(Archived document, may contain errors)

203

September 26, 1994

HOW 11NCPM%CMAV'HEALTH CMIE RUORM WOLTLDMAKETMGSWOME: TMROWLAND-BUMUMBRL

INTRODUMON With the Clinton-style health bills apparently dead on Capitol Hill, at least for this year, the majority leadership now seems intent on winning passage of an incremental bill in the final weeks of this Congress. One such incremental measure, being proposed as an amendment to the House majority leadership bill, is a revised version of a bipartisan bill introduced earlier this year by Representatives Roy Rowland (D-GA) and Michael Bilerakis (R-1FL).1 The major provisions in the Rowland-Bifirakis proposal are drawn from elements of the Affordable Health Care Now Act of 1993 (H.R. 3090), sponsored by House Minority Leader Robert Michel (R-EL), and the Managed Competition Act of 1993 (H.R. 3222). sponsored by Representative Jim Cooper (D-TN). 2 Incremental reform, such as the Rowland-Bilirakis bill, is widely promoted as an appropriately cautious approach-taking just one step rather than legislating radical and uncertain change. But while it is wise to be cautious when designing health care legislation, at this late stage in the legislative session it is virtually impossible for even an incremental bill to be crafted and enacted without disastrous unintended s for health care and the taxpayer. The "new" Rowland-Bilirakis bill, in short, is a classic example of a dis- aster waiting to happen. There are several reasons why pushing for incremental reform within the next month invites disastrous results. Among them:

Reason 1: The pending incremental bills are far from modest-they would revolutionize the health care system. Congress has spent the last few months debating sweeping changes to one-seventh of the entire economy. Some lawmakers feel that an incremental measure would be so modest in comparison that a bill could be enacted safely in a few weeks. They are very wrong. To- day's "incremental" bills are small steps only when compared with the sweeping changes involved in the Clinton-style plans. Yet these bills would dwarf any other health reform in the last generation. They would rechannel potentially hundreds of billions of dollars each year and change the very nature of the health care system. As the co-authors, of the Rowland-Bilirakis bill conrcdy describe their legislation: 7bese reforms may appear modest when set against the lofty and overly ambitious objectives of other proposals. But taken on their own, these provisions embody the most far-reaching changes in our nation's health care system since the creation of the Medicare and Medicaid programs. 3 Revolutionary changes are best considered carefully, not rushed through in the last four weeks of a congressional session.

Reason 2: In the complex health care system, even small changes that are not considered carefully can have unexpected side effects that are serious and costly. When a complex and subtly intertwined sector of the economy such as health care is altered by laws or regulations, the result is usually very different from that expected. History has shown that even small changes to the health care system can alter incentives and behavior in ways that produce often costly side. effects. Example. The Clinton Administration's modest Vaccine for Children program sought to ve the distribution of vaccine to children but has aimed into an embarrassing fi o. Notes Senator Dale Bumpers (D-AR), "A fairly simple law designed to benefit a relatively small of uninsured children was transformed into a bureaucratic nightmare.7oup Example. After a small change in the law governing federal payments to hospitals with a disproportionate state of uninsured and Medicaid patients, state officials discovered loopholes allowing states to be eligible for payments for activities not foreseen by Congress. The result: federal payments under the program shot up from $500 million in 1990 to $10.8 billion in 1992. White House officials were stunned. "Nobody Imew what happened to the money," says former White House official lbomas Scully. "It was money going out the back door when nobody was watching."5

The Rowland-Bilirakis bill contains several features that, unless crafted very carefully, could lead to runaway costs or create damaging incentives. Among them: V The bill would require younger individuals to pay much more in premiums than their coverage is worth while older Americans Will get a bargain. This "modified community rating" would lead many younger Americans to avoid cover- age, although how many would do so under the bill is unknown. V The bill requires insurance companies to enroll families irrespective of their health status. Since coverage is voluntary, and modified community rating artificially raises the cost of coverage for the young, there will be a new incentive for many younger, healthier Americans to forego insurance until they become sick and then to buy it. The degree to which this would happen under the Rowland-Bilirakis bill and its likely impact on the stabffity of the insurance industry are unknown. How- ever, the Health Insurance Association of America, which represents insurers, indicates that a 20-year-old nude could experience a rate increase of 44 percent under the bill. Most problematic of all, the Rowland-Bilirakis bill, like other measures, would establish a huge new welfare program in the form of a generous subsidy sys- tem for lower- and moderate-wage workers. The subsidy could be over $6,000 for a family of four and is phased out as income grows. But the phase-out itself introduces enormous problems. According to the Congressional Budget Office and calculations by Heritage analysts, the combination of income taxes and the withdrawal of health subsidies and other benefits would mean workers earning $20,000425,000 could face effective marginal tax rates of over 100 percent. In other words, low- and moderate-wage families would receive generous health benefits but would lose virtu- ally all incentive to work harder to improve their situation. In fact, if a family of four's income rose hom $16,000 to $30,000, taxes and the phase-out of benefits mekn that its actual take-home income would rise by just $91 -an effective tax rate of 99.42 percent.6 These problems are hardly minor. Possibly they can be solved. But they cannot be solved in a few weeks-nor even their possible implications assessed with confidence in so short a time. There first must be careful studies and hearings, and the legislation must be meticulously designed before it is enacted.

Reason 3: Congress and the public must understand where step-by-step legislation is to lead and approve of the ultimate goal. The majority leadership, apparently frustrated in its attempt to enact comprehensive changes in the health care system, now argues that Congress should at least enact the 11first step" this year. But this begs the obvious question: the first step to what? The debate over health care this year demonstrates clearly that the American people are unsure which direction they do want to take. The only thing they have made crystal clear is that they do not want to travel down the path charted by the Clinton White House and the majority leadership in Congress. Yet the leadership proponents of incremental reform state quite openly that they want an incremental bill that will move the health care system decisively toward the same goal of federal control that has been rejected this year by the American people. The political danger with congressional consideration of incremental reform this year thus is that supporters of Clinton-style reform simply will hijack an incremental bill, such as Rowland-Bilirakis, and use it to lay down the foundation needed to achieve in stages what they have not been able to achieve when their ultimate goal is made clear. Unfortunately the Rowland-Bilirakis bill would be a vehicle easily hijacked. The sensible course now for the congressional leadership is to recognize that it badly misjudged both the complexity of health care reform and the desires of the American people and not try to rush a bill through before November. Lawmakers then should take the opportunity presented by the election to begin a long-overdue national discussion with the American people in which the costs and trade-offs of alternative reforms are made clear. When this has been done, and the American people have given explicit guidance to their representatives, health reform can proceed in the next Congress with less chance of an- other political fiasco.

PROBLEMS M= THE ROWLAM-BILIRAKM BELL The Rowland-Bilirakis bill is a well-intentioned attempt to address some of the problems in today's health care delivery system and does make . in certain areas. But, unfommately, many of the bill's provisions likely would trigger new problems-some of them severe. At the very least, these provisions need to be considered carefully and revised. The bill makes some changes that would improve the existing health cam system. For ex- ample, Title 11, Section 2 101 resuscitates the 25 percent deduction for health insurance payments that until this year was available to the self-employed and in the year 2000 raises it to 100 percent..The section also introduces a new 25 percent deduction for health insurance for employees not covered by an employer-sponsored plan. Further, the bill creates opportunities for Americans to open medical savings accounts, limits the ability of health insurance companies to drop coverage or raise premiums because an individual has been sick, and makes it easier for small employers to obtain less expensive coverage. These provisions would nudge the health care system toward a situation where tax relief for medical care is not tied to an individual's place of work. Today the only way for most Americans to obtain tax relief for health insurance is to be enrolled in a health plan owned by their employer. Thus when an employee moves jobs, the policy stays with the employer and the employee must enroll in a new plan-usually chosen by the new employer. With full tax relief for health care coverage available for a plan chosen by an individual, insur- ance would be truly portable because a family could own its plan and not have to change it when changing jobs. Unfortunately, the Rowland-Bilirakis bill does not provide full tax relief, but only a de- duction for 25 percent of the cost of insurance. Worse still, individuals eligible for a com- pany-provided plan are denied a deduction for an alternative plan of their choice. In short, the bill does nothing to give Americans real choice over their health insurance.

The medical savings account provision (Section 1014) also is highly restrictive and of lit- tle real value. The account must be set up though an employer, and only the employer- not the employee-can make payments into the account. The employer contribution, more- over, is limited to the difference in premium price between a standard plan and cata- strophic coverage. In addition, this catastrophic plan must include a standardized benefits package approved by the federal government, and the employee cannot use the account to purchase supplemental insurance. The medical savings account in the bill thus does little more than permit the employee to pay cash for some medical care rather than file an insurance claim. Indeed, the account would be far more restrictive than the "cafeteria!' or "flexible spending" plans available un- der current law. These latter plans permit employees both to make their own tax-free contri- butions to the plans and to purchase supplemental insurance of their own choice as well as deductibles and copayments and medical services not included in their insurance. The bill makes it easier for groups of small employers to band together to purchase lower-cost insurance for their employees. It is difficult to fault this objective (although it would be much better to open up similar purchasing opportunities to other organizations as well, such as unions or churches), but there are several dangers in the approach taken by Rowland-Bilirakis. One is that large, regulated employer groups could evolve into the bu- reaucratic health alliances of the Clinton plan. Another is that such groups could become unstable if the insurance reforms in the bill also are enacted. The reason for this is that such pools are likely to be more attractive for employers with generally older and sicker work- ers than for firms with younger and healthier employees. This raises the risk that some pools could become unstable as low-cost firms withdraw to seek better rates outside the pool while higher-cost firms gravitate to the pool. But while the bill does make these somewhat beneficial reforms, albeit reforms associ- ated with many problems, it contains other provisions that would be bad policy or that could prove disastrous if enacted. Among these:

1) The low-income subsidy program, in reality a huge new welfare program, would eliminate the incentive to work for millions of Americans. Sections 2101-2135 of the Rowland-Bilirakis bill establishes a new subsidy program to help low-income Americans afford health insurance. This would be a huge new welfare program on top of the myriad of other income-support programs for welfare recipients and the working poor. But most problematic, the design of the program would have devastat- ing effects on work effort. Under the bill, families below 200 percent of the poverty level (or 240 percent in the case of children and pregnant women) receive assistance to buy the standardized health plan. If the individual's family income is 100 percent of the poverty level or less, the sub- sidy is equal to the full premium cost of the standard health plan. The subsidy amount is phased out on a sliding scale as income climbs to the maximum eligibility level. The amount of federal subsidy also would be reduced by the amount of any contribution to the cost of health insurance made by the individual's employer. One problem with this, as the Congressional Budget Office noted of a similar set of pro- visions in the bill introduced by Senate Majority Leader George Mitchell (D-ME), is the enormous and costly burden placed on states. The states would have to determine efigibility for the subsidies and to figure out how much each eligible individual should receive. The CBO describes this as "an enormous task for the states" and points out that there would be a strong incentive for states to avoid administrative costs b simply signing households of even questionable eligibility for the federal subsidies. Another problem is that the bill would create an enormous incentive for many employers to drop health insurance coverage for employees. This incentive would arise because the amount of proposed federal subsidy would be reduced, dollar for dollar, by the amount of any employer contribution. Thus, many firms could achieve considerable savings by aban- doning their health coverage and allowing the government to pick up the tab for the firm's low- and moderate-wage workers. True, any firm wishing to eliminate health coverage gen- erally would have to eliminate it for all employees, not just for lower-wage workers. But even if a minority of workers were low-paid, this would be a logical step, since for higher- wage workers who would not receive government subsidies, the firm could convert current health coverage into cash wages, permitting these employees to purchase their own care. 8 Since lower- and moderate-income workers would receive a subsidy as well as the cash, they could be far better off-at no extra cost to the employer. Moreover, firms which de- cided to retain health benefits for employees would have a very strong incentive to con- tract out most functions performed by lower-wage employees to other firms which did not provide health coverage. The bottom line: potentially millions of low- or moderate-wage employees in the United States would lose their employer-provided health coverage. Even if they received the equivalent value in extra cash, many doubtless would choose to forgo even basic catastrophic health insurance. However, the most devastating problem is that the Rowland-Bilirakis bill calls for a mas- sive expansion of the welfare state. By piling huge new health subsidies on top of dozens of already existing welfare programs and then phasing all these out as income increases, the bill would effectively obliterate the rewards for work and extra education for roughly one-third of U.S. families. This happens because phasing out a subsidy as income rises is equivalent to a special tax on additional income--each additional dollar of income leads to the loss of some part of the subsidy. But other subsidy programs also would be phased out alongside the health subsidy, including food stamps and the earned income tax credit. These multiple-subsidy phase-outs, together with federal and state income taxes on addi- tional income, result in very high effective tax rates. Due to this effect, the Rowland-Bili- rakis bill would create devastatingl; high combined marginal tax rates on most families earning less than $40,000 per year.

In its examination of the subsidy provisions in the h1itchell bill, which are very similar to those of Rowland-Bilirakis, the CBO declares: Senator Mitchell's proposal would discourage certain low-income people from working more hours or, in some cases, from working at all, because subsidies would be phased out as fan-dly income increas . 10

Adds the CBO:

In 2000, the effective marginal tax rate on labor compensation (wages and benefits) could increase by as much as 30 to 55 percentage points for workers with family income in the phase-out range. Moreover those levies would be added to the explicit and implicit marginal taxes that such workers already pay through the income tax, the payroll tax, and the phase-out of the earned income tax credit. In the end, some low-wage workers would keep as little as 15 cents of every additional dollar they earned. I I [The phase-out could produce effective] marginal tax rates of more than 95 percent for some workers. The nmdated system would also discourage some people who have spouses working at covered firms from participating in the labor force or at least from taking a job at a firm with more than 25 employees. If those people took a job at a covered firm, their wages would be reduced by the additional cost for insurance but they would receive no additional benefits. 12 Analysis of the Rowland-Bibrakis bill by Heritage Senior Policy Analyst Robert Rector indicates that the bill contains confiscatory marginal tax rates similar to those in the Mitchell. bill. The results of this analysis are presented in Charts 1 through 3.13 Chart I displays the welfare benefits which would be available to a family of four under . 14 the Rowland-Bi&akis bill. In addition to the proposed health benefits, families with in- con= of less thaii $20,000 per yew would receive aid from at lent three existing welfare for each extra $100 earned. A "combined marginal in raW shows &c combined effects of both taxes and the benefit reduction rates (BRR) of government aid and subsidies. For exaniple, an individual might face marginal taxes (such as income and Social Security taxes) of 30 pacent on earned income but may also receive goverturient welfare benefits which are reduced by 40 cents fbr each dollar of added earnings. Such an individual thus would face a combined marginal tax rate of 70 percent. For each added $100 earned, he would receive a net increase of just $30 in total income. 10 CBO, "A Preliminary Analysis of Senator Mitchell's Health Proposal," p. 16. 11 Ibid. 12 Ibid., p. 17. 13 The assumptions used in the analysis are presented in the appendix to this paper. 14 Charts 1. 2. and 3 illustrate what would happen if Rowland-Bilirakis were fully implemented in the year 1995. in reality, the bill phases in its health ca subsidies during a period leading up to the year 2004. The phase-in of health cam subsidies cream even lwgcr marginal tax rate problem in the intemning years.

programs: food stamps, the earned income tax credit (EITC), and the school lunch pro- gram. These three programs form the broad foundation of the current welfare system. Most working families with children and incomes below $20,000 per year are eligible for aid from these programs and currently receive benefits from them. (Many families receiving aid from these programs also receive benefits from other welfare programs such as Aid to Families with Dependent Children [AFDC], public housing, and the Women, Infants and Children Food Program [WIC]. However, because various fiscal and legal restrictions limit the number of low-income families which participate in these extra welfare programs, they have not been included in Charts 1-3. If the extra welfare programs were included, the mar- ginal tax rate problems would be even more severe dm those presented.) As Chart I shows, under Rowland-Bilirakis, a family of four earning $10,000 per year would receive over $14,000 in welfare once the proposed health benefits were added to food stamps, the EITC, and school lunch subsidies. ' The combined post-tax income for this family from earnings and welfare would be around $23,000. As earned family income rises, welfare benefits are reduced. Overall there would be a reduction in welfare of roughly $600 to $700 for each $ 1,000 in extra earnings in the $15,000 to $30,000 range. However, working families in this income range also must pay federal income tax, state income tax, and Social Security taxes. The effect of these taxes combined with the phase- down schedule of welfare benefits would be to reduce the net return to labor to nearly zero for millions of families.

Chart 2 displays the combined effect of taxes and welfare under the Rowland-Bilirakis bill. The chart shows the net increase in total family income that would result from an extra $ 1,000 in family earnings. For example, a family raising its earnings from $25,000 to $26,000 would achieve a net increase in total income of only $144. In many cases the re- duction in welfare actually exceeds the gain in post-tax earnings, creating a net loss in fam- ily income from added pay. For instance, a family whose earned income increased from $16,000 to $17,000 would lose roughly $1,500 in welfare. The result: an addition of $ 1,000 in earnings generates a loss of $867 in net family income.

Chart 3 shows the overall effect of Rowland-Bilirakis on family incomes. Once the pro- posed health benefits are piled on top of existing welfare programs (EITC, food stamps, and school lunches), the bill achieves a dramatic income leveling for families with earn- ings; between $15,000 and $40,000 per year. In this income range, the bill either entirely obliterates or drastically shrinks the rewards for improved education, greater work effort, and higher productivity. Overall, for families within the affected income range of $10,000 to $40,000, the Row- land-Bilirakis bill creates a combined marginal tax rate of roughly 75 percent. Thus, an in- crease in annual earnings of over $10NO will generate an increase in net financial well-being of only $2,500. For certain segments of the population, the effects are even more se- vere. For example, a family of four which cams $16,000 would have a net income of $25,81 8i combining post-tax earnings with welfare. By contrast, a family which earned $30,000 would have a net income of $25,899. Thus, an extra $14,000 in earnings would yield only an $81 increase in overall family income. Because of these strong work disincentives, the bill would have immediate and destruc- tive effects on the economy. In families mnicin less than $40,000, workers could be as- sumed to stop overtime work. Many women in dual-eamer, married couple families would quickly leave the work force since their employment would result in no income gain for the family. Husbands makin $35,000 per year or less would discover that they could ob- tain almost the same net income from working half a year as from working a full year. Mil- lions of workers would begin to work part-time or part-year. The special subsidy components of the Rowland-Bilirakis health care bill and similar provisions in the Mitchell bill would be among the most destructive pieces of legislation ever considered by Congress. If enacted, the Rowland-Bilirakis bill would impose a strait- jacket on U.S. society, leveling economic rewards for mom than a third of the population. To this population, the bill would send a very clear message: working hard, striving for a promotion or a better job, and studying in school will provide little or no financial return. This is a message the U.S. governmrit cannot afford to send.

2) The bill is an incremental step toward employer mandates and a standardized comprehensive benefits package.

Hospital benefits. Prescribed drugs, medicines, Surgical benefits. and prosthetic devices. In-hospital medical benefits. Other medically supplied services. Obstetrical benefits.

The Rowland-Bilirakis, bill would extend employer intrusion even deeper into decisions over a family's medical benefits. The bill also mandates employers to offer coverage and introduces the principle of some standardization into the cost and range of employer-pro- vided benefits. These provisions make dangerous concessions, increasing the chances that future congressional action or regulation would lead to a standardized comprehensive bene- fit system and an employer mandate to pay for at least part of that package. Specifically, Title 1, Section 1201 of the bill requires each employer to "make available' to each employee a group plan that includes at least a "standard" benefits package for the employee and his or her family. Actuarially equivalent plans are permitted (see below). These plans, as stipulated in Section 1101 of the bill, must be offered by all insurers wish- ing to do business with small employers. Exemptions are allowed for employers who have been in business for less than one year or for firms with no more than two employees. Sections 1101, 1102 and 1103 of the bill define the standard and catastrophic package. The standard package must include the following benefits, which are based on the standard plan available under the Federal Employee Health Benefits Program A health insurance plan providing this standard coverage may limit the amount, scope, and duration of preventive services if this is reasonably consistent with recommendations and periodicity schedules developed by appropriate medical experts. This standard package includes no specific limitation on deductibles. The task of setting deductibles is left to the Secretary of Health and Human Services. An employee has the right to choose a high-deductible version of the standard plan. In the case of the cata- strophic package, the bill does not specify any limits on the insurees decision to unpose co- payments or deductibles; the only requirement is that the deductible leads to a plan that is 20 percent less than the actuarial value of the standard plan. While specific benefits can vary somewhat in the standardized and catastrophic pack- ages, the bill requires each standard plan to have a standard actuarial value. This means that within rules to be certified by the federal government. substitutions of benefits beyond a minimum is allowed-provided that the cost of the package is the actuarial value of benchmark coverage for each geographic area of the specified benefits package under the standard or catastrophic plan. In other words, the requirement on employers is to provide or offer an insurance package with at least a certain price tag equal to over $4,800 for a family of four. Any substituted services must be "essential and medically necessary:' which means that, for instance, hair transplants could not be substituted for emergency hos- pital care. However, it is the federal government, in the form of the Secretary of HHS, who ultimately decides what substitutions are permissible.

This requirement on employers not only does nothing in itself to open up new, non-em- ployer group health options for Americans, but also requires all employers to intrude in health care decisions that should be in the hands of families. By mandating an extension of the function of an employer in this way, it is bad policy and opens the door to those who wish to push health care reform in the direction of greater government control. Specifically: V The provisions require employers to offer a comprehensive group plan of a certain dollar value and standard benefits. Even though the benefits in the bill's package can be varied, them will be strong pressure in Congress from those who prefer the Clin- ton plan for variations to be severely limited. While employers are free to offer additional plans, the paperwork and complexity in- volved in offering multiple plans means that most small and mid-sized firms likely would offer only the plan they am required to. The danger is a one-size-fits-all bene- fits package for most workers in such fums. 7be provisions introduce a health care mandate on all employers. To be sum, them are no requirements to pay for care, but the danger is that a requirement to offer a standard plan today will evolve into a mandate to pay for a certain level of compre- hensive care in the future.

3) The bill Will dramatically raise the cost of coverage for many youn& healthy Americans, encouraging them to drop coverage. The Rowland-Bilirakis bill introduces several reforms of the insurance market. Some of these, such as a requirement that insurers must renew coverage even if the health of an en- rollee or group of enrollees deteriorates, would be welcome. Others would be retrograde steps. One of the most troubling would be the requirement that insurers use "fair rating practices." Like most uses of the word "fair" in a bill, the practices to be required under Rowland-Biliralds would be decidedly unfair to millions of Americans. Tide 1, Section 1021 of the bill would permit insurers to adjust premiums to reflect age, geographic area (to account for medical costs), and family size. However, insurers could not vary premillms, fully with age, but only (after a phase-in period) by a ratio of 3-to- I for that element of the premium. When computing the age element of a prendum, in other words, an older, generally less-healthy individual could not be charged mom than three times the amount charged a much younger, generally healthier individual. Placing this restriction on premium-setting means that young Americans generally would have to pay much more for coverage than the true insurance value of their plan, while older Americans would pay much less. According to the Congressional Research Service, the current maximum variation for the age element of premiums among commer- cial insurers is 9.3-to- 1, and among Blue Cross/Blue Shield plans it is 8-to- 1. Thus, limit- ing the variation to 3-to- I would introduce significant distortions into the market. The financial effects and their implications could be severe. The Health Insurance Asso- ciation of America (HIAA) estimates that in the individual insurance market, a 20-year-old male on average would experience a 44.4 percent increase in rates while a 60-year-old male would see his premium fall by 3.7 percent. While a worker within an employment group typically would not see his rates change this much, if he worked for a firm with predominantly young workers or old workers he and his employer would face premium changes broadly in line with these figures. 16 Besides the issue of simple equity, limiting the ability of insurers to adjust premiums fully for age would have some serious side effects. One is that if premiums for young indi- viduals or groups of young workers are raised artificially in this way, there will be a strong incentive for many young Americans to become uninsured - especially if the cost of in- surance substantially exceeds their expected cost of paying for care directly. Older indi- viduals and groups, on the other hand, would have a greater incentive to acquire insurance. It is unclear what the net effect of these contrasting incentives would be on the net level of insurance coverage. The likelihood is that it would reduce the proportion of Americans with coverage. When New York State introduced strict community rating last year (mean- ing everyone had to be charged exactly the same premium), the number of insured in the small group market increased slightly while the number of New Yorkers with individual coverage fell by 44,000- or 12 percent - within nine months. The net reduction in cover- age in the state was 1.2 percenL The Rowland-Bilirakis bill's 3-to- I ratio for the age ele- ment is less severe than Now York's policy. On the other hand, Rowland-Bilirakis limits the waiting period for coverage of a pre-existing condition to six months, while New York permitted insurers to withhold coverage for one year. 7bus, younger individuals would be more likely under Rowland-Bdirakis than in New York to remain "outside' the insurance system until a medical problem developed, thereby reducing the incentive to insure.

CONCLUSION The overwhelming majority of Americans have health insurance. They are satisfied with their coverage generally but fear the loss of that protection. Americans thus want the peace of mind of knowing that should they lose their job, change jobs, move across the country, or have a preexisting condition, the choice of health beneflits and plans will be them-not subject to the arbitrary and exclusive will of their employer, a bureaucrat in Washington, D.C., or even an elected blember of Congress. 17 But reforming America's enormous health care system to give families this security is no easy task. Health cam represents one-seventh of the U.S. economy and is one of the most complex sectors of that economy. When even small changes are made to the system, many unintended side effects - sometimes with severe results - usually follow. If not crafted meticulously, an incremental approach can make things worse as well as better. The history of health care legislation shows that even small and carefully designed reforms can end up as disasters. But@ like most of the other incremental bills@ the Rowland- Bilirakis legislation is no tiny reform. It would be the most sweeping change in the health system for decades. Moreover, several of its well-meant features, such as the new subsidy program for low-income families, could prove extremely hamiful to those the bill is in- tended to help. These immense problems with the Rowland-Bilirakis bill, like the problems associated with other incremental bills, underscore the need to move cautiously on health care reform, not to rush through a bill in the final days of this Congress. Americans do want health care reform, but they want it done right. Stuart M. Butler Vice President and Director of Domestic and Economic Policy Studies

John C. Liu Policy Analyst

Robert Rector Senior Policy Analyst

APPENDIX

Computations for the income effects and marginal tax rates for the Rowland-Bilirakis bill demonstrate, for purposes of illustration, the results which would occur if the bill's pro- visions were fidly in effect in the year 1995. (The bill actually phases its health care subsi- dies in over time, completing the phase-in by 2004. During the phase-in period leading up to 2004, the marginal tax rate problems would be as great or greater than those presented in this paper.) Under current law and the Rowland-Bilirakis bill, the poverty threshold, tax brackets, welfare benefits, and'health benefits that determine the combined marginal tax rates dis- cussed in the text are all indexed to inflation. Thus, in constant dollars, the marginal tax rates presented in the text will not change over time. The income data and combined mar- ginal tax rates in 2004 (when the Rowland-Bilirake plan is fully implemented) will be ex- actly the same as those shown in Charts 1-3, except that nominal incomes will be some 20 percent higher due to inflation. AD computations in the paper represent a family of four husband, wife, and two chil- dren. Welfare benefits included in the calculations are food stamps, school lunch subsidies for two children, and the EITC. The EITC is set at the fall rate of 40 percent of earnings which is scheduled to come into effect in 1996 under current law. Taxes incorporated in the calculations include federal income tax, employee share of So- cial Security tax, employer share of Social Security tax, and estimated state income taxes. State income taxes are based on Ohio, which provides avenge tax rates for the relevant in- come range. The family of four is assumed to be eligible for a subsidy to purchase health care which would have a maximum value of $6,177 in 1995. This figure represents the CBO estimate of the average cost for families of community and experience-rated plans providing the level of coverage specified in the bill. The family is assumed to receive separate subsidies for parents and children. The value of the parents' subsidy is estimated to be two-thirds of the value of the ftdl family subsidy. The parents' subsidy is phased down linearly as family income rises above 100 percent of poverty and reaches zero when family income reaches 200 percent of poverty. The children's subsidy is estimated to equal one-third of the value of the full-family subsidy and is phased down linearly as family income rises over 100 per@- cent of poverty, reaching zero at 240 percent of poverty. According to the bill, families with earnings below 100 percent of poverty would receive an extra cost-sharing subsidy for out-of-pocket medical expenses. This cost-sharing subsidy is estimated to be worth 20 percent of the health insurance premium. In all calculations, the family itself is assumed to pay for the portion of the health insur- ance premium which is not subsidized directly by the government. According to the bill's provisions, the family would receive an income tax deduction equaling 25 percent of the cost of this self-purchased portion of the health insurance premium. The calculations as- sume that wage earners would be employed in fu= which do not provide employer health coverage.