in which the citizen has the right each year during an "open period" to choose between several health care insurance options. A recent (2007) such plan is the Healthy American's Act suggested by Senator Ron Wyden of Oregon. A feature of this plan is that an HFA + high-deductible health care policy option may be offered. This plan must, however, be "actuarially equivalent" to the more standard plans which are to be much like a standard PPO policy.

Wyden's plan has the desirable feature of removing insurance from the domain of the employer, and of placing health care on the same tax status as other goods and services.

Industry is finding that the HFA option typically offers care at less expense for their employees. Hence it is likely that those choosing the HFA option will accumulate the marginal "actuarially equivalent" money in their HFA. Unfortunately, under the Wyden plan, rational employees will also likely put off expensive procedures until they have switched to a more comprehensive policy during the annual open season. Then, after they have undergone the expensive procedure, they will switch back to the HFA plan in the next open season. This will likely result in an instability or death spiral in which most citizens end up in the HFA option.

The instability might be reduced if citizens were allowed to switch back into an HFA policy only if they had stayed in a PPO policy for several years.

Wyden's plan might be regarded as a first step toward the plan advocated here if it were modified to require, not just allow, an HFA option to be offered. Many citizens would take up HFAs, and likely a substantial market would emerge at the procedure level. Many of those who initially chose standard PPO packages would, over time, see benefits to themselves of the HFA option.

Representative Paul Ryan (R-WI) has proposed the "Patient's Choice Act" (H.R. 2520 in 2009) which has many of the desirable features of the Wyden plan, but which also has a particular same defect in that there is annual open period for guaranteed issue together with no requirement to purchase insurance. Hence many individuals will not buy insurance until they become sick.

This defect could be corrected in both plans by government encouragement of suitable use of guaranteed-renewable, incentive compatible private policies. Such policies would eliminate the need for underwriting for most individuals.

As discussed above, one of the principal benefits of the HFA voucher approach is the resistance of HFA holders to fraud. There may be a tendency for poor individuals to be encouraged by criminals to select comprehensive insurance options so that they may receive a share of the fraudulent proceeds.

Senator Obama during his presidential campaign suggested that health insurance be mandatory only for children; this is similar to the plan suggested in this proposal.

President Bush offered in early 2007 a change in tax policy under which "families with health insurance will not pay income or payroll taxes on the first $15,000 in compensation, and singles will not pay income or payroll taxes on the first $7500. At the same time, health insurance would be considered taxable income. This is a change for those who now have health insurance through their jobs. Many of these ideas were also found in Senator McCain's health plan suggestions during his 2008 presidential campaign.

Governor Schwarzenegger has proposed (January, 2007) a universal health care program for California. His program does not seem to offer an HSA-type option for individuals. If it did so, then his program could serve as a state-experimental step toward the national approach outlined here.

David Goldhill, in the September 2009 issue of The Atlantic has published an article with a good analysis of the defects of our existing health care system and with recommendations for reform with similarities to the approach outlined here. Perhaps our agreement can be traced to the facts that both of us come from backgrounds completely disjoint from health care, and that both of us were motivated by the real-life experiences of relatives. (top)

PHILOSOPHICAL CONSIDERATIONS

The HFA is not transferable because it seems that allowing transferability would place too great a moral burden on individuals to gamble that they would not need health care and give the HFA away. Also, then the donors themselves may fall into the safety net and the total cost to society will be put off to a later time. Of course, one is free to give any other non-HFA money one may have. In the case of parents, for example, they may mortgage their house and take out loans if they are willing to sacrifice for their children's health care to this degree. This option might not be available in the Canadian system where all expenses are required to be paid by the state in the name of equality.

Such a problem exists today in the US in the case of kidney dialysis, where it is difficult to use personal funds to obtain the higher level of dialysis which provided in Europe and Japan and which is known to increase life expectancy over that which is available under the US dialysis practices funded by Medicaid. (top)