Lessons From Tennessee's Failed Health Care Reform

April 7, 2000 About an hour read Download Report

Merrill Matthews

Resident Scholar, Institute for Policy Innovation

...

When President Bill Clinton first ran for office in 1992, he made health care reform a preeminent domestic policy issue. At the time, there were about 37 million uninsured Americans, and many of those who had health insurance lived in fear of losing their coverage, especially if they changed jobs. One year later, the President submitted a complex proposal that he said would fix the problems of the system.

Congress and the American people soundly rejected the highly regulatory and expensive Clinton health plan in 1994. Despite the weaknesses of that plan, however, several states adopted variations of the proposal in hopes of achieving universal health coverage or at least making health insurance more accessible and affordable for their citizens. Among the most ambitious of these plans at the time was Tennessee's TennCare program, which state officials proudly described as a less ambitious version of the Clinton plan.

The drive to implement universal health insurance reforms began to wane shortly after the President's plan was defeated in Congress. State legislators began to pass limited and incremental, rather than comprehensive and sweeping, health reforms. Now some legislators in states that enacted Clinton-style plans are expressing a desire to reconsider these programs.¹ The reason: Instead of reducing the number of uninsured, the state-level "reform" efforts have increased costs, and these cost increases have contributed to rising numbers of uninsured.²

Because a track record on the Clinton-style health reform efforts at the state level is available, policymakers around the country have a body of data from which to learn the results they have achieved. The results thus far can teach them very specific lessons about what to do and not to do to reform their health care systems. Before legislators in other states pass comprehensive health care reforms aimed at providing universal coverage, they would do well to look at the struggles Tennessee faces and the problems its legislators created.

Among the key lessons of TennCare:

Don't rush to "do something" for health care without considering the inherent limits of state policymaking and the damaging impact of higher health care costs, which would result from their changes, on individuals and families and on the numbers of uninsured.
Recognize that managed care is not some sort of panacea for saving money and that one cannot cut reimbursement of doctors or hospitals without compromising the quality of care.
Acknowledge that "comprehensive" health benefit packages will not also be "affordable." In any case, if the legislature wants to provide government assistance, efforts should focus on those who most need the help.
Realize that the only way to achieve high-quality care at a lower cost is to assure consumer choice so that patients will have a reason to become prudent health care shoppers.

State legislators should understand that the major problems of health insurance markets stem not from inadequate government regulation, but rather from an outdated set of federal and state tax policies that grew out of social and economic conditions in the 1940s and 1950s. These policies provide unlimited tax relief for persons who get their health insurance through the workplace but deny equal treatment to those who wish to buy health insurance on their own.

Not only are these outdated tax policies inequitable and unfair, but they also profoundly distort the health insurance market. They limit Americans' health insurance options and thus frustrate consumer choice of plans and benefits. They also undermine the opportunity for millions of Americans to find alternative forms of health insurance coverage, such as coverage obtained through their associations, trade and professional groups, and religious and fraternal organizations.

TennCare can teach valuable lessons, including the recognition that the program should be radically restructured to make it a safety net for low-income people who qualify for Medicaid and those who, because of a medical condition, cannot obtain health insurance on their own. Moreover, it shows that, rather than adapting portions of President Clinton's ill-conceived health plan to fit the unique conditions in each state's health insurance market, state legislators should urge their federal representatives to make serious changes in the tax code. Implementing reforms that enable a competitive market to develop and flourish--one in which individuals and families enjoy real consumer choice and insurance companies compete directly for their dollars--is the best approach.

TENNESSEE'S ATTEMPT TO EXPAND HEALTH CARE

In 1993, Tennessee quickly passed legislation to create the TennCare program. The speed with which the legislation flew through the assembly was unprecedented. Only a few months later, Governor Ned McWherter, a Democrat, received the necessary federal Medicaid waivers from the Clinton Administration to put the plan, which he had proposed, into effect beginning on January 1, 1994.

TennCare did not go through a normal legislative process to give proponents and opponents, special interests, and the public an opportunity to comment on it and allow the legislature to make needed changes based on that information,³ and this rush to "do something" was unfortunate. Legislators had hoped that, by enacting TennCare, the state could offer one health insurance safety net to cover all poor or uninsured residents at a lower cost than the state would spend on Medicaid, the federal-state program providing health care to the poor.

Governor McWherter and supporters of TennCare in the legislature--like President Clinton and congressional supporters of his plan--had argued persuasively that the state's current patchwork health insurance system was expensive and inefficient. The key to improving the problems was to shift everyone into managed care and provide close state-government oversight. The seductively simple theory was that, once the state controlled the money and the care provided, existing inefficiencies would disappear. Everyone would be covered and the state would save money.

In order to implement the plan, Tennessee applied for and received a five-year federal waiver from the Health Care Financing Administration (HCFA), the federal agency that runs the Medicare and Medicaid programs. This Section 1115 waiver permitted the state to leave the Medicaid program but use that money to fund TennCare. In return for the waiver, the federal government required the state to cover all Tennesseans who qualified for Medicaid, plus the uninsured and the uninsurable--or those who could not obtain coverage because of a pre-existing condition.⁴ Initially, TennCare covered 1.1 million people--the 766,000 residents then enrolled in Medicaid and an additional 340,000 who were uninsured or uninsurable. Today, it covers 1.3 million people--about the same number of Medicaid-qualified residents plus 550,000 others.⁵ (See Table 1.)

The federal waiver from HCFA set the stage for TennCare to become a model for state health care reform across the country. The Clinton Administration and other states watched with hopeful anticipation to see whether the "universal vision" of this highly regulated health care system would work.

TennCare's Track Record

One of the reasons the Clinton health care proposal failed is that the Administration and its congressional supporters had not promoted the provision of basic coverage--the "bare bones" benefits that would assure an adequate level of care. Such policies cover major medical expenses but not all the options that would make a plan more comprehensive as well as more expensive. As a result, health insurance is more affordable and thus more accessible, especially for middle-income families.

Instead, supporters of the Clinton plan pushed for a "Cadillac" type of plan that included many government-mandated "options" (which, because they were mandated, ere not really options at all).⁶ TennCare's sponsors wanted the plan to be comprehensive as to both who and what it covered.

Eligibility.
Eligibility criteria have changed since TennCare was passed. Currently, the program is available to:

People who are eligible for Medicaid;
Uninsured children under the age of 19 who do not have access to health insurance and children with access to health insurance whose family incomes fall below 200 percent of the poverty level;
Workers who lost coverage due to the closing of their employer or the expiration of their benefits under the Consolidated Omnibus Budget Reconciliation Act (COBRA) of 1985, and those with limited coverage (with no income restrictions);
The "uninsurables," or those who have been denied health insurance coverage because of a medical condition (with no income restrictions); and
Uninsured people with incomes below 200 percent of poverty (a group that has been closed to new entrants since 1995), although some people such as displaced workers can still enter as uninsured persons.

Managed Care Coverage.
Although most states are beginning to move their Medicaid populations into managed care, Tennessee is the only state to have done so with all of its Medicaid recipients.⁷ TennCare beneficiaries who do not qualify for Medicaid are also enrolled in managed care. The state contracts with managed care organizations, provides a capitated payment system for each patient, and expects physicians and hospitals in the program to provide comprehensive, high-quality health care.

TennCare covers inpatient and outpatient services, physician services, prescription drugs, and medical supplies, as well as laboratory tests and x-rays. It covers home health care, hospice care, and ambulance charges. In addition to physical health care benefits, TennCare contracts with behavioral health organizations to provide comprehensive mental health benefits. Out-of-pocket expenses are limited for low-income participants. Others, such as the uninsured, the "uninsurables" who are unable to purchase health insurance because of a pre-existing condition, and those who have lost coverage, pay premiums based on a sliding scale.

Problems for Providers.
Currently, nine managed care organizations participate in the program, but this will soon change. Blue Cross Blue Shield of Tennessee, which covers half of TennCare's recipients, recently announced that it will drop out of TennCare next summer.⁸ In addition, Xantus HealthPlan of Tennessee, Inc., recently went into receivership, and other TennCare managed care organizations claim that they cannot continue to provide services for the low reimbursement rates they receive in the program, even after the legislature provided an additional $190 million to help make up the shortfall.⁹ As a result, it is likely that far fewer managed care plans will be participating in the future.

UNINTENDED CONSEQUENCES

Although TennCare was designed specifically to control costs and expand coverage, the program faces enormous problems that prevent it from improving health care for many Tennesseans.

PROBLEM #1: Rising Subsidies.
Federal subsidies are addictive. It is almost impossible for lower levels of government to forgo them after relying on them in the past--a fact well demonstrated by TennCare. The reason: The federal government matches the state's contribution to the Medicaid program, but in most cases the federal share exceeds that of the state. This disproportionate match induces state and local governments to spend more than they otherwise would, which certainly helps states that are struggling to expand Medicaid coverage. It also, however, makes it very difficult to repeal those benefits in the future.

For example, for every Medicaid dollar Tennessee spends, the state puts up about 37 cents. The federal government pays the other 63 cents. This means that for the state of Tennessee to cut Medicaid costs by one dollar, it must be willing to cut three dollars from the program's funds--an almost impossible task when doctors and managed care organizations clamor that the program is underpaying them.

The perverse incentive that this creates, both politically and financially, is to throw more state money at Medicaid because the state gets so much bang for its buck. However, if the state is not careful, it will find its budget increasingly dominated by Medicaid to the detriment of other programs and the taxpayers.

PROBLEM #2: Higher Costs and Higher Taxes.
At a time when other states are experiencing record budget surpluses, the Tennessee legislature recently held a special session to consider imposing a state income tax to offset the budget shortfall created by TennCare. Governor Don Sundquist, a Republican and one of the income tax's strongest backers, proposed lowering the state sales tax to offset some of the increase in the income tax. Nonetheless, the purpose would still be to obtain more revenue for the state, not to break even.

The state needs the increase in funds because TennCare is so expensive--it consumes $4.3 billion in the current budget--and costs are growing rapidly. According to a new study of Tennessee's economy, health care expenditures increased 69 percent between 1992 and 1999, while personal income increased only 38 percent.¹⁰

Supporters argue that TennCare, which spends about $3,300 per person per year, has saved Tennessee money over what the state would have spent on Medicaid.¹¹ But they are comparing the program's costs with a projected Medicaid growth rate that is too high. When TennCare was created, proponents compared it with Medicaid expenditures in 1992--one of its more expensive years because Congress was expanding the program--to project future Medicaid expenses.¹² Such comparisons should use projections for at least two years; even more than that would provide a significantly more accurate picture.

Moreover, as shown in Table 2, the growth in Medicaid spending nationwide moderated in the mid-1990s, in part because of the strong economy and the success of welfare reform.¹³ Tennessee was unable to enjoy even this budgetary reprieve because TennCare was already draining excess revenue from the state budget. Had the state addressed the Medicaid problem differently in 1993, Tennessee lawmakers, like Members of Congress and so many other legislators in other states, might today be debating what to do with a budget surplus rather than a budget deficit.

There are several reasons why TennCare costs the state so much money. A primary reason is its ambitious nature. Currently, 24 percent of the state's population is enrolled in the program. TennCare's creators intended it to be ambitious and designed it to be more aggressive than other states' programs in covering the poor and uninsured. They succeeded, making the program uncontrollably large compared with those of most other states. For example, in 1997:

Only the seven states with significantly larger populations have surpassed Tennessee in the number of people they covered in their state-run programs.
Nearby states that are comparable in size to Tennessee, such as Kentucky, Missouri, Mississippi, and Alabama, have about one-third to one-half the number of Medicaid participants enrolled in their programs.

TennCare enrolls a higher percentage of adults than children; and adults, on average, require more health care. Whereas in most states, children make up a little more than half of the people on Medicaid,¹⁴ in TennCare, children represent about 41 percent of enrollees.

Another contributing factor is that TennCare provides a very generous benefits package that is better in some respects than the packages some private-sector employers offer their employees. For example, although many employers do not offer drug coverage, 22 percent of TennCare's budget covers its prescription drug benefits.¹⁵

In addition, TennCare includes the state's uninsurables with pre-existing medical conditions. Middle- and upper-income uninsurables pay monthly premiums to TennCare, but this type of program never raises enough in revenue to cover costs, and therefore loses money. TennCare's uninsurable population is disproportionately large, and this imposes a huge financial pressure on the system.

Finally, the ease with which sick people can enter the system encourages people to wait until they have a medical condition before they obtain coverage. Uninsurables need only prove that they have been turned down for private insurance coverage before they enter the program.¹⁶ Given such ease of entry, it is curious that healthy Tennesseans are still paying for private health insurance coverage.

There are other factors the state cannot control. Although health care spending moderated in the mid-1990s, every indication is that higher health care cost increases will become the norm for both the public and private sectors in the future.¹⁷ Several factors will cause this increase.

The growing use of prescription drugs is a trend that will hit Medicaid populations especially hard since the poor, on average, have more health problems;
There are new medical technologies that permit hospitals and physicians to do more than they could in the past;
New federal and state mandates will require Medicaid to cover products and services, such as Viagra, that it might not otherwise cover;
New restrictions on managed care plans that provide services to many Medicaid recipients nationwide and all of TennCare's population are driving up the cost of care; and
The gradual aging of the population will mean that more people, especially the poor, will face more health problems in the future.

The combination of such factors means that Americans can expect health care to be more expensive for the population as a whole, but particularly for Medicaid recipients.

PROBLEM #3: Declining Quality of Care.
Since its inception, TennCare has faced quality problems.¹⁸ The reason: Chronically low reimbursement rates for doctors and hospitals have led to rationed care, which means less care in most cases.

A March 1999 actuarial review by PricewaterhouseCoopers found that managed care organizations reimbursed providers at a rate of about $11 per member per month (about 10 percent) below what would be considered an "actuarially sound" level. The report estimated that the managed care organizations needed to institute a capitation rate increase of 5 percent to 35 percent, with a 20 percent increase as its best-guess estimate.¹⁹ At the end of the 1999 session, the state legislature followed the recommendation to increase the capitation rate, providing an additional $190 million in state and federal matching funds--most of which is required to go to providers. This increase brought TennCare's current budget to $4.3 billion.²⁰

Financial cutbacks or low reimbursement rates usually do not produce an immediate drop in the quality of care. Doctors and hospitals look for alternatives first. They may absorb the losses for a while. They may also cost-shift money from private-sector or Medicare patients to help make up the difference. Over time, however, prohibitively low reimbursement rates drive the best doctors out of a program or force the increasingly cynical physicians who remain to find other ways to be compensated for their services.

Patients may not recognize the decline in quality immediately. Hospitals and physicians may not even recognize it themselves at first. But eventually, they come to realize that choices between money and patient care are being made.²¹ Add to that scenario the standard managed care practice of allowing physicians to limit the amount of care they provide and the result is a prescription for rationing care.

PROBLEM #4: Fraud and Abuse.
Between 1988 and 1993, federal Medicaid spending grew at an average annual rate of 19.6 percent. The rate of increase has slowed, but Medicaid is still expected to grow by an average annual rate of 7.9 percent from 1999 to 2004.²²

The program also has been plagued with fraud. Even before managed care was adopted, many states (especially those with large inner-city populations) experienced significant problems with what came to be known as "Medicaid mills"--doctors who routinely saw 60 or 70 patients a day, most of whom only wanted a prescription for drugs subsidized by Medicaid.

The prevalence of these two factors--explosive growth and abuse of the system--led states to begin placing a significant portion of their Medicaid population in managed care. Tennessee was the only one to place all of its Medicaid beneficiaries in managed care.

While there is some reason to believe that managed care can reduce the amount of fraud that occurs when beneficiaries get unneeded medical products and services, or when providers prescribe goods or services that patients do not need, other types of fraud are likely. TennCare's rich benefits package and ease of entry make the program a magnet for abuse. According to news accounts:²³

TennCare spent $6 million covering 14,000 dead enrollees;
TennCare covered 16,500 enrollees who lived out of state;
An analysis of 98,000 enrollees found that 20 percent were ineligible to be in the program; and
450 of those ineligible were state employees who had access to the state employees' health insurance plan.

The legislature may be able to implement reforms to address these types of fraud, but incremental steps will not solve TennCare's growing financial problems. These problems are systemic and will continue to invite abuse. The system needs fundamental reform.

TEN LESSONS FROM TENNCARE

Legislators across the country can learn from TennCare's track record and avoid making the mistakes Tennessee's legislators made in their attempt to improve health care. The following ten lessons are key to enacting good reform.

Lesson #1: The states have limited options for fixing the health care system.

Tennessee officials hoped to do at the state level what President Clinton was unable to do at the federal level.²⁴ They were not successful.

TennCare's experience demonstrates that, under the current system, universal coverage at the state level--even just coverage for those who cannot or will not obtain coverage through their employer or government program--is economically impossible unless the state is willing to commit a huge amount of its limited resources to the program. Such a commitment may be politically impossible because there is limited willingness among taxpayers to pay for such programs, and Tennessee may have reached the limit. But there are additional reasons that states are limited in what they can do--reasons that are deeply rooted in the very structure of the health insurance market and are a product of federal, not state, policy.

The Federal Tax Impediment.
The current health insurance system is largely a result of federal tax policy. People who obtain health insurance through an employer receive a tax exclusion for that benefit; that is, they do not record the money their employer spends on insurance as income and thus pay no taxes on it. The self-employed get a partial tax deduction, but those who work for employers who do not provide health insurance receive no tax break on the money they spend to obtain health coverage.²⁵ Since states are unable independently to change federal tax law, it is almost impossible for state legislatures to alter the current tax incentives. Moreover, whatever reforms they make must be compatible with federal tax policy.

The Employee Retirement Income Security Act of 1974.
Many states are targeting the business community for health insurance reform. They reason that if they could require businesses to provide employees with health insurance, they could pass on to the private sector the expense of and responsibility for creating universal coverage.²⁶ However, the Employee Retirement Income Security Act (ERISA) places the health insurance programs of employers who self-insure (i.e., who pay the health insurance bills instead of paying an insurance company) under federal labor law rather than under state insurance laws. As a result, states have virtually no control over the health insurance plans of self-insured companies.

For example, state legislatures that want to require employers and insurers to cover treatment for drug and alcohol abuse (a benefit "mandate") can impose the law only on companies and people buying from health insurers operating within the state. Such purchasers usually are small companies and individuals buying their own policies. Self-insured employers--whose employees make up about half of the workforce--need not comply. Thus, this process puts smaller businesses at a competitive disadvantage against larger companies, which can pick and choose what to cover or choose very few options in order to keep the cost of policies down.²⁷

The inability to impose health insurance laws on all insurance in the state leaves a huge escape hatch for businesses. Employers dissatisfied with state regulations may look for a way to self-insure or may choose not to provide any health insurance coverage. States have been challenging the ERISA preemption in courts, but they have experienced little success so far.

If states want to expand coverage, they must first realize that their options to enact fundamental change are limited. They have little ability to do good and a lot of opportunity to do harm, as TennCare demonstrates.

Lesson #2: Don't rush to create a radically new program.

"Haste makes waste," and in TennCare's case, the haste with which the system was created led to significant waste. A state's health care system, including its health insurance markets and the interaction of those private markets with public health care programs, is complex. Bad policy in such a complex environment predictably will produce even worse results. Badly designed policies, even if enacted with the best of intentions, often have disastrous consequences. If there is one area in which state legislatures should tread carefully, it is health care policy.

Tennessee, however, did not do this. Governor McWherter proposed TennCare on April 8, 1993. By May 5, the legislature had approved the outline and authorized the governor to proceed. By June, the state submitted a request to HCFA for a waiver that would permit TennCare to move forward. The U.S. Secretary of Health and Human Services had reservations about some of the provisions and met with the governor in the fall to discuss them. By November, HCFA had granted the waiver, which allowed TennCare to begin operations on January 1, 1994.²⁸

Thus, only about eight months had transpired from TennCare's inception to its implementation, with minimal input from the elected representatives of the taxpayers of Tennessee. The political objective appears to have been to get a complex regulatory structure in place and then to allow the "experts" to work with a minimum of interference from the legislature.

Of course, America's Founding Fathers inspired a much different approach to the legislative process. As evident from even a casual reading of The Federalist Papers, James Madison, John Jay, and Alexander Hamilton believed that an elected legislature--an essential feature of a republican form of government--should govern through careful deliberation. The Founders intentionally created a legislative system that is slow and cumbersome in order to minimize the chance that damaging legislation would be rushed through to enactment. Slowly moving legislation, especially on weighty matters, gives the people a chance to learn about it, analyze it, and determine whether any aspects of it could cause harm or work poorly, which would enable legislators to change or improve the proposal.

TennCare bypassed this slow and deliberate legislative pace. Much of the approval process took place after the legislature had been dismissed for the year and under the shadow of the Clinton health care plan, which created momentum for doing something quickly. The governor of Tennessee capitalized on that momentum, but instead of becoming a solution to a real and difficult problem, TennCare became part of the problem.

There is a growing desire in some states to reevaluate the possibility of creating a system of universal coverage based on the Clinton model, but state legislatures should move slowly and cautiously, giving ample opportunity for vetting the components of any proposal for such reform. Otherwise, they will have to revisit the enabling legislation, time and again, in an effort to reform the "reforms" that fail to work.

Lesson #3: Include medical professionals and other stakeholders in developing the reform.

Reforming something as extensive and complex as the health insurance system requires input from all stakeholders. Insurers, the medical community, elected representatives and other government officials, and especially the taxpaying public should have a say--or at least an opportunity to raise issues and concerns--about the direction or elements of reform.

This did not occur in Tennessee. The system is based largely on the proposal of Governor Ned McWherter and his political allies. The professional medical community and the private insurers had little input.

Politicians often remark that by excluding special interests--which usually means the people a proposed piece of legislation will affect the most--they can create public policy that will work because it has not been corrupted by those whose profits would be affected by the legislation. However, those same special interests are often the ones that have the deepest understanding of how the system operates and what will and will not work.

Tennessee's politicians are not alone in pursuing this unproductive politics of exclusion. For example:²⁹

Kentucky decided to push through health insurance legislation in 1994 and ignore the comments of the insurers who would have to operate under the system. Within just a few years, 45 of the state's insurers dropped out of the system, leaving only Anthem Blue Cross and KentuckyKare.
New Jersey decided to reform its individual health insurance market (policies bought by individuals and self-employed people) by making policies more accessible and affordable. As a result, the state imposed a "guaranteed issue" requirement so that people could obtain a health insurance policy regardless of their health status, in addition to a "community rating" requirement that charges everyone the same price, again regardless of health status. Because healthy people can go other places and get cheaper premiums, guaranteed issue creates a system in which only sick people are in the pool and premiums are very expensive. Currently, a person in New Jersey purchasing an individual policy with a $500 deductible and a 20 percent co-payment will pay between $25,000 and $30,000 a year for most of the policies available.

In both Kentucky and New Jersey, insurers and other "special interests" tried to warn state legislators that their plans would not work. Tennessee likewise excluded the experts, and now Tennessee legislators are being forced to consider how to reform TennCare before it breaks the state budget. This time, they should include stakeholders in their deliberations on the substance, not just the outline, of this complex program.

Lesson #4: Don't overprice the plan through a comprehensive benefits package.

Besides their desire to enact reform legislation quickly, ignoring the special interests who know what works and what does not, state lawmakers who push massive reform typically want to provide a comprehensive package of benefits. Like the reasons put forth for guaranteed issue and community rating requirements, the rationale advanced by proponents of comprehensive benefits assumes that insurers, employers, and health plans will limit what they cover because they "put profits above patients." The state-level reformers want to prove that they can create a health insurance package that is easy to obtain, comprehensive, and affordable for a lot less than the system provided by profit-motivated employers and insurers.

This approach has not worked. Insurers know that if the government creates a subsidized health insurance policy with a richer package of benefits than most people obtain from their employers, then more people will try to drop their employer-provided coverage and join the public system. This predictable process is called the "crowding-out effect."

This is precisely what has happened in TennCare. Tennessee made TennCare a very attractive insurance policy, including, for example, prescription drug coverage. When the plan was open to anyone who was uninsured, people who could purchase policies in the private sector flocked to the program, forcing TennCare to limit access to the uninsurable population.

Besides the impression that insurers and employers limit benefits to save money, there is another myth guiding policymakers: that many of the people who have health insurance are underinsured. That is, the policy may not cover all medical services, experimental procedures, preventive care, and prescription drugs.³⁰

But the truth is that most people are overinsured. That is, they have health insurance coverage for medical products and services, such as preventive care, that they could easily pay for out of pocket. The "underinsured mentality" leads lawmakers to try to impose on insurance much more comprehensive--and therefore expensive--policies than most people need. Comprehensive health insurance packages can quickly break a state's budget, as TennCare has demonstrated. Not even Medicare is comprehensive. It leaves a number of gaps, which is why the vast majority of seniors obtain some type of supplemental coverage.

It is understandable that lawmakers would want to provide comprehensive coverage, especially to the poor, but that desire must be balanced against the limited funds available and the problem of luring people to the program who could obtain coverage in the private sector. Tennessee ignored that balance and is now paying the price; other states should learn from that mistake.

Lesson #5: Managed care is not a panacea.

Governor McWherter and other supporters of TennCare sold the program to the public in part by claiming that managed care would save the system so much money that the state would be able to cover not only the Medicaid population, but the uninsured and uninsurables as well.

Studies have shown that managed care can indeed save states money, especially in a program as chronically inefficient as Medicaid, but those savings are limited. In both the public and private sectors, there appears to be a one-time savings of between 2.3 percent and 9.6 percent when shifting to managed care. After that, however, costs begin to grow at relatively the same rate as under traditional insurance.³¹ Although many states have managed to squeeze additional savings out of Medicaid managed care, they have been able to do so in large part because they are arbitrarily cutting reimbursement rates below providers' costs.

Managed care may be able to hold down some costs, but there is a limit. Below that limit, cuts will begin to threaten the availability of services and the quality of care. TennCare has demonstrated that while states can use managed care to provide health care for less money, proponents of universal coverage are too optimistic in thinking that managed care is the panacea that will let them obtain "universal" coverage with no additional cost.

Lesson #6: Keep reimbursement rates adequate.

TennCare was meant to be more than managed care; it was meant to be managed competition. Like the Clinton health care plan, it would have the state offer a flat amount of money for each enrollee, and managed care companies would compete against each other for customers, keeping prices low and quality high. Once again, state lawmakers failed to understand just how insurance markets work.

As with many managed care arrangements, TennCare imposes a capitation rate, meaning that providers receive a predetermined amount of money for each person enrolled. Currently, managed care organizations receive an average of $132 per enrollee per month, while behavioral health organizations receive about three times as much per enrollee.³² Those patients who, on average, use less than $132 a month in health care services are profitable. Those who use more than $132 per month cost providers money.

Thus, providers have an incentive to attract patients who are healthy rather than sick. Even though TennCare and other such plans impose restrictions prohibiting what is known as "cherry picking," it is not unusual or hard for health plans to find a way around the restrictions.

As a so-called public-private partnership, TennCare would try to foster competition by providing enrollees with a choice of plans. However, reimbursement rates have been kept so low that no new managed care organizations have joined, several have dropped out, and Blue Cross Blue Shield, which covers about half of the TennCare population today, has said that it intends to exit the program in June 2000.³³

This is a health care disaster. A fundamental rule of the marketplace is that price controls cannot foster competition. This is especially true if the government sets the prices too low. Companies will not compete to win a money-losing contract, which is precisely what TennCare has become. It is probably fair to say that the only reason some of the managed care organizations have remained as long as they have is a sense of obligation to provide health care. But even that obligation has its limits, especially if capitation rates become too low.

The obvious solution for Tennessee would be to raise reimbursement rates, as PricewaterhouseCoopers and others have suggested, and the state recently complied. But there are reasons why politicians may want to avoid the obvious solution. The state determines what the capitation rate will be, and that decision can be guided more by politics than by good patient care.

Part of the political problem is that there are other pressing claims on the state's limited resources. For example, if health care providers are in desperate need of a reimbursement increase but teachers clamor more loudly, then any additional state money available may go to pay teachers rather than doctors and hospitals.

But the bigger problem is saving political face. TennCare was promoted as a way to cover more people and save money. If politicians raise reimbursement rates above a predetermined level, they become targets of critics who say they lied when the program was sold to the state. Politically, it is much easier to keep reimbursement rates low, say they are adequate, and claim that complaints emanate only from greedy doctors and health plans.

In November 1999, the Tennessee legislature convened a special session to consider adopting an income tax in addition to the state's sales tax so that the state could meet its financial obligations, created largely by TennCare's drain on its resources. The legislature failed to do anything but study the problem yet again. It may be that Tennesseans have reached their limit on how much they are willing to spend to help middle-class people buy government-run health insurance.

On the subject of uninsurance, taxpayers should be treated to a healthful dose of honesty. Other states are considering ways to cover all of their uninsured. State officials should decide up front whether their legislature is willing to ask taxpayers to pay for the insurance they want to provide. Unless there is an honest and truthful discussion about the true cost of a program, states are likely to promote reform with artificially low cost projections. Years--and millions of dollars in debt--later, proponents will try to avoid explaining how they could have misjudged the costs so badly while they look desperately for new ways to meet the shortfall.

Lesson #7: Create a real high-risk pool for the uninsurable.

One of TennCare's safety-net provisions is its coverage of the state's uninsurable population--those who are unable to get a health insurance policy because of a pre-existing medical condition. This part of the program is very costly and inefficient, and reforming this high-risk portion of TennCare would lower program costs significantly. Fortunately, there are other state models that show how to establish an effective high-risk pool.

Currently, 27 states have implemented some form of high-risk pool to address the needs of their uninsurables.³⁴ Although these programs vary significantly, most states have created an insurance safety net for this group of people. Typically, people qualify by proving that they have been denied a health insurance policy because of a pre-existing medical condition. If accepted into the high-risk pool, they are able to purchase a standard health insurance policy, but usually at a rate of between 25 percent and 50 percent above the cost of a standard policy.

There are two reasons for the additional premium. First, uninsurable people cost more to cover. Second, and perhaps more important, states want to discourage people from waiting until they get sick to get an insurance policy. A 50 percent increase appears to be a fairly reasonable deterrent.³⁵

However, TennCare charges about 22 percent more for the uninsurables who have incomes at or above 400 percent of poverty. This means that the state loses money, even on those who could easily afford higher premiums. For example, a qualified family falling in the uninsured category that makes between 400 percent and 749 percent of poverty pays about $489 per month in premiums. An uninsurable family pays about $595 for the same coverage. Using TennCare's rate for the uninsured as the standard, charging the uninsurables 50 percent more than the standard rate would mean that a family would pay about $733 a month.³⁶

Is that unaffordable? Remember, these are families with incomes between 400 percent and 749 percent of poverty--or between $62,400 and $116,800 a year for a family of four. While $733 a month is not inexpensive, it also is not unreasonable, especially for those in the upper-income category who already have medical conditions that may cost the plan thousands of dollars in medical expenses.

Even if the state raised the premium cost to 50 percent more than the standard premium, the state would still need to subsidize the program, but the amount of money needed would be much less than is required under the current system. Although other states with high-risk pools also lose money, those losses are manageable. (See Table 3.)

Another problem is that TennCare, oddly enough, seems to favor the rich. Under current rates, an uninsurable family making 750 percent of poverty or above pays only $18 more a month than those in the 400 percent to 749 percent category. This is a very small increase, given the huge income differences.

Pricing high-risk insurance too low and making TennCare very easy to enter has created an unsustainable program that encourages people to remain uninsured until they get sick and enter TennCare after being denied coverage by a private insurer. Insurers are all too willing to deny coverage, since they do not want to pick up the expenses of a high-risk patient.

As a result, about 114,000 people are classified as uninsurable in Tennessee. The uninsurable in all other 27 states with high-risk pools combined amount to slightly more than 100,000 people.³⁷ As Figure 1 shows:

California, the most populated state in the country, only has 21,429 people in its high-risk pool.
States with populations close to Tennessee's have drastically fewer people in their high-risk pools.

Legislators should keep in mind that the uninsurables are not necessarily the same as the poor. For the uninsurables, the primary goal should be to make health insurance accessible, not affordable. While charging 50 percent above the standard premium can make a policy expensive, charging too little is an open invitation to people to wait until they get sick before obtaining insurance.³⁸

Lesson #8: Means test the public health care program.

Most states across the country, including Tennessee, have seen significant declines in their welfare caseloads. Tennessee reached its caseload peak of 112,597 in November 1993, just a few months before TennCare's initiation.³⁹ By the middle of 1999, the caseload had dropped by nearly half to 57,710. (See Figure 2.) This decline has left the state with a surplus of $91.5 million in funds for the Temporary Assistance for Needy Families (TANF) program.⁴⁰

For many states, the drastic decline in welfare caseloads also reduced the number of people on Medicaid.⁴¹ Between 1995 and 1997, according to a September 10, 1999, analysis by the U.S. General Accounting Office, welfare caseloads declined nationwide by 23 percent, while Medicaid declined by 7 percent. Of the 21 states sampled (Tennessee was not one of them), Wisconsin recorded the largest Medicaid caseload decline of 19 percent, while Delaware experienced a 26 percent increase.⁴² States with welfare caseload declines comparable to Tennessee's appear, on average, to have had 10 percent to 15 percent declines in their Medicaid caseloads.

An examination of Tennessee's Medicaid population for approximately the same time period shows that the state experienced a 2 percent to 3 percent increase. Tennessee Medicaid enrollment reached its peak in 1996. Calculating the caseload decline from 1996 to 1999 shows that the state experienced a 9 percent decline, but this is still probably 10 percent to 15 percent below where the state should have been, considering that the most rapid decline in welfare occurred in 1997.

One of the explanations for why Medicaid caseloads have not declined as rapidly as welfare caseloads is that many states permitted those moving from welfare to work to remain on Medicaid. This policy helped address the problem of welfare recipients leaving welfare, which provided coverage, to take a job that did not offer health insurance. However, even if some of these families are permitted to stay on Medicaid for a limited period of time after joining the workforce, Tennessee's 50 percent decline in caseloads should have produced a larger drop in its Medicaid population.

Although states want to ensure a smooth transition from welfare to work, they should make sure that those who have the option of employer-provided health insurance do not continue on the Medicaid rolls.

Lesson #9: Let federal legislation handle the "portability" problem.

At the time TennCare was created, many Americans faced the loss of their employer-based health insurance when they changed jobs. One of TennCare's goals was to stabilize this portability problem by giving the uninsured--some of whom had lost their coverage in a job transition--the ability to get health insurance from the state.

However, in 1996, Congress passed the Health Insurance Portability and Accountability Act (HIPAA) to enhance portability. Prior to enactment of this legislation, insured workers who worked for companies with 20 or more employees had been permitted under federal law (the Consolidated Omnibus Budget Reconciliation Act, or COBRA) to continue obtaining that coverage for up to 18 months after leaving that job. Insured workers who were employed by small employers (those with under 20 employees) did not have that option. Under HIPAA, workers who have been insured continuously and who work for an employer not covered by COBRA may purchase a policy in the individual market immediately after leaving a job without having to face a pre-existing condition waiting period. Those with COBRA benefits may purchase a policy in the individual market after COBRA benefits have been exhausted.

In other words, there is now a federal law addressing a part of the uninsured problem TennCare sought to address. Insured workers in job transition and needing insurance should simply exercise their options under HIPAA. A few people may fall through the cracks (for example, a worker who was insured by an employer but not long enough to qualify for HIPAA), but broad-based reform to create insurance portability in TennCare is no longer a pressing problem.

Although HIPAA has helped in some areas, it has caused a lot of problems in others. Even its attempt to solve the portability problem is only a second-best solution. The best answer to the portability problem is a change in the federal tax laws so that individuals and families can obtain tax relief for the purchase of health insurance regardless of their place of employment. This would facilitate the personal ownership of health plans and policies, which is the best way to create true portability.

Lesson #10: Include real incentives to control utilization of medical services.

A goal of former Governor McWherter in creating TennCare was to give enrollees incentives to lower their utilization of medical services. That goal remains unfulfilled. Were TennCare to shift its enrollees--or at least those with incomes 300 percent of poverty or higher--into a system of medical savings accounts (MSAs), that goal could be realized.

Medical savings accounts give people a good reason to be prudent health care shoppers: They will benefit financially.⁴³ With an MSA plan, people would take the same money they spend on a low-deductible health insurance policy and purchase a less expensive policy with a high deductible of, say, $3,000. Because the high-deductible policy costs less, they can put the premium savings they accrue in a personal MSA to use for preventive and routine care during the year. Money left over at year's end belongs to the individual. Thus, an enrollee would have a high-deductible policy to pay for care in case of a major accident or illness, but most health care expenses would be paid from the funds saved in the MSA.

One of the reasons why TennCare costs have exploded is that patients have no incentive to control their health care spending. Giving people an MSA is one of the best ways to achieve both goals of providing coverage and giving patients the proper incentive to obtain the best value for their health care dollar.

Another option is to use Medicaid monies to give low-income families vouchers they can use to pick and choose the kinds of plans they want in a genuinely competitive market in which plans and prices respond to the pressures of consumer choice and competition.⁴⁴

CONCLUSION

Since its inception as a model for health insurance reform at the state level, TennCare has encountered problems related to costs, fraud, and the quality of care. True, the state of Tennessee has managed to cover more persons with health insurance than it would have had it done nothing. But simply covering more people should not be the only goal.

State health care reform should create a program that effectively covers the poorest and those unable to purchase health insurance without either undermining coverage in the private sector or imposing huge costs in the public sector. In other words, a solid program of health reform should target those who need help the most. It should not undermine private-sector coverage for the middle class. It should create a safety net, not a hammock. By that standard, TennCare has failed miserably.

TennCare can and should be a model for lawmakers in other states, but not the sort of model its creators intended. Of the many lessons to be learned from TennCare's experience, lawmakers should gain a clear understanding of what not to do.

Merrill Matthews, Ph.D., is a health care policy analyst who specializes in federal and state health care policies. He is currently a visiting scholar at the Institute for Policy Innovation in Lewisville, Texas.

1. See, for example, Robert Cihak, M.D., Bob Williams, and Peter J. Ferrara, "The Rise and Repeal of the Washington State Health Plan: Lessons for America's State Legislators," Heritage Foundation State Backgrounder No. 1121/S, June 11, 1997; Rachel McCubbin, "The Kentucky Health Care Experiment: How Managed Competition Clamps Down on Choice and Competition," Heritage Foundation State Backgrounder No. 1119/S, June 6, 1997; and Dale Snyder, "Building Bureaucracy and Invading Patient Privacy: Maryland's Health Care Regulations," Heritage Foundation Backgrounder No. 1168, April 16, 1998.

2. On the relationship between aggressive state regulation of the health insurance market, price increases, and the rise in the uninsured, see Grace-Marie Arnett and Melinda L Schriver, "Uninsured Rates Rise Dramatically in States with Strictest Health Insurance Regulations," Heritage Foundation Backgrounder No. 1211, August 14, 1998.

3. For the most thorough analysis of TennCare and its legislative history, see James F. Blumstein and Frank A. Sloan, "Health Care Reform Through Medicaid Managed Care: Tennessee (TennCare) as a Case Study and a Paradigm," discussion draft (revised), August 1999, funded by the Robert Wood Johnson Foundation.

4. Indeed, HCFA required everyone to be in managed care. At that time, the Administration saw managed care as part of the solution to the health care "crisis," not the problem.

5. TennCare has an enrollment cap of 1.5 million people.

6. See Merrill Matthews, Jr., and John C. Goodman, "The Cost of Health Insurance Mandates," National Center for Policy Analysis, Brief Analysis No. 237, August 13, 1997.

7. John Holahan, Suresh Rangarajan and Matthew Schirmer, "Medicaid Managed Care Payment Methods and Capitation Rates: Results from a National Survey," Urban Institute, June 1999.

8. Blue Cross and Blue Shield, "Blue Cross Withdraws from TennCare Effective June 2000," press release, December 15, 1999. According to the release, Blue Cross told Governor Sundquist that "it could not accept the financial risk for fiscal year 2000 because the company's estimate of losses in the TennCare program were $51 million to $96 million." The state recently said that it will require Blue Cross and Blue Shield to remain in TennCare until the end of the year.

9. John Commins, "$190 Million Asked for TennCare," Times & Free Press, May 11, 1999.

10. Steve Adams, Tennessee State Treasurer, "Tennessee Tax Policy: What's at Stake?" presentation before the Senate Finance, Ways and Means Committee, April 6, 1999, pp. 4, 11.

11. Dividing the $4.3 billion annual budget by the 1.3 million people in the program yields $3,307 per person.

12. Sandra Hunt et al., "Actuarial Review of Capitation Rates in the TennCare Program," PricewaterhouseCoopers, March 1999, p. 17.

13. See, for example, "State Budget Surpluses and Maximum State Spending Under CHIP," Children's Defense Fund, March 15, 1998. The budget surpluses for many states have grown significantly since this analysis.

14. See "Medicaid: A Primer, An Introduction and Overview," Kaiser Commission on Medicaid Basics, August 1999, p. 7.

15. David Flessner, "TennCare Drug Coverage Cost Is a Growing Concern," Times & Free Press, March 14, 1999.

16. "Some companies including Blue Cross and Blue Shield of Tennessee, the state's largest health insurer and a major player in TennCare, actually help people to qualify for TennCare by selling them `denial letters' to prove they've been denied coverage." Bill Snyder and Dorren Klausnitzer, "Increase in `Uninsurables' Clogs TennCare," The Tennessean, March 8, 1999.

17. For example, CalPERS, the California pension and benefits fund that covers about 1 million state employees and their dependents, recently announced a 9.7 percent increase for health insurers in 2000. "CalPERS: Announces 9.7% Increase for Insurers," American Health Line, May 20, 1999. See also Julie Appleby, "Health Care Premiums on Rise Again," USA Today, May 17, 1999.

18. For an early analysis of TennCare, see Terree P. Wasley, "TennCare: Health Care Reform Dream or Disappointment?" Heritage Foundation State Backgrounder No. 1021/S, February 28, 1995.

19. Hunt et al., "Actuarial Review of Capitation Rates in the TennCare Program," p. iv.

20. See David Kushma, "TennCare's Ills Can Be Cured, Director Says," The Commercial Appeal, September 26, 1999.

21. Keith Snider, "Hospitals Lost Millions on TennCare," The Tennessean, August 31, 1999.

22. "Medicaid: A Primer," p. 5.

23. See Bonnie M. de la Cruz, "Audit: TennCare Paid $6 Million to Insure Dead People," The Tennessean, July 9, 1999, and Paula Wade, "TennCare Finds 16,500 Ineligible," The Commercial Appeal, August 11, 1999.

24. Indeed, discussion surrounding state-level health care reform in 1993 and 1994 generally assumed that state reform programs eventually either would merge with or be superseded by the Clinton plan once it passed Congress. As a result, reformers often perceived their plans as short-lived.

25. There is significant support in Congress for redressing these discrepancies. Congress already has passed legislation that lets the self-employed deduction rise until it reaches 100 percent, and there is support for giving the employed but uninsured either a tax deduction or a tax credit. Indeed, Congress included that legislation in a $792 billion tax cut bill passed in 1999 but vetoed by President Clinton.

26. Hawaii is the only state with an employer mandate that requires businesses to provide insurance. Congress granted Hawaii a special waiver so that it could institute its mandate.

27. As a practical matter, however, most large self-insured companies voluntarily include a number of coverage options that are required by the states.

28. See Wasley, "TennCare: Health Care Reform Dream or Disappointment?"

29. Merrill Matthews, Jr., "Government Rules Are to Blame," USA Today, October 1, 1998.

30. Such impressions are exacerbated by studies from companies such as Blue Cross that identify the number of people who do not have health insurance and add to that the number of people who are underinsured.

31. See David W. Emmons, "The Impact of Managed Care on National Health Spending: A Critical Review of the Literature," American Medical Association, Center for Health Policy Research, Discussion Paper 95-2, June 1995.

32. This figure varies somewhat between five age categories, with newborns and older participants receiving a larger allotment.

33. Blue Cross and Blue Shield, "Blue Cross Withdraws from TennCare Effective June 2000." However, the state recently stated that it will require Blue Cross and Blue Shield to stay in TennCare until the end of the year.

34. There are 28 states if TennCare, which is not considered a high-risk pool, is included. The reason TennCare is sometimes included is that it does provide coverage for the uninsurables.

35. An analysis of the cost of individual tax credits for health insurance by the actuarial firm Milliman & Robertson found that an increase of the tax credit by 50 percent above standard rates would be appropriate. Mark Litow and Peter Hendee, "Tax Credit Estimates for Under Age 65 Population (Revised)," prepared for the Council for Affordable Health Insurance, March 22, 1999.

36. TennCare's uninsured pay a premium of $489 a month, or about $5,870 annually, which is probably reasonable given the heavy managed care element in the program.

37. "Comprehensive Health Insurance for High-Risk Individuals," Communicating for Agriculture, Minnesota, 1999, p. 9.

38. People in high-risk pools do not represent all of the uninsurables in a given state. Many people who might be deemed uninsurable if looking to purchase individual coverage may have coverage through an employer or through a spouse. Indeed, high-risk pools are fluid, and people move out because they find insurance through an employer or spouse, their medical condition improves, they turn 65 and become eligible for Medicare, or they die. Milliman and &Robertson estimate that one can expect about 2 percent of the under-65 population for middle- and upper-income workers to fall into the uninsurable category. See Litow and Hendee, "Tax Credit Estimates for Under Age 65 Population (Revised)."

39. Robert E. Rector and Sarah E. Youssef, The Impact of Welfare Reform: The Trends in State Caseloads, 1985-1998 (Washington, D.C.: The Heritage Foundation, 1999), pp. 86-87.

40. Ibid. TANF, the welfare-to-work program established by Congress in 1996, replaced the 60-year-old Aid to Families with Dependent Children (AFDC) program.

41. See, for example, Families USA, "Losing Health Insurance: The Unintended Consequences of Welfare Reform," May 1999.

42. U.S. General Accounting Office, Medicaid After Welfare Reform, September 10, 1999, p. 2.

43. For a discussion of medical savings accounts and current legislation, see Greg Scandlen, "Medical Savings Accounts: Obstacles to Their Growth and Ways to Improve Them," National Center for Policy Analysis Policy Report No. 216, July 1998.

44. This option, too, would take a federal waiver. On the value of using federal Medicaid waivers to promote consumer choice options in the states, see Richard Teske, "How States Can Use Federal Waivers to Help the Poor and Test Health Reforms," Heritage Foundation Backgrounder No. 1337, November 2, 1999.