Cascade Policy Institute

This proposal is one of ten winning reports from the 1996 Oregon Better Government Competition. The 1994 and 1996 Competitions were organized by the Portland-based Cascade Policy Institute. Opinions expressed are those of the author(s) and not necessarily those of Cascade staff or advisors, nor should they be construed as an attempt by Cascade Policy Institute to influence any election or legislation.

Improving Oregon's Medicaid Program:
Why Not Try Means-tested
and Risk-adjusted Vouchers?

by State Representative Patti Milne and Jim Seagraves

Woodburn and Oregon City, Oregon

Executive Summary

Oregon's Medicaid expansion program, which extends managed care benefits to 115,000 people with low incomes, has a serious problem. Costs will rise at unacceptable rates because of two perverse incentives: the program causes some people to keep their income below the poverty level, and once on Medicaid, people have little incentive to economize on the use of services. This problem can be alleviated with a system of vouchers based on income and health risks.

The vouchers may be used to buy approved health-care coverage including high-deductible policies and medical savings accounts. Individuals may supplement the voucher with their own money. If an employee chooses coverage provided by an employer, the voucher would go to that insurer. The employer and employee would decide how to pay any difference.

Means-tested and risk adjusted vouchers, plus a refundable tax credit for individual outlays for insurance and medical savings accounts, will: Make it easier for people to move off welfare; Increase the number of Oregonians with health coverage; Reduce spending per person on health care; allow for increased choice and competition; and, Improve the quality of health services.

About the Author

Patti Milne is currently serving as State Representaive for District 38, and will be starting her third term in January. Jim Seagraves, PhD. is an economist and a Professor Emeritus from North Carolina State University.

I. INTRODUCTION

Oregon's Medicaid expansion program commands a great deal of National attention. Extending health insurance to everyone below the poverty level -- an additional 115,000 Oregonians -- while restricting the covered treatments to those that are most cost-effective seems practical. Actually, the main way costs are controlled is by maximizing the use of managed-care organizations (MCOs) and limiting the amount paid per capita.

However, the costs of Oregon's Medicaid expansion program will continue to rise at unsustainable levels because of two perverse incentives: 1) some people keep their reported income below the poverty level just to retain their health coverage, and 2) once on Medicaid, people have little financial incentive to economize on the use of medical services or to maintain a healthy life-style.

Means-tested and risk-adjusted vouchers can alleviate these problems. In Addition, two other measures proposed here will help many people with lower incomes -- replacing regressive tax deductions with refundable tax credits and giving everyone more choice, including a high-deductible policy plus a medical savings account (MSA). An MSA is a separate, tax-exempt savings account similar to an IRA. Money in an MSA may be used to pay any medical expenses. Earnings in the account accumulate tax free.

Benefits expected from these changes include a reduction in the number of families in poverty, an increase in the number of people with health benefit plans, a reduction in spending on health care, an increase in price competition and choice, and a healthier population. About 313,000 persons with incomes under 200 percent of the poverty level have health benefit plans. The proposed changes would cause an 18 percent increase in the number insured.

II. GOALS AND ASSUMPTIONS

Disputes about policy often stem from a lack of clarity about goals. Before trying to choose among alternative policies it is important to agree on goals. We assume these goals for health reform: 1) to guarantee that everyone will have access to high-quality health care and financial protection from the cost of catastrophic illnesses, 2) to improve the fairness and efficiency of the existing system, and 3) to devise a system that is as simple and as self-regulatory as possible. We hope the reader will accept these goals, at least for now.

"Oregon's goals for health care reform are stated in statute: 'universal access to an adequate level of high quality health care at an affordable cost.' While there has been significant progress through voluntary efforts and through state funding to support health insurance for the poorest citizens, a significant number of Oregonians [about 400,000] remain without health care coverage. ...There is no perfect answer. ... After all of the data, reports and studies reviewed by this office, it is hard to avoid the conclusion that money and mandates probably provide the answers." (Oregon Health Plan Administrator (OHPA), Mar. 1995, p. 79, emphasis added.)

The statutory goal is access to high-quality health "care," but under current policy the most important method of achieving that access is by extending "coverage." Coverage has come to mean a basic health care package, preferably through an MCO, with no co-payment per visit. Coverage has become an unstated goal. It is not surprising that "money and mandates" are the answers.

Important assumptions that we share with the administrators of the Oregon Health Plan (OHP) are that many people who are now poor want to lift themselves out of poverty, that subsidies are needed to help these people buy health care, and that the lower the income and the more serious the health problems the larger the subsidies need to be. With these goals and assumptions in mind, our purpose is to design an incentive-based alternative which will help people in the Medicaid expansion program become more self-supporting.

III. OREGON'S MEDICAID EXPANSION PROGRAM AND SOME ALTERNATIVES:

A. Oregon's Medicaid Expansion Program

The overall Medicaid program covers about 375,000 Oregonians with a budget of about $850 million per year. It is funded with approximately 62 percent federal and 38 percent State money. Our proposal and pilot project only pertain to the 115,000 "non-categoricals," who are enrolled in the Medicaid expansion program simply because their income is below the federal poverty level (FPL). Even though they represent 30 percent of the people on Medicaid, they only require about 23 percent of the budget.

Many people who qualify still do not know about the program. They learn about the Medicaid expansion program when they end up in an emergency room. The hospital asks about their income and savings, and, if they qualify, enrolls them on the spot. Individuals who enroll are insured on a fee-for-service basis for about a month while their income is being checked. Once their application is approved, they must choose an MCO.

Many other states also are using MCOs to control Medicaid costs. See the story about Tennessee in The National Weekly Edition of The Washington Post. For Oregon's Medicaid as a whole, the goal is to enroll 90 percent in MCOs (The Oregonian, 6/21/96). For the expansion program, the goal must be to enroll about 100 percent in MCOs. The MCOs provide those enrolled with at least the "Basic Health Care Package."

"The Basic Health Care Package" is based on a priority list of health services created by the Oregon Health Services Commission. As of January 1, 1996, the package covers the first 581 of 744 condition/treatment pairs. The package stresses prevention, including services such as maternity and newborn care, immunizations, well-child exams and preventive dental care. It also covers all major diseases of women and children, reasonable diagnostic services, dental services, prescription drugs, many transplants, hospice care, [and] mammograms. The package also covers outpatient chemical dependency services. Mental health services are covered for about 25 percent of OHP-eligible (to be phased into 100 percent after July 1997).

"The Basic Health Care Package" does not pay to treat conditions that get better on their own; conditions for which home treatments are effective; cosmetic conditions; and conditions for which treatment is generally ineffective (such as advanced cancer) although in such cases comfort care is provided." (Office of Medical Assistance Programs, OMAP, May 1996.)

Oregon and many other states are searching for ways to reduce the rising costs of Medicaid. To stay within the 1995-97 budget and cope with a projected cost overrun of 2.9 percent ($18.4 million on a State budget of $638 million for the biennium), the legislature's Joint Interim Task Force (June 1996) authorized a number of measures, including accelerating the transfer of the elderly and disabled people from fee-for service into managed care, delaying slightly the eligibility date of new enrollees, reducing the fee-for-service reimbursements by 5 percent, postponing increases in managed-care rates, and reducing the asset limit from $5,000 to $2,000 for new enrollees. Most of these measures affect the expansion program.

In January 1996, the State instituted a policy of "cost sharing" in the form of small premiums to be paid by those enrolled in the expansion program. Premiums range from $6 to $28 per month depending on income and family size. The current collection rate is 59 percent, instead of an anticipated 75 percent (OHF, August 1996). Sharing the cost of a fixed premium will do nothing to reduce excessive use of managed-care services by those who remain in the Medicaid program.⁽¹⁾

There is extreme resistance at the national level to the idea of co-payments for those on Medicaid. However, many MCOs find that co-pays are effective in reducing frivolous visits by their non-Medicaid clients. Also, MCOs currently seem to resist the idea of deductibles.

B. Alternatives for Oregon to Consider The Office of Health Plan Administrator prepared a "Report on Alternatives for Increased Health Coverage" for the State Legislature (OHPA, March 1995). One alternative is called the "voluntary option." This involves a large increase in the number of children and pregnant women covered by managed care which is achieved by raising the income threshold from 133 to 185 percent of the FPL. A net increase of 95,000 in the number of people covered is achieved at a combined private and public cost of zero. The changes in annual costs offset one another as follows: government spending increases by $140 million, employers save $20 million, and individuals reduce spending by $,120 million. Part of the explanation of this remarkable result is the bargaining power of the State; another part is that the MCOs are less costly than privately purchased health care and fee-for-service Medicaid.

That report also defines an "employer/individual mandate." It has one element in common with our proposal -- sliding-scale subsidies. Their subsidies apply to people earning from 100 to 250 percent of the poverty level. Individuals are required to have insurance. If they are employed more than 10 hours a week, their employer is required to contribute something toward the cost of coverage. If they are employed more than 30 hours per week, their employer must pay half of the premiums for employees and their families. The State will continue to fully subsidize coverage of individuals and families below the FPL. The changes in total costs are as follows: employers add $190 million, individuals save $133 million, and the governments increase spending by $186 million. Assuming, as OHPA does, that the 418,000 people now without insurance will be covered under this set of mandates, the added social (private plus public) cost is only $581 per person. The average total cost falls from $2,412 per person covered under the current system to $2,171 per person under this set of employer and individual mandates.⁽²⁾

The voluntary program we propose below would cost much less than OHPA's employer/individual mandate because our sliding scale goes from 50 to 150 percent, instead of 100 to 250 percent of the FPL. Ours has no mandates, is based on risk-adjusted vouchers, and will encourage price competition in the markets for individual coverage and other health services.

The Health Education Alliance (HEAL) in Montana has a comprehensive plan for health-care reform which includes optional high-deductible policies and MSAs for those on Medicaid who are not aged, blind, or disabled. The State would provide personal health-care credits (PHCCs) to persons on Aid to Families with Dependent Children (AFDC) who are at or below the FPL. At the outset, PHCCs would equal "the average state premiums for the age and benefit matched non-Medicaid groups/individuals in the state" (HEAL, p.41). The State pays insurers to provide participant with coverage and also pays co-payments and deductibles from each person's PHCC. If, at the end of the year, part of the PHCC is not spent, the State will send half of that amount to that individual's MSA. If the amount in the MSA exceeds twice the annual deductible, the balance may be withdrawn for non-medical purposes and taxed as regular income.

One attractive feature of the Montana proposal is that in order to qualify for the PHCC system individuals "would have to attend a session instructing them in their options, the use of credit and MSAs, how to use the price comparisons, how to check bills, how to obtain refunds [for finding errors in bills, and] what preventive services offer refunds" (HEAL, p.22). Another feature is that Medicaid recipients who qualify for Montana's high-risk pool could send their PHCC plus their share of the premium to the pool.

A recent study of Montana's Medicaid population indicated that "recipients do not lack the incentive to take care of themselves, just the resources." Many AFDC recipients "accept the value of working hard to support themselves," and "do not intend to stay on the program." Also, "any program that will move them toward self-sufficiency, and not punish them for such movement, will probably meet with considerable success" (HEAL, pp. 31 & 32).

The HEAL proposal is patterned after a 1994 Indiana bill which passed the Senate but not the House. Milliman & Robertson estimated the Indiana proposal would save about 14%, and recently concluded that HEAL's plan for AFDC would save Montana 4 to 11 percent. We assume that the recipients also would save about the same amount.

The HEAL proposal is as close to ours as any we have found. Both involve risk-adjusted payments to insurers, MSAs, and a sharing of the money saved between the individuals and Medicaid. Our mean-testing idea might appeal to Montana, and their tightly controlled lines of credit might work better for us than vouchers. Publications by Cantwell and by Mittler and Matthews have also recommended MSAs for those on Medicaid.

IV. PROPOSED REFORMS, BENEFITS, IMPLEMENTATION PLAN, AND PROBLEMS

A. Proposed Reforms

We propose using the federal and State funds now allocated to the Medicaid expansion program to extend means-tested and risk-adjusted vouchers to those with incomes below 150 percent of the FPL. Initially, we recommend a pilot project in two of the five regions of the State to test the program and its assumptions.

To illustrate the proposed reforms, we assume a 1997 FPL of $16,000 per year for a family of four. If a family reports an income last quarter below 50 percent of the FPL, or less than $2,000, they will be given a voucher equal to the full value of their expected cost of Medicaid (ECM). Each family's ECM would be risk-adjusted so that it depends mainly on the age and gender of family members, but also on any special health problems of individuals in that family.⁽³⁾

For families with higher incomes, the voucher's percentage of the ECM declines on a sliding scale until, for those with incomes above 150 percent of the FPL, or $6,000 per quarter, the voucher is zero. Using a straight-line sliding scale, people having an income equal to 100 percent of the FPL during the last quarter will receive a voucher equal to 50 percent of their ECM throughout the current quarter.

Medicaid vouchers are first applied to the purchase of approved health-care coverage.⁽⁴⁾ Individuals may supplement the voucher with their own money. If an employee chooses coverage provided by an employer, the voucher would go to that insurer. The employer and employee would decide how to pay any difference.

If the voucher is for more than the cost of health coverage the remainder is either deposited in the recipient's medical savings account or reverts to the State. Recipients of Medicaid vouchers whom the Department of Health and Human Services determines, according to rules, are capable of making appropriate health-care decisions may purchase a basic high-deductible health benefit plan. At least three-fourths of the difference between the ECM and the cost of coverage must be deposited in the recipient's MSA each year. Individuals may deposit more than that amount of their own money in their MSA, but the insurer who receives a Medicaid voucher will deposit no more than three-fourths of that difference in an insured's MSA. Any surplus value in the voucher after paying for a high-deductible policy and depositing the minimum required amount in the individual's MSA will be returned to the State.⁽⁵⁾

As an incentive for everyone to maintain health-care coverage and make deposits to an MSA, we also propose a 9-percent refundable Oregon tax credit.⁽⁶⁾ Refundable tax credits replace income-tax deductions. This new 9-percent tax credit will apply regardless of whether one itemizes deductions or owes any taxes. To receive a voucher and/or the tax credit, every member of the household must be covered by at least an approved high-deductible health benefit plan.

Individual's may purchase a generous list of health-related items from an MSA -- all those described in section 213 (d) of the Internal Revenue Code. Trustees will be in charge of administering MSAs and may help account holders decide which withdrawals are legitimate medical expenses. Other features of MSAs include tax-free gains on balances in these accounts, and withdrawals for non-medical purposes would be subject to limits and penalties (See Appendix 1).

Some of those in the Medicaid expansion program may not be capable of making decisions regarding their own health care. The option we are proposing -- freedom to choose a high-deductible policy and to responsibly manage an MSA -- is only appropriate for those who are capable. Part of the purpose of the pilot project is to determine how many of those in the Medicaid expansion program have the necessary motivation and ability.

Critics of MSAs claim they can't help the poor who have nothing to save and they will appeal mostly to the healthy and the wealthy. The experience of Golden Rule Insurance Company supports the notion that lower-income workers are more cost conscious and more likely to choose MSAs than their higher paid counterparts.⁽⁷⁾

It is also said that high-deductible policies and MSAs will appeal mainly to healthy people and, through "adverse selection," cause an increase in premiums for those who are less healthy. That notion is based on a mistaken assumption that community rating is somehow desirable in the sale of health plans to individuals.⁽⁸⁾ Here, risk-adjusted vouchers will cause insurers to charge premiums which are also based on risk. Also, vouchers and refundable tax credits will cause some MCOs to offer high-deductible policies, a price list for their services, and MSAs.

For example, a family of four might have an ECM of $350 per month. If their income last quarter was $4,000 which is the same as the FPL, they will receive Medicaid vouchers equal to half their ECM, or $175 per month. Let's assume they buy a high-deductible policy for $200 per month. The difference, $25, comes from their own pocket. For the insurance costs they pay themselves and any deposits they make to an MSA, they receive a 9-percent refundable tax credit when they file their taxes. They are obligated to deposit $112.50 (three-fourths of $350 - $200) per month in an MSA. Their out-of-pocket medical cost would be $137.50 per month, or $1,650 per year less the 9-percent tax credit.

Following this example further, the family's annual income is $16,000, but they receive an earned-income tax credit of $2,143, pay a Social Security tax of $1,202, pay a State income tax of $388, and receive a medical tax credit of $148 (9-percent of $1,650). Their after-tax income is $16,701. Of the $1,350 per year they must deposit in an MSA, about one-third, or $450, might be saved in a typical year and carried over to future years to meet unexpected health-related problems not covered by their high-deductible policy. Their net cost of health care is $1,650 - $450 - $148 = $1,052, or 6.3 percent of $16,701.⁽⁹⁾

The administration of health benefits will be affected only slightly by these changes.⁽¹⁰⁾ Currently, clients choose an MCO which is then in charge of most decisions about delivery. Under our proposal, MCOs may charge co-payments for visits and prescriptions, post price schedules, charge risk-adjusted premiums, sell high-deductible policies, and administer MSAs for their clients. Clients are free to choose a wide range of insurance policies and health-related services and pay for them from their MSAs. Given all these choices, the State or the MCOs will want to develop programs to help people make good choices; and to advise some people to stick with conventional MCO policies. In the next section, we assume that one-third of those receiving Medicaid vouchers will choose to continue their full coverage with an MCO.

B. The Expected Benefits

Our proposed reforms offer five main benefits: 1) an increase in the number of people who move "out of poverty" because they will no longer be penalized if they earn more, 2) an increase in the number of people voluntarily insured, 3) a reduction in total spending on health care, 4) more competition and choice, and 5) a gradual increase in the quality of health care. At this time, our estimates of these benefits are based on assumptions, many of which are relative matters. One reason for conducting a pilot project will be to get a better understanding of these assumptions.

To illustrate these benefits, we compare the current system with our means-tested vouchers for people with incomes between 50 and 150 percent of FPL. These numbers are in 1996 dollars and pertain to the whole State as if vouchers and refundable tax credits had been fully implemented for some time.

Table 1. Results for Two Systems

Appendix 2 lists the assumptions behind these estimates.

The reader will see that it is easy to alter the assumptions. The most critical are that two-thirds of the people will choose high-deductible policies and spend three-fourths as much as at present. We project an 18 percent increase in the number of people covered including 18,800 with incomes between 150 and 200 percent of the FPL. There is only a one percent increase in cost of the program to the two government but a 24 percent increase in combined private and public costs.

B. 1. Added Incentive to Work

If these reforms are implemented, a "sharp edge" will no longer exist at the FPL. People will have little reason to work less to retain their Medicaid. They will tend to retain the same medical protection as their income fluctuates. People won't even notice it when they "go off Medicaid." If this quarter their income exceeds 150 percent of the FPL, then next quarter their insurer won't receive a Medicaid voucher. The insurer's bill will reflect this. Also, no stigma will be attached to "going back on Medicaid."

B. 2. An Increase in the Number Voluntarily Insured

Means-tested vouchers will give employers an incentive to provide health benefits to more low-paid and part-time employees and to help employees retain such benefits when their incomes increase. When an employee's family income is close to the poverty level, the present system gives an employer and employee a strong incentive to agree that the family should be enrolled in Medicaid instead of employer-provided insurance. They can agree to adjust the hours worked to achieve this result.⁽¹¹⁾

Also, many people now without insurance are under the age of 30 with incomes between 100 and 200 percent of the FPL. Many are self-employed or independent contractors, and their income fluctuates so that at times they will receive vouchers. People in this category are virtually left out of current subsidies for health-related expenditures. Most of what they would spend for health insurance is not exempt from income and Social Security taxes. Many of them can afford a catastrophic policy and would buy one if it were competitively priced and they had the refundable tax credit. Also, many would put some tax-free money in an MSA. In the above table we assumed that progressive vouchers for those between 0 and 150 percent of the FPL plus refundable tax credits for everyone will cause 21 percent of those from 100 - 150 percent of the FPL plus 8 percent of those from 150 - 200 percent of the FPL to acquire approved coverage.

B. 3. Reduced Spending on Health Care

We assume that managed care has reduced health-care spending by about 15 percent and that MSAs can reduce spending by another 25 percent.⁽¹²⁾ We assume that about 152,800 persons will have incomes below 150 percent of the FPL and accept medicaid vouchers. If two-thirds of them opt for high-deductible policies plus MSAs, the combined savings is $31.5 million per year.

B. 4. Enhanced Competition and Choice

Oregon is virtually forcing everyone who enrolls in Medicaid to join an MCO. The State pays the same flat amount per person or per family to each MCO; it's called "price control." Evidence that the price may be fixed too low is that most of the MCOs are not now actively seeking Medicaid clients; at least one, Prime Care, has dropped out.⁽¹³⁾ OMAP is considering a system of competitive bidding for Medicaid contracts, which will probably result in fewer MCOs operating in each region and even less competition. The "winning" firms will become even more dependent on Medicaid.

In our proposal, risk-adjusted vouchers are the key to stimulating genuine competition. MCOs and fee-for-service insurers will compete for clients in each risk class. Premiums will not be controlled and may be risk-adjusted based on each MCO's judgement as to the risk of serving an individual. The State will not attempt to regulate the quality of service but simply report on it. The State Insurance Division may promote using a few standard contracts. That too will enhance competition.

Choice in health care will be enhanced by increasing competition, by giving people an opportunity to buy high-deductible policies, and by having a wide range of services that may be purchased from an MSA.

B. 5. A Gradual Improvement in Health

Under our proposal, people of all income groups will have continual financial incentives to take good care of themselves and to use preventive measures. First, if they don't take good care of themselves, they will spend the savings in their MSAs on medical expenses that could have been avoided. Second, if they don't exercise common sense in caring for themselves, competitive insurers using risk-adjusted premium classes may "down-grade" them or raise their insurance premiums. Under our proposal, smoking would be expensive for both of these reasons. In contrast, under Oregon's current Medicaid program, the main financial incentive not to smoke is the tobacco tax.

For people with incomes below 150 percent of the FPL, three of the above-mentioned benefits -- the greater incentive to work, the increase in the number insured, and the enhanced choice -- can contribute indirectly to long-run good health. By taking a more responsible attitude toward their own health, many also will end up with money in an MSA and more freedom.

On the other hand, some argue that high-deductible policies and MSAs will cause people to use fewer preventive services, such as immunizations; and that charging low-income people for such things would be a mistake. These services now are free to Medicaid recipients, and all the MCOs must emphasize prevention. It is argued that failure to use preventive measures early often results in complications and higher costs later on.

We assume that some competitive MCOs will offer high-deductible policies in the future and find it advantageous to explain the value of preventive services to all their clients. The evaluation of our pilot project must test assumptions regarding the use of "free" versus "purchased" preventive services and the motivation to take good care of one's self. The evaluation should determine what percentage of lower-income people benefit from incentives to be more self reliant.

C. The Implementation Process - Time Line

We propose a pilot project in which our vouchers will be tried in two regions. An independent research institution will evaluate the results. The 1997 legislature would authorize a Joint Interim Task Force to determine regions for the pilot project and the institution to do the evaluation after considering the recommendations of the Department of Human Resources and the Governor.

As part of the evaluation, preliminary measurements or benchmarks should be established for each of the five Medicaid regions for at least the following: 1) costs of administration, 2) quality of service, 3) consumer satisfaction, and 4) health-related spending by different demographic groups. Annual reports will compare data from regions with and without the vouchers. Changes from the benchmark year also will be noted for all five regions. Some method of measuring the change in competitiveness -- such as the price of given services, new products offered, the Herfindahl index, or the four-firm concentration ratio -- should be attempted.

The implementation has this optimistic time line:

1) In the fall of 1996, a) the proposal is explained and debated, b) the OHPA includes an explanation of something similar to this proposal as a "voluntary option" in a report to the 1997 legislature which summarizes alternatives, and c) the Governor indicates support of this voluntary Medicaid expansion and the pilot project.

2) In the first half of 1997, a) the legislature passes an MSA bill with refundable tax credits, passes another bill authorizing the first 15 months (October 1, 1997 - January 1, 1999) of a Medicaid pilot project trying out means-tested and risk-adjusted vouchers, and appoints members to a Joint Interim Task Force, b) the State applies for federal permission to amend its 1115 demonstration waiver to allow for the pilot project, and c) the permission is quickly granted. (Oregon's 5-year demonstration, the Medicaid expansion, ends January 1, 1999.)

3) In July 1997, the regions for the pilot project are selected and the independent evaluator begins assembling the preliminary measurements.

4) October 1, l997 - June 30, 1999, a) the pilot project begins October 1, b) it looks like it will be both interesting and valuable, c) the Joint Interim Task Force votes to extend it to January 1, 2002, and d) the 1999 legislature concurs.

D. Problems with Acceptance and Implementation

We envision four small problems: those who favor community rating may fight our risk-adjusted vouchers, MCOs might fight the idea of high-deductible policies and competition, the defining of risk classes, and appeal procedures for contested ECMs.

First, community rating (CR) is widely accepted making it hard to explain the advantages of the opposite -- risk-adjusted premiums in the sale of individual policies to voucher holders. Activists may object strongly to risk-adjusted vouchers, causing some politicians to refuse to even consider the idea.

It might be helpful to explain the popularity of CR. It is based on the widespread adoption in the U.S. of group health plans offered by employers. These usually involve the same payment per person or CR to the insurer. This tradition is based on very generous tax breaks to employer-provided benefits which began during World War II. Such health benefits now enjoy a 40- to 50-percent tax advantage over insurance that other individuals must purchase with after-tax dollars. Not surprisingly, many people are convinced that employer-provided health benefits are the best "pill" for everyone. However, concluding that CR is desirable in the sale of policies to individuals just because it is widely used and subsidized in group policies is not logical.

Defining two types of community rating might help:

1) CR means that everyone pays the same insurance premium even though the benefits expected, provided, and needed differ widely, and

2) CR taxes people in a variety of ways but has the State pay the same amount per capita to MCOs.

The first type represents a flat tax per person, which we know to be regressive. States that have tried it for health insurance have proven the obvious; that making individuals pay the same amount for widely different insurance benefits is a good way to kill private insurance.

Oregon is applying the second type of CR in its Medicaid expansion program.⁽¹⁴⁾

It is a simple way to achieve "to each according to his need." It forces the prepaid health plans to redistribute income from less costly to more costly clients. The problem is that CR plus competition eventually will force each MCO to find subtle ways to avoid covering less healthy individuals. The MCOs that treat these people fairly won't be able to survive.

Vouchers based on age, gender, and special health problems will make it possible for competitive MCOs to want to serve everyone, including the chronically ill. The competitive response will include risk-adjusted premiums and a variety of policies appealing to different groups.

When governments provide insurance directly to individuals, as with fee-for-service Medicare or nationalized single-payer systems, CR is irrelevant. It is understood that the government is responsible for redistributions from the healthy and the wealthy to the less healthy and the poor. However, people who recommend CR for individual insurance contracts, including people on Medicaid choosing an MCO, should understand why it destroys private coverage and leads either to a single controlled provider or to the socialization of health insurance. If policy makers don't want that result, they should recognize that CR is but a simplistic, short-term solution to a problem and look for a way to encourage real competition. Risk-adjusted vouchers and premiums are the answer.

Second, MCOs form a powerful lobby. They could resist our proposal because they think they are better off with more government subsidies and regulations and less competition. However, what one government gives, another may take away. Accepting competition and promoting MSAs is a logical way for MCOs to protect themselves in the long run.

Third, disputes over how to define risk classes could be a problem. The experiences of insurance companies, managed-care providers, actuaries, and other states and countries need to be reviewed prior to creating risk groups and ECMs.

Fourth, individuals asking for higher ECMs might become an administrative headache. One way to cope with contested ECMs would be to help people use their vouchers to shop among managed-care and other providers. If several insurers explain why a person's risk-adjusted ECM must be much higher, then the State should consider adjusting it upward.

V. CONCLUSIONS

Oregon's Medicaid expansion program has two serious problems: a strong incentive for people to keep their income below the poverty level and very little financial incentive to take good care of themselves. These problems will lead to increasingly higher costs and a gradual deterioration in the quality of service to the chronically ill -- the people who need it most.

A common-sense way to improve Oregon's Medicaid program and reduce its cost is to offer means-tested and risk-adjusted vouchers plus an opportunity for many people in this program to choose the type of coverage that suits them. Some will buy a high-deductible policy and gradually save money in their own MSA. Choice and competition will help everyone secure the quality of care they want and allow the State to gradually reduce its regulatory functions and the cost of this program.

The first and most important benefit of this proposal will be an improvement in Oregon's welfare programs by removing a strong incentive for people to keep their income below the poverty level. Other benefits will be an increase in the number of people who have health coverage, a reduction in cost by about 16 percent per person, expanded competition and choice, and a healthier population.

REFERENCES

Barchet, Stephen, "Medical Savings Accounts, A Building Block for Sound Health Care," Evergreen Freedom Foundation, PO Box 552, Olympia, WA 98507, 1995.

Cantwell, James R, Reforming Medicaid, NCPA Policy Report No. 197, August 1995. The address of the National Center for Policy Analysis is 12655 N Central Expy, Suite 720, Dallas, TX 75243-1739.

Cutler, David and Jonathan Gruber, "Does Public Insurance Crowd Out Private Insurance?" NBER Working Paper No. 5082, summarized in The NBER Digest, May 1995.

HEAL, the Health Education Aliance of Montana, MEDI*CHOICE, P.O. Box 111, Great Falls, MT 59403, 1995.

The Joint Interim Task Force, "Final Report on the Oregon Health Plan to the Emergency Board," June 1996.

Mittler, Brant S., and Merrill Matthews Jr., "Can Managed Care Solve the Medicaid Crisis?" NCPA Brief Analysis No. 155, April 10, 1955.

The National Weekly Edition of The Washington Post, July 15 - 21, and July 22 - 28, 1996.

OHF, the Oregon Health Forum, P.O. Box 2942, Portland, OR 97208.

OHPA, Office of Health Plan Administrator, "Alternatives for Increased Health Coverage," March 8, 1995.

OHPA, "The Uninsured in Oregon, Who Are They?", December, 1995.

OMAP, Office of Medical Assistance Programs, "Oregon Health Plan, Highlights/Fact Sheet," May 1996,

The Oregonian, 6/21/96.

APPENDIX 1. MEDICAL SAVINGS ACCOUNT BILLS FOR OREGON

Part of the Kassebaum/Kennedy health insurance reform bill of 1996 provides federal income-tax breaks for MSA's. But these are limited to a demonstration with 750,000 policies over a four year period. Policies will be available only to: firms with fewer than 50 employees, the self-employed, and the "uninsured."

In order to make MSAs available to all Oregonians, the next legislature should consider two bills. One which extends to those who get into the federal demonstration a parallel exclusion from Oregon's taxable income. That should be part of the 1997 "re-connect bill." The second bill will allow all other Oregonians to have similar (federal-like) MSAs, but with only the State income-tax exclusions. The main features of the second bill are:

1.) A generous list of health-related items may be purchased from MSAs. The list includes all items described by section 213 (d) of the Internal Revenue Code. [The federal bill has complex restrictions on which types of health insurance may be purchased from an MSA and when. We do not recommend that those regulations be included in the second bill.]

2.) People with MSAs must also have a high-deductible health benefit plan which is approved by Oregon's Department of Consumer and Business Services. The deductibles must be between $1,500 and $2,250 for individuals, and between $3,000 and $4,500 for families. HMOs may offer high-deductible plans. [The federal bill specifies maximum out-of-pocket expenses for agreed-upon deductibles and co-payments, $3,000 for individuals and $5,500 for families. We would not include these regulations in the Oregon bill.]

3.) Deposits to MSAs are encouraged by letting employers exclude them from employees' taxable income, or by granting individuals a 9-percent refundable tax credit on their own deposits. Tax-favored deposits are limited to 65 percent of the actual deductible chosen by an individual and 75 percent of deductible chosen by a family. Total annual deposits (tax-free and non-tax-free) are limited to 75 percent of maximum allowable deductibles, or to $1,687 for individuals and to $3,375 for families. Tax-favored contributions may be made either by an employer or an employee, but not both in any one calendar year. Employee contributions must be "comparable" for all participating employees. Funds may be distributed from one MSA to another (rolled over) once a year without being subject to tax, penalty, or contribution limits, provided they are deposited within 60 days of the distribution. [Under the federal rules, tax-free contributions can only be made by or for people under 65 years of age and others who are not on Medicare. We don't feel that restriction should be part of the Oregon bill.]

4.) Trustees must administer MSAs. Trustees may be a financial institution, an insurance company, an employer with a self-insurance plan, or an entity approved by the U.S. Secretary of the Treasury. They make annual reports to the State Department of Consumer and Business Services regarding each accounts' deposits, withdrawals, earnings, and balances. Trustees may issue MSA debit cards and extend lines of credit to customers.

5.) Interest and gains on balances in MSAs are not taxed at the State level.

6.) Withdrawals for non-medical purposes are taxed and subject to a penalty of 15 percent (either federal or State, but not both). However, the penalty does not apply to those who are over 65, disabled, or deceased.

7.)Upon the death of an account holder, balances that are transferred to the MSAs of a spouse will not be taxed. Balances transferred to an estate are taxed as income but not subject to inheritance taxes.

APPENDIX 2. ASSUMPTIONS USED IN ESTIMATING NET BENEFITS

To estimate the benefits of means-tested vouchers listed in Table 2 we make these simplifying assumptions:

1.) ECMs are the same for everyone and equal to current managed-care Medicaid premiums: $1,644/person/year.

2.) High-deductible policies with deductibles of $1,500 for individuals and $3,000 for families cost one-half of their ECMs.

3.) Those who receive vouchers and choose high-deductible policies deposit three-fourths of the difference between their ECM and the cost of the high deductible policy in their MSA but only spend in an average year two-thirds of what they put into the MSAs.

4.) Individuals get a refundable tax credit of 9 percent of whatever they pay out-of-pocket for insurance and MSA deposits.

5.) After the program is in place, two-thirds of those affected choose high-deductible policies plus MSAs and the other one-third stay with the basic MCO policy.

6.) The numbers assumed and calculated for each income group are listed in Table 2. Voucher Program by Income groups on following page.

ACKNOWLEDGEMENTS

We wish to thank the many people who helped with the preparation of this paper without implicating them in our recommendations or any errors that remain.e about. We are grateful to a number of public servants for accepting our revolutionary perspective and making constructive suggestions; especially we thank Hersh Crawford, Director of OMAP, and Chuck Sigmund and Bob DiPrete of the OHPA. Dr. Steve Barchet and Dr. Paul Smith offered valuable counsel. Early on, Ian Timm, Director of the Oregon Primary Care Association, also helped shake our assumptions. Dexter Johnson and Gwen Dayton helped tighten up our ideas while shaping them into legislation. Francisco Perez was a good companion and a help with literature searches and editing. Sharon Barrell made many valuable editorial and structural suggestions. Diane Lund, Justin Zimmerman, Merrill Matthews Jr., Dr. Brant Mittler, Dr. Paul Gorsuch, Elise Bernadas, Mary Lou Cooper, and Aimee Charette helped us with references and contacts.

ENDNOTES

1. Also, we anticipate that cost-sharing and efforts to enforce it will result in a small decline in the number of people enrolled.

2. From the tables on pages 71, 73, and 78 of OHPA, 1995.

3. We assume that ECMs for individuals based on age and gender alone would vary about 4 to 1 and are still seeking good references to average risk-adjusted premiums when MCOs sell individual policies.

4. We suggest that a special task force with representatives from the Health Services Commission, the Department of Human Resources, and the sectors affected be appointed to advise the Department of Consumer and Business Affairs on the definition of minimally acceptable policies.

5. State legislation (LC 1247) has been drafted to define the vouchers and the pilot project. That bill, as it stands, does not provide for refunds to the general fund.

6. Oregon's Legislative Council prepared three bills (LC 253-1 provides for MSAs in Oregon, LC 253-2 provides for refundable tax credits, and LC 253 provides for both) prior to the passage of the federal health insurance reform bill. That bill provides for a limited MSA demonstration. Appendix 1 compares the features of LC 253 with the provisions for MSAs in the federal bill and discusses current alternatives for Oregon.

7. Barchet's (1995) intensive study of 17 of Golden Rule's MSA programs resulted in some important qualitative conclusions as well as the numbers given in Endnote 12 below. These include "greater, not less use of preventive services, choice of MSA programs by both chronically sick and healthy people, likely choice of MSA programs by low income employees, and strong employee participation, acceptance and satisfaction" (pages 54 and 55).

8. See the discussion of community rating below in section D.

9. Many readers will ask: "What happens to families who cannot pay their share of the cost even if it is only 6 percent?" The Department of Human Services could lend them money. We estimate that 10,000 persons now on medicaid with incomes between 50 and 100 percent of FPL will choose not to be insured. Others might well ask, "What happens to people who have exhausted their MSA and are not yet covered by their high deductible?" The Department might lend them money, or, if they are improvident, deny them the privilege of using their voucher to buy a high-deductible policy.

10. Many people now learn about Medicaid when they go to an emergency room. The hospital enrolls them temporarily in fee-for-service Medicaid if their income is less than the FPL. The same could occur under means-tested vouchers. The hospital would simply tell patients about Medicaid's approved fee for that service and explain what part of the bill they will have to pay. The present system provides 6 months of Medicaid each time a person qualifies. Our proposal would rely on quarterly reporting of income, reset voucher levels quarterly, and use various methods to verify income.

11. Expansion of publicly-funded health care coverage logically causes a reduction in private and employer-provided insurance. From 1990 to 1994, employer-based health insurance declined from 60 to 52 percent in Oregon while government-provided coverage increased from 21 to 27 percent (OHPA, Dec. 1995, page 11). Cutler and Gruber ( 1995) studied the national expansion of Medicaid from 1987 to 92. They found that a 1.5 million increase in coverage of children led to a decline in the private coverage of 0.7 million children. "For women of childbearing age, there was a Medicaid increase of 0.8 million, but a reduction in private coverage of the same amount."

12. Barchet (1995) found a 40-percent savings with MSAs over the cost of traditional family coverage (page 10). That translates into about a 30-percent saving over managed-care policies. He also cites a KPMG Peat Marwick study of 1,037 firms that had MSAs, which found that projected net costs of MSA plans with deductibles of $2,000 for families were 55 percent of the costs of conventional MCO plans (page 11).

13. OHF, July 1995, p. 9.

14. Oregon pays MCOs about ten different rates for the different categorical groups, such as aged, blind, disabled, pregnant women, and children. Within the Medicaid expansion program, rates are differ for geographic reasons and through an adjustment for pregnancy.