Skip Navigation

For nearly three decades, lawmakers fought persistent federal budget deficits. Over the past few years, however, those deficits have disappeared and been replaced with large and growing surpluses. The surprising speed with which those surpluses have emerged has caused a major shift in the public and political debate over federal budget policy.

The budget debate is no longer dominated by initiatives to reduce the deficit; it now focuses on various alternatives for using projected surpluses. This Congressional Budget Office (CBO) memorandum, which first appeared as Chapter 1 in CBO's March 2000 Budget Options volume, examines a range of major policy initiatives for expanding the scope of federal activities in health, retirement, and education. (In a few instances, it has been updated from the March 2000 version to reflect revised cost estimates and changes in federal laws.)

The memorandum was prepared by the Health and Human Resources Division under the direction of Joseph Antos. Sandra Christensen and Joseph Antos wrote the section on Medicare. Ralph Smith prepared the section on Social Security. Joseph Antos, Jennifer Bullard, and Rekha Ramesh collaborated on the discussion of health insurance coverage. Joseph Antos and Stuart Hagen wrote the section on long-term care. Bruce Vavrichek, Michelle Jewett, and Nabeel Alsalam prepared the section on education. Arlene Holen, Thomas Bradley, Paul Cullinan, Jeanne De Sa, Douglas Hamilton, Steven Lieberman, Audra Millen, and Kathy Ruffing provided helpful comments and suggestions.

Sherry Snyder edited the memorandum, and Christine Bogusz proofread it. Ronald Moore prepared earlier drafts, and Kathryn Quattrone prepared the memorandum for publication. Georgia Brown prepared the electronic versions for CBO's World Wide Web site (www.cbo.gov).
 
 

MEDICARE

• Expanding Benefits


• Long-Term Reform

SOCIAL SECURITY

• The Social Security Budget Story in Brief


• Proposals for Increasing Retirement Income


• Long-Term Reform

HEALTH INSURANCE COVERAGE

• Overview of Policy Approaches


• Expanding Government Insurance Programs


• Providing Tax Incentives for the Purchase of Insurance


• Expanding Private Coverage

LONG-TERM CARE FOR THE ELDERLY

• Use of Long-Term Care Services by the Elderly


• Current Financing of Long-Term Care for the Elderly


• Options for Financing Long-Term Care

EDUCATION

• Expanding Preschool Education


• Improving Elementary and Secondary Education


• Promoting Greater Investment in Higher Education

Every year, policymakers weigh the benefits and costs of undertaking new federal initiatives or expanding the scope of current programs. They also review existing spending and revenue options. With the recent dramatic change in the fiscal outlook--from projections of persistent deficits to ones of large and growing surpluses--policymakers are now considering more ambitious initiatives than they may have in the past. Some of the proposals that have been widely discussed would greatly expand federal activities, affecting millions of people and costing billions of dollars. Other proposals, though more modest, could also have a significant impact on the economy and society in general.

This memorandum analyzes some general themes for expanding the scope of federal activities that have received considerable public attention in recent months. Those themes include:

• Expanding Medicare benefits while ensuring the long-term financial health of the program;
• Raising the incomes of the elderly and preserving Social Security for future generations;
• Increasing the number of people who have health insurance coverage;
• Improving the financing of long-term care services for the elderly; and
• Expanding federal support for education at all levels, from preschool through college.

The discussion in this memorandum is intended to provide a broad perspective on several issues, including the nature of the policy problem, the extent of current federal programs, and the approaches that have been proposed to expand federal funding or involvement. Because the number of specific options that have been proposed is large, the memorandum does not reflect a comprehensive set of proposals. Nor does it provide detailed cost estimates; instead, it discusses the principles and major factors that would influence the costs of any specific proposal.

The inclusion or exclusion of a particular proposal does not imply an endorsement or rejection of that proposal by the Congressional Budget Office. As a nonpartisan Congressional staff agency, CBO does not make policy recommendations.
 

Medicare

Medicare, which is the second largest entitlement program after Social Security, provides health insurance coverage to people who are aged or disabled. In 2000, the federal government will spend about $220 billion to finance the health care of 39 million beneficiaries. Medicare spending has grown dramatically since the program began more than three decades ago, and that growth has been of increasing concern to policymakers. Between 1975 and 1995, Medicare spending grew faster than the economy, rising from 1.1 percent of gross domestic product (GDP) to 2.6 percent.

Following years of rapid growth, spending for Medicare has slowed considerably in the past few years. Indeed, spending was actually lower in fiscal year 1999 than in 1998. Likely reasons for the slowdown include the cost-reducing provisions of the Balanced Budget Act of 1997 and the reactions of providers to enhanced federal efforts to combat billing errors and fraud. Although the slowdown is a temporary phenomenon, projected baseline spending (net of premium receipts) for Medicare in fiscal year 2000 is now $8 billion, or nearly 4 percent, below the levels that CBO projected as recently as last summer.

The good fiscal news has not been limited to Medicare. According to current projections, the federal budget will have a surplus in each year of the 10-year budget window. If discretionary spending increases at the rate of inflation, the on-budget surpluses, which exclude Social Security and the Postal Service, are projected to total more than $800 billion over the 2001-2010 period.

Those fiscal developments have led to a greater focus on proposals to expand Medicare benefits, particularly to add coverage for outpatient prescription drugs and limit out-of-pocket expenses for beneficiaries. Medicare beneficiaries often incur substantial costs for prescription drugs, and many of them have little or no insurance protection. Moreover, unlike typical private insurance plans, Medicare does not cap beneficiaries' cost-sharing liabilities, leaving them without financial protection against high costs for services that the program covers.

Last year, the Bipartisan Commission on the Future of Medicare considered a number of ways to address those two deficiencies, although it reached no agreement on recommendations. The President has proposed a new prescription drug benefit for Medicare, as have several Members of Congress. Others have developed alternative proposals to have the program provide a more comprehensive set of benefits.

Concerns have been raised, however, that proposals to expand Medicare benefits could exacerbate the program's long-term financing problem. The leading edge of the baby-boom population will become eligible for Medicare in 2011, and program costs are certain to increase rapidly thereafter under current law. Demand for Medicare services will grow dramatically over the next few decades, while the number of people in the labor force will grow much more slowly. Between 2000 and 2030, for example, the number of Medicare beneficiaries will almost double, compared with an expected increase of about 13 percent in the number of workers paying payroll taxes. For that reason, some fundamental reform of Medicare's financing may be necessary whether benefits are expanded or not. One such reform is the premium-support proposal that members of the Bipartisan Commission developed last year.

Expanding Benefits

Compared with the typical health insurance plan offered by employers, Medicare's benefit package is limited in important ways. The program covers basic services--hospital stays, postacute care, physicians' services, and other outpatient care--but it does not cover other services that are important in the treatment of disease. Perhaps the most notable omission is coverage for outpatient prescription drugs, which represent a significant expense for many beneficiaries. In 1996, spending on prescription drugs accounted for over 10 percent of the cost of health services for noninstitutionalized Medicare beneficiaries. Almost half of that cost of approximately $25 billion was paid for out of pocket rather than through some type of insurance coverage.

Beneficiaries are potentially liable for significant costs even for the services covered by Medicare. For example, beneficiaries must pay a deductible equal to $776 in 2000 for an inpatient hospital stay, and hospital stays of more than 60 days require a substantial copayment. Care in skilled nursing facilities is also subject to substantial copayments after the first 20 days. Most outpatient services are subject to a $100 annual deductible, after which the patient is responsible for 20 percent of covered expenses (plus any additional amount that the physician is allowed to charge).

In part because Medicare leaves beneficiaries at risk for very large out-of-pocket costs, most beneficiaries seek some kind of supplementary coverage through employment-sponsored retiree health plans, private medigap plans, health maintenance organizations (HMOs), or Medicaid (for those whose income and assets are low enough to qualify). But such a patchwork arrangement generates a number of problems. First, it leaves unprotected a group of people who do not qualify for Medicaid or coverage under a retiree health plan and who cannot afford a private supplement. Second, the coverage available from private supplements is eroding. The share of employers offering health coverage to their retirees has been declining in recent years, and the supplementary benefits offered by HMOs are also being cut back in response to lower payment rates from Medicare. Further, because most medigap plans do not offer coverage for drugs, those that do so experience adverse selection (attracting enrollees who are more costly than average), resulting in such high premiums that few medigap enrollees purchase those plans. Third, the costs of administering insurance supplements are high because of the need to market to individuals and to coordinate benefit payments with Medicare.

Making Medicare's coverage more comprehensive would reduce or eliminate the need for private insurance supplements, but it would also mean that some of the costs now paid by beneficiaries, their employers, or state Medicaid agencies would be paid by Medicare. Expanding Medicare's benefits would also probably slow the shift of enrollment from Medicare's fee-for-service sector to risk-based Medicare+Choice plans because those plans are currently one low-cost way in which enrollees can supplement Medicare's coverage. It might also accelerate the decline in employer-sponsored retiree health benefits.

Covering Prescription Drugs. The President has proposed adding a drug benefit to Medicare. In the proposal first presented last summer (and included in the President's fiscal year2001 budget), the benefit would be offered under a new Medicare program--Part D. The new program would pay 50 percent of enrollees' total drug costs, up to a maximum annual benefit that would eventually reach $2,500. People enrolling in Part D would pay a premium that would be set to cover half of Medicare's cost for the new benefit.

The value of the Part D benefit would depend on how much an enrollee spent on prescription drugs. For example, someone whose total drug spending was $5,000 when the program was fully phased in would pay $2,500, and the program would pay the remainder. Enrollees whose total drug spending exceeded $5,000 would pay all of the costs above that amount. CBO estimates that the President's prescription drug proposal would add about $8 billion to Medicare's net costs in 2003, its initial year of operation. Annual costs of the drug proposal would increase to $30 billion in fiscal year 2010.

Because benefits would be paid from the first dollar of an enrollee's spending on drugs, the President's proposal would provide some benefit to nearly 85 percent of Medicare beneficiaries if all of them chose to enroll. However, because benefits paid to any given enrollee would be capped, Part D would leave those with the largest drug expenditures--about 20 percent of enrollees--unprotected once they spent $5,000. Although the cap on benefits would enable Medicare to better control costs for Part D, it would also limit the protection enrollees would have from large out-of-pocket costs.

Part D enrollees could be offered better insurance protection, but that would impose greater financial risks for Medicare. One option would provide fewer benefits to enrollees whose drug costs were low and greater protection to those whose costs were high. For example, a deductible could be required for enrollees' drug costs, but the benefit cap could also be eliminated so that Part D would pay 50 percent of all drug costs above the deductible. The deductible would have to be quite large, however, for that option to have federal costs that were no greater than those of the President's proposal.

A 50 percent coinsurance rate would give enrollees in Part D a strong incentive to be price-conscious. Other, more generous options could weaken that sensitivity to drug prices, which might loosen whatever pricing restraints drug manufacturers now face; if so, that could substantially increase federal costs for the new benefit. An example of such an option would be one that eliminated any further cost sharing for enrollees who spent more than a specific amount during the year.

Although competition among manufacturers might hold down prices for drugs that have generic equivalents or therapeutic substitutes, many new and unique drugs are patented and are protected from such competition. The price of those drugs could rise sharply if Medicare offered an unrestricted drug benefit. In that case, new mechanisms might ultimately be needed to control program costs. Such mechanisms might include giving administrators the power to deny coverage of some new drugs if they were too expensive or limiting the prices that drug manufacturers initially charge for new drugs.

Capping Cost-Sharing Requirements. Medicare provides substantial protection for millions of beneficiaries against the cost of health care services. But the insurance protection Medicare now provides against high out-of-pocket costs could be significantly improved if its cost-sharing requirements for currently covered services were limited to a maximum annual amount for each enrollee. Such stop-loss protection is typical in private insurance plans.

Adding stop-loss protection would increase Medicare's costs unless other aspects of the program were modified. For example, if enrollees' cost-sharing expenses were capped at $2,000 in 2001 with no other changes in current law, Medicare's net costs for the year would be about 6 percent higher. One option to limit costs would be to increase the cost-sharing requirements that Medicare beneficiaries would pay until they met an annual cap on those expenses. Combining the new stop-loss protection with cost-sharing requirements that include a single $800 deductible and a uniform coinsurance rate of 20 percent of amounts above the deductible, for instance, would lower Medicare spending by about 1 percent in 2001.⁽¹⁾ That alternative might be unpopular, though, because 70 percent of all beneficiaries would face somewhat higher cost-sharing expenses whereas only 10 percent would have their cost-sharing expenses fall because of the stop-loss protection.

Long-Term Reform

A booming economy and the prospect of large federal budget surpluses, in part because of the recent slowdown in Medicare spending, have given policymakers confidence that the program will be adequately financed over the next decade. But over the long term, Medicare spending will grow much faster than the rest of the economy.

Medicare costs will increase dramatically after 2010, when the first of the baby boomers reach age 65. The number of beneficiaries will double by 2030, and the growth rate of costs per beneficiary witnessed in the past may well accelerate with the aging of the Medicare population and continuing improvements in medical practice and technology.

If a balance between spending pressures and revenues is to be maintained in the long term, policy actions must be taken. Those actions might include options to increase premium revenues, change eligibility conditions to reduce the number of beneficiaries, or reduce costs per beneficiary.

The most direct way to maintain the balance between Medicare spending and federal revenues would be to move from the current defined benefit program, which mandates a broad set of benefits and provides unlimited federal contributions, to a defined contribution approach. Such a plan would limit the federal contribution to Medicare, allowing that contribution to grow at some rate that could be sustained in the long run (such as the growth rate of the overall economy). If the total cost of Medicare-covered services grew faster than the federal contribution, those additional costs would be borne by beneficiaries rather than by taxpayers.

Since beneficiaries typically are not working and have limited income, a defined contribution approach would sharply limit the financing available for health care. Unless other program changes were instituted that increased the efficiency of Medicare providers and thus slowed the growth in costs, some beneficiaries could ultimately have difficulty paying for basic Medicare services under such an approach.

An alternative to a defined contribution plan would be to shift some risk to beneficiaries while introducing mechanisms that would encourage more price competition among plans and providers. A premium-support mechanism, like the one developed by members of the Bipartisan Commission on the Future of Medicare, is one example. Important elements of such a plan for Medicare would include:

• Multiple plans in each geographic area, each offering at least the basic Medicare benefit package. All plans would compete for enrollment on the basis of price, quality, and (perhaps) benefits beyond the basic package.
• A contribution by Medicare to each enrollee's plan premium. The contribution could be set at or somewhat below the premium for the lowest-cost plan in each geographic area, or at some average of the premiums of the low-cost plans.

Under this approach, beneficiaries would be able to enroll in at least one plan for which they would pay no more than a modest premium. Because beneficiaries would pay the full additional premium of a more expensive plan, they would have a financial incentive to seek out low-cost plans. Competition among plans for enrollment would help induce plans to provide adequate service at the lowest cost possible. If Medicare's current fee-for-service plan was to continue, it would have to compete for enrollment on the same basis as private plans; otherwise, enrollees' incentive to seek out low-cost plans would be diluted.

How effective the premium-support approach would be in reducing growth in Medicare costs over the long term is uncertain. For one thing, the approach could not be implemented in areas where the Medicare population was too small to support multiple plans. In such areas, the traditional fee-for-service plan might have to be retained, and reforms to make that plan more efficient would also be important. Even in areas populous enough to support competing plans, extensive regulatory oversight would probably be necessary to ensure that plans were competing fairly, that enrollees were well informed, and that access and quality of care were maintained. Finally, it is unclear whether managed competition causes only a one-time reduction in cost for each enrollee who moves from fee-for-service care to a more efficient managed care plan or whether it can also slow cost growth once all beneficiaries who will switch to managed care have done so.
 

Social Security

This year, the Social Security program will pay about $400 billion in benefits to about 45 million retired and disabled workers, their families, and their survivors. Nearly all workers and their employers now pay Social Security payroll taxes, and most people over age 65 (as well as many younger people) receive monthly benefits from the program. This section presents basic information about the Social Security program and its financial outlook. It also examines several approaches the Congress could take in the near term to increase the income of the elderly. Any approach taken in the short term that increased the federal government's total financial commitments, however, could make the long-term budget problem more difficult to fix. This section concludes with a brief review of the proposals being discussed for funding Social Security over the long term.

The Social Security Budget Story in Brief

Social Security is, by far, the federal government's largest program, playing a critical role in supporting the standard of living of its beneficiaries. In recent years, people age 65 or older received about 40 percent of their cash income from Social Security. Elderly people whose cash income was relatively low were particularly reliant on Social Security. Families who had at least one member collecting Social Security benefits and who were in the lowest income quintile of elderly families received almost 90 percent of their income from Social Security, compared with only 25 percent for those in the highest income quintile.

Short-Term Budget Outlook. Spending for Social Security has been growing at roughly the same pace as the overall economy in recent years and will continue to do so throughout the next decade. The share of the economy devoted to Social Security has been between 4 percent and 5 percent of gross domestic product for the past quarter of a century and is expected to remain below 5 percent until 2013, according to the Social Security Administration.⁽²⁾ Meanwhile, revenues from Social Security payroll taxes have increased rapidly as the economy booms. The Congressional Budget Office projects that Social Security revenues will exceed program outlays by between $150 billion and $300 billion in each of the next 10 years.⁽³⁾

Long-Term Budget Problem. Once large numbers of the baby-boom generation begin receiving benefits, spending on Social Security (as well as on other programs for the elderly) will consume an increasing share of national income. The Social Security program's trustees project that under the current benefit structure, total spending will rise to almost 7 percent of GDP in 2030.

The expected increase in Social Security spending as a share of GDP results from the aging of the population born during the 1946-1964 baby boom. As that cohort retires and becomes eligible for Social Security benefits (starting in 2008), the ratio of beneficiaries to workers is expected to surge. By 2030, there will be 48 beneficiaries per 100 workers covered by Social Security, compared with only 30 today, according to estimates from the Social Security Administration. The number of beneficiaries is expected to increase somewhat faster than the number of workers thereafter, as life spans continue to lengthen.⁽⁴⁾

Much attention has been focused on the outlook for the Social Security trust funds. Last year, Social Security tax revenues, together with interest and other intragovernmental payments, exceeded expenditures by about $130 billion, bringing total Social Security trust fund balances to almost $900 billion at the end of December 1999. Projections show those balances rising steadily over the next two decades, peaking at $4.5 trillion at the beginning of 2022 and then diminishing until the balances are exhausted in 2034. Once the funds are exhausted, the payroll taxes collected for the Social Security program will equal only about 70 percent of benefits owed.

But the size of trust fund balances bears no relationship to Social Security's obligations or to the country's ability to fund benefits. Once Social Security benefits begin to outstrip payroll tax collections, the federal government eventually will need to reduce Social Security benefits or spending on other federal programs or raise taxes--regardless of the size of the trust funds. To fulfill the nation's promises to Social Security beneficiaries, the government must acquire resources from existing production when benefits are due. Actions taken now to boost capital accumulation, enhance productivity, and increase work effort could help build a larger economy in the future, which in turn would expand the capacity to fund future Social Security benefits, other federal commitments, and other claims of the elderly on the economy.

Proposals for Increasing Retirement Income

Despite the large amount spent on Social Security benefits, many elderly people still have low income. In the most recent year for which data are available, 1.0 million elderly men (7.2 percent of men age 65 or older) and 2.4 million elderly women (12.8 percent) had income below the poverty threshold.⁽⁵⁾ Many others have income slightly over the poverty line. As the number of elderly people increases, the number with low income is likely to rise as well.

The Congress could take several approaches in the short run to improve the lives of the elderly by increasing their income, particularly for those with low income, although that need not be the only goal of federal policies. To help raise the income of the elderly, the government could:

• Provide them with more income from Social Security or other public programs once they were no longer working;
• Encourage current workers to save more for their retirement by contributing to pensions, individual retirement accounts (IRAs), or other types of retirement plans; and
• Encourage people to work longer.

Numerous proposals in each of those areas have been made in recent years.

Increase Benefits. The first approach would be to target additional federal resources toward low-income elderly people. The Social Security program already does so by using a progressive benefit formula through which retired workers with a history of low wages receive benefits that replace a higher percentage of their preretirement earnings than the percentage replaced for other retired workers. The program also bases benefits for widows on the benefits for which their husbands had qualified, if that provides them with higher benefits than they would receive on the basis of their own past earnings. Both of those features could be strengthened, or new provisions could be enacted to specifically focus on beneficiaries with low family income. If those provisions were successful, some of the additional Social Security expenditures would be offset by reductions in outlays for Supplemental Security Income (SSI) and other means-tested programs.

For example, the "special minimum benefit" provisions in the current Social Security program could be revamped to increase benefits for people who worked many years at low wages. Fewer than 200,000 people receive Social Security benefits under the current rules for special minimum benefits, and the average benefit they receive is below the poverty line.⁽⁶⁾ Some Social Security reform plans call for a new provision that would raise the minimum benefit above the poverty line for retired workers who had worked most of their adult life at low wages.

One problem with trying to modify the Social Security system to strengthen its role in providing adequate income to retired workers who would otherwise have low income is that it is difficult to make such changes "target efficient." That is, it is hard to specifically focus additional Social Security benefits on people who are in low-income families. For example, some people who receive low Social Security benefits have pensions and other sources of retirement income or have a spouse who has high benefits. Likewise, although a widow has a much higher likelihood of being poor than does the average elderly person, a policy that focused on improving the benefits of widows could also help those with higher income as well and could miss the majority of the low-income elderly.

A more direct method of helping low-income elderly people would be to increase both the number who receive SSI and the amount of their monthly benefits. This year, that means-tested program will provide over 6 million recipients with almost $30 billion in federal benefits. (In addition, most states supplement the federal benefits.) About one-third of those recipients are age 65 or older; the others will qualify on the basis of their disabilities. Many people now eligible for SSI do not participate.

Increasing maximum monthly SSI benefits would raise the income of current recipients and could bring other low-income elderly and disabled people into the program. (The maximum monthly benefit for an individual with no other income is currently $512; for a couple, it is $769.) Increasing benefits could, of course, substantially add to SSI program costs, especially if more people participated in the program.

Increase Savings. Another approach for increasing the income of the elderly would be to subsidize or otherwise encourage or require people to save more for their old age. That approach could increase the resources available to future retired workers and their families, but it would not help people who had already retired.

The federal government encourages workers to save for their retirement, largely through various tax incentives. For example, workers can receive favorable tax treatment of earnings that they and their employers put directly into qualified pension plans, such as the commonly used 401(k) plans. They can also receive favorable tax treatment for money they invest in IRAs.

Additional incentives could be provided by broadening the eligibility for existing plans, increasing the amounts that workers can contribute, or developing new types of plans. For example, last year's proposal by the Administration to establish Universal Savings Accounts (USA accounts) would have provided eligible workers with matching contributions plus a flat contribution for lower-income workers to encourage them to put money into a retirement plan. Several of the proposals for partial privatization of the Social Security program (discussed below) would also encourage or require workers to put money into investment accounts that they could not withdraw from before age 62.

A key issue in assessing any proposal of this sort is whether federal spending (directly or through reduced revenues) would actually increase overall saving or merely substitute for saving that would have occurred without the proposal. The majority of workers already save something for their retirement through pension plans, IRAs, and other investments. If the federal government subsidized or required workers to put aside money in a specific type of plan, they might put less into other accounts. If such substitution occurred, any federal subsidy would reduce federal saving without an offsetting increase in private saving. Proposals that focus the subsidy on workers whose income is relatively low would suffer less from that problem because those workers are less likely to have pensions and other savings.

Increase Employment. Encouraging workers to delay retirement would also increase the income of the elderly. At age 62, most workers become eligible for Social Security benefits and must make two decisions:

• Should they continue to work and, if so, how much?
• Should they apply for Social Security benefits?

Within a year of becoming eligible for benefits, a majority of workers have stopped working (or substantially reduced their earnings) and a majority have filed for benefits. One consequence of those actions is that most of those workers subsequently have a smaller income than they would have had if they had postponed retirement. For example, workers who stop working and begin collecting benefits at age 62 receive monthly Social Security benefits that are at least 20 percent below the amount they would have received if they had delayed retirement and receipt of benefits until age 65. Moreover, if they instead continued to work, fewer years of retirement would need to be financed out of whatever private savings they had already accumulated, and they might be able to save more for their retirement. Likewise, the size of any private pensions they had might increase somewhat. (The relevant Social Security rules are described in Box 1.)

Box 1. Eligibility for Social Security and the Earnings Test

Workers can begin receiving Social Security retirement benefits as early as age 62, but the monthly benefits they receive will be lower than if they postpone filing. From age 62 to the full retirement age (also known as the "normal" retirement age), each year postponed adds about 8 percent to monthly benefits. For example, a worker eligible for a monthly benefit of $800 at age 62 could wait until age 65 and receive a monthly benefit of $1,000. Likewise, workers who delay collecting benefits beyond the full retirement age receive a credit for doing so. Each year delayed adds 6 percent to the monthly benefits of workers turning age 65 this year; the size of that credit is scheduled to gradually increase to 8 percent for subsequent birth cohorts. Until this year, the full retirement age was 65. Starting with workers born in 1938 (that is, workers who become eligible for retirement benefits this year), the full retirement age gradually increases from 65 to 67. For workers born in 1938, the full retirement age is 65 years and 2 months. For most practical purposes, that increase in the full retirement age simply reduces monthly benefits below what they would have been in the absence of the change; it does not change the age of eligibility for benefits. For example, when the full retirement age was 65, the benefits of workers who began collecting them at age 62 were permanently reduced by 20 percent. When the full retirement age becomes 67, workers will still be eligible to collect benefits at age 62, but they will incur a 30 percent reduction. (Workers who begin collecting retirement benefits this year at age 62 will receive about 1 percent less than they would have received had the full retirement age remained at 65.) The rules requiring the withholding of Social Security benefits if beneficiaries under age 70 have earnings in excess of certain exempt amounts--the "retirement earnings test"--are complicated and easily misunderstood. In 2000, the benefits of workers who are under the full retirement age are reduced by $1 for each $2 they earn above $10,080. (The earnings threshold for workers below the full retirement age automatically increases each year according to the annual increase in a national average wage index.) Workers whose benefits are reduced because their earnings exceed the threshold will subsequently receive higher monthly benefits--ultimately, about 8 percent higher for each year in which benefits are entirely withheld because of the retirement earnings test. The increase in benefits in many cases will be even more than 8 percent because the additional earnings can raise the earnings base on which benefits are calculated. In short, even though the retirement earnings test is often portrayed as a tax on work, it is more accurately described as a means of deferring benefits until the workers no longer have substantial earnings.

One way of encouraging people to work longer would be to eliminate Social Security's retirement earnings test so that people could begin to collect Social Security benefits at age 62 while they continued to work. Under current law, retirement benefits are reduced by $1 for each $2 that beneficiaries under age 65 earn above a specified threshold ($10,080 in 2000). Although those workers can later receive substantially higher monthly benefits as a consequence of that reduction, many people apparently are not aware of that and treat it as a simple benefit reduction. As a result, some people either stop working before they would have in the absence of the retirement earnings test or, at least, keep their earnings below the threshold.

Eliminating the retirement earnings test would be quite costly initially, because doing so would encourage workers who were already eligible for Social Security benefits to claim them. But the effect on Social Security spending would be small in the long run, according to the Social Security Administration's Office of the Actuary, because the earlier receipt of benefits would result in lower future monthly benefits.⁽⁷⁾ Moreover, if eliminating the earnings test caused some people to increase their earnings, the federal government would gain from added tax revenues.

Proponents of eliminating the earnings test contend that it is unfair and counterproductive to penalize people who want to work. Workers ages 62 through 64 who are otherwise eligible for Social Security benefits may think they are facing a 50 percent tax on their wages if they earn more than the threshold amount. That tax rate is in addition to the payroll taxes and income taxes they already must pay. Although these workers may be mistaken, proponents of abolishing the earnings test can point to strong evidence that some people are working less to avoid any reduction in their Social Security benefits.⁽⁸⁾

Opponents argue that the main effect of eliminating the earnings test would be to provide Social Security benefits to workers who already have a higher income than do many Social Security beneficiaries. The only people who would receive higher Social Security benefits if the earnings test was eliminated would be workers who earned above the threshold amounts. For example, 63-year-old workers who had earnings above the threshold this year and were otherwise eligible for the average Social Security benefit for workers their age would need to have a total income (earnings plus benefits) of more than $18,000 before their benefits would be reduced.⁽⁹⁾ Moreover, the annual earnings threshold for older workers is already scheduled to nearly double over the next two years. Another drawback of eliminating the earnings test is that the workers who decided to claim benefits while still working would receive lower benefits after they stopped working than they would have received if they delayed filing for them. Thus, encouraging people to claim benefits at an earlier age could subsequently increase the number of elderly retired workers and their survivors who have low income.

An alternative approach for raising the income of the elderly is to increase the earliest eligibility age for Social Security retirement benefits. Several proposals for slowing the growth in Social Security spending include provisions that would raise the earliest eligibility age from 62 to 65 and then link subsequent increases to changes in life expectancy. Such proposals would make people below the new eligibility age worse off by delaying their eligibility but would help ensure that they had higher income later. Unlike proposals to eliminate the retirement earnings test, this approach would initially reduce Social Security spending because workers would need to wait longer to become eligible for benefits. In the long run, however, raising the earliest eligibility age without making other changes in the program probably would have little impact on Social Security spending because the workers would ultimately become eligible for higher benefits. Nonetheless, if that approach resulted in people working longer, federal tax revenues would increase.

Proponents argue that the federal government should no longer be helping people retire at age 62, for several reasons. First, with the coming shift in the age distribution of the population, it makes little sense to give up the productive capacity and revenues that would result from more people working longer. Second, as life spans have increased and the average job has become less physically demanding, most people can work longer. Third, by enabling workers to trade lower future Social Security benefits for early access to benefits, the current rules for early retirement contribute to the higher poverty rates experienced by people who live to a very old age.

Opponents of raising the earliest eligibility age contend that it would be especially harmful to people who have little or no choice about when they stop working and who have few resources other than Social Security.⁽¹⁰⁾ Those opponents argue that many low-earning workers are in physically demanding or unpleasant jobs and that by age 62, if not earlier, they have worked long enough.⁽¹¹⁾ Moreover, by that age, opportunities for those workers are not very plentiful if they lose their job, particularly if the labor market is weak. Another argument made by opponents is that raising the earliest eligibility age would be unfair to workers with a below-average life expectancy, especially if they left no survivors eligible for benefits.

Long-Term Reform

Both the Congress and the Administration are interested in addressing the problem of funding Social Security over the long term in a timely fashion, before the baby boomers begin drawing benefits. But policymakers sharply disagree about how to do so.

Benefit Reductions and Revenue Increases. Slowing the growth in spending for Social Security would be one way of reducing future budgetary pressures. Previous CBO reports have reviewed a wide range of options for doing that. For example, the formula used to calculate benefits for newly eligible beneficiaries could be altered to reduce their initial benefits; the age at which full benefits became available could be increased; or the cost-of-living adjustments beneficiaries receive could be reduced.⁽¹²⁾

By itself, each option for slowing the growth in benefits would leave some beneficiaries worse off than they would be if they received the benefits scheduled under current law and the benefits were paid for in some other way. If the changes were made in a way that preserved the benefits of those with the lowest amounts, then larger reductions would need to be made in the benefits received by other retired workers. That is, the benefit structure would need to be made more progressive.

Benefit reductions could be avoided by increasing Social Security revenues. The Social Security program's trustees project that the gap between spending and revenues in 2034 will be about 5 percent of taxable payroll. Thus, an increase in the combined payroll tax on workers and their employers from 12.4 percent to 17.4 percent at that time would be an alternative way of dealing with the shortfall.⁽¹³⁾

Privatization. Numerous proposals have been made to pair a reduction in the Social Security program with the establishment of mandatory individual investment accounts that are owned and directed by the workers themselves. Such proposals, often referred to as privatization, would give workers control over how their money was invested. Most privatization plans contain at least four elements:

• Reducing Social Security benefits below the amounts specified under current law;
• Requiring (or at least giving a strong financial incentive to) workers to put a certain percentage of their earnings into individual investment accounts;
• Allowing workers generally to decide for themselves how their accounts are invested; and
• Prohibiting withdrawal of money from those accounts until the worker reaches a certain age.

Privatization proposals raise a number of issues concerning their potential consequences for the economy and for the income of workers and their families after the workers retire, become disabled, or die. Proponents of plans to replace all or part of future Social Security benefits with income from mandatory defined contributions contend that doing so would increase national income and enable workers to receive much higher returns on their investments than they could get by putting their money into the Social Security system. Opponents argue that those claims are exaggerated and that even partial privatization could subject workers, particularly low-wage workers, to unnecessary risks.

Although mandatory accounts would not resolve the projected shortfall between revenues earmarked for Social Security and program costs, they would provide an alternate source of income for former workers and their families if Social Security benefits were scaled back. The proposed USA accounts, cited earlier, could serve a similar purpose. Replacing part of Social Security with individual accounts would shift some financial risk, now borne collectively, onto the workers themselves, but at the same time it would offer workers the potential to increase their income in retirement.
 

Health Insurance Coverage

Despite significant economic growth over the past decade and the lowest unemployment rates in 30 years, the number of people who lack health insurance coverage continues to increase. The number of uninsured rose from about 35 million in 1991 to more than 44 million by 1998. The percentage of the population that was uninsured also increased over that period, from 14.1 percent to 16.3 percent. Lack of health insurance coverage is primarily a problem of the nonelderly, since Medicare covers people over the age of 65. In 1998, 18.4 percent of the nonelderly population was uninsured.⁽¹⁴⁾

The majority of nonelderly Americans had health insurance in 1998. Two-thirds participated in health plans through their employer, and nearly 7 percent purchased insurance in the individual market. About 14 percent of nonelderly people obtained coverage through a public program, primarily Medicaid. Some people were covered by more than one type of insurance during the year, and some were covered part of the year and were uninsured for the remainder.

Although policymakers have focused considerable attention in recent years on the lack of insurance coverage among children, adults account for most of the uninsured population. About 15 percent of children lacked health insurance coverage in 1998, compared with almost 20 percent of nonelderly adults.

The percentage of adults without insurance varies according to employment and income characteristics. In general, workers who are self-employed or who work in small firms are less likely to have health insurance than workers in large firms. Small firms may have higher health insurance costs than large firms because of smaller risk pools and higher administrative and marketing costs. Health insurance status is also correlated with income. More than a third of nonelderly people with income below 150 percent of the poverty threshold lack health insurance, compared with 12 percent of those with income above 200 percent of the poverty line.

The continuing growth in the percentage of people without health coverage stems partly from rising costs for private health insurance. Although private insurance premiums grew relatively slowly during the mid-1990s, premium increases of 7 percent or more--substantially greater than general price inflation--are expected this year. Those rising costs, coupled with less generous benefits, may have led an increasing number of employees to decline coverage offered by their employers.

Another factor in the continuing increase in uninsured people is the drop in Medicaid enrollment during the mid-1990s as the economy improved and welfare reform was implemented. Between 1993 and 1998, the percentage of nonelderly people covered by Medicaid fell from 12.7 percent to 10.4 percent. However, some people not currently enrolled in Medicaid continue to be eligible, even though they may no longer receive cash welfare benefits. Those people might obtain Medicaid coverage if they became ill and sought medical care.

The uninsured remain an important focus of concern among policymakers. People without health insurance are less likely to receive basic health care services than are those with insurance. They may delay treatment until a condition becomes serious, which can result in costlier treatment than would otherwise have been necessary. Hospitals and physicians are often uncompensated for the care they provide to uninsured people. As health care markets become increasingly competitive, providers have more difficulty covering those costs. As a result, less health care may be available to the uninsured.

Overview of Policy Approaches

Several policy approaches could increase the number of people covered by health insurance. They include:

• Expanding the scope and funding of government insurance programs. Policymakers have recently focused on broadening eligibility for existing programs rather than creating a new government insurance program.
• Providing tax incentives for health insurance purchased in the private market or from an expanded government insurance program.
• Regulating the private market to expand options for the purchase of lower-cost health insurance.

An alternative approach, not discussed here, would increase the direct provision of health services to people without insurance. That could be accomplished by expanding government funding for public health clinics and other providers.⁽¹⁵⁾

Expanding government programs or providing new tax incentives could involve substantial new federal spending or forgone revenues, particularly if those policies were intended to significantly reduce the number of people without health insurance. Subsidies approaching the full cost of insurance might be necessary to induce most low-income people who were uninsured to purchase coverage or participate in a government program. In contrast, expanding insurance options by changing the regulatory environment of the private market would not require significant new government spending but by itself might have limited effectiveness.

The effectiveness of any of those strategies in reducing the number of people without insurance depends both on the specifics of the proposal and on the broader policy environment. Recent debate over the cost-containing actions of managed care plans, for example, has raised legislative interest in imposing new mandates on health plans that would affect access to specialist care, payment for specific services, coverage of certain benefits, and portability of insurance. If such mandates were enacted, they would increase the cost of private insurance and ultimately increase the number of people who do not have private coverage.

Health insurance coverage increases the likelihood that an individual will receive basic health care services. But lack of insurance may be only one of the barriers to appropriate treatment. Low-income people, in particular, may not have access to physicians' offices near their home, may lack transportation, and may risk significant income losses (including loss of employment) if they take time off work to seek treatment for themselves or their children.

Expanding Government Insurance Programs

Three government programs--Medicare, Medicaid, and the recently enacted State Children's Health Insurance Program (SCHIP)--offer health insurance to elderly, disabled, or low-income people. Some 60 million people are expected to participate in those programs in 2000 at an annual federal cost totaling more than $300 billion.

Of the three programs, Medicare is the only one that is completely financed and run by the federal government. Both Medicaid and SCHIP are partnerships between the federal and state governments. The federal government sets basic standards for insuring populations and guidelines by which states will be reimbursed for a portion of the expenditures they incur for insuring individuals, but the administration of both Medicaid and SCHIP is left to the states. A federal initiative to expand coverage in those programs is thus not simply a matter of providing more federal funds. States may also require more flexibility in how they may use those dollars to better accommodate the needs and circumstances of their populations. Even then, some states may not expand their programs enough to make full use of the additional funds.

Making Medicaid Eligibility Broader and More Uniform. Medicaid is an entitlement program that provides medical assistance to low-income people who are aged, blind, disabled, or members of families with dependent children. It also covers certain other pregnant women and children. The program is funded jointly by the federal and state governments, with federal payments ranging from 50 percent to 83 percent of total expenditures. Outlays for Medicaid in 2000 are expected to be about $115 billion for the federal government and nearly $90 billion for the states. About a third of Medicaid spending is for long-term care services.

Medicaid is the principal source of health insurance for low-income people, but that coverage varies among states. Federal eligibility requirements are complex, and states have wide latitude to set their own eligibility standards above federally mandated levels. States must cover pregnant women and children under age 6 with family income below 133 percent of the federal poverty level. By 2002, states are required to phase in coverage of all children under age 19 with family income below the poverty line.

Beyond those requirements, states vary widely in the populations they cover under Medicaid. At their option, states may cover pregnant women and infants (under the age of one) whose family income is at or below 185 percent of the poverty threshold; about 30 states do so. Although some states have not covered all people whose income is below the poverty level, other states have chosen to enroll particular groups of people with income considerably above the poverty line, using options available under current law or through waivers granted by the Health Care Financing Administration. There is no guarantee that states will expand their programs even if federal funding is increased and federal restrictions on the uses of those funds are loosened, although some states surely would.

The number of low-income people who are covered by insurance could be increased by broadening federal eligibility requirements for Medicaid, making them more uniform among states for people facing similar economic circumstances. Options might include, for example, requiring all states to cover pregnant women and children with family income up to 185 percent of the poverty threshold or to cover all people up to some income level. Permitting or requiring states to cover groups who are not traditionally covered under Medicaid is another way to expand coverage.

Such policies would probably increase the number of people with insurance, but not all people targeted by the policy would enroll. Some people might wish to avoid the perceived stigma of enrolling in a welfare program. Others might delay enrolling in Medicaid until they needed services. Still others--who, before the passage of welfare reform in 1996, might have been automatically eligible for Medicaid as recipients of Aid to Families with Dependent Children--might not realize that they were eligible for the new benefit. Special outreach efforts might be required for the program expansion to be effective.

Other people (particularly those with higher income) who enrolled in an expanded Medicaid program would have had insurance even without that expansion. Some of them would have purchased individual coverage but would choose Medicaid because of its lower out-of-pocket costs, broader benefits, or both. Others would have had employment-based coverage. Some employees would refuse that coverage if they became eligible for Medicaid when the program expanded. Some employers would also have an incentive to drop the benefit if most of their workers could obtain coverage elsewhere, although that might leave some workers uninsured.

Broadening federal eligibility requirements for Medicaid would have a differential impact on states, depending on the generosity of their current programs. Less prosperous states tend to have relatively narrow eligibility rules, at least partly because they are less able to pay for large programs. Those states might argue that broader national eligibility requirements would impose an unreasonable fiscal burden on them.

Expanding the Scope of SCHIP. The State Children's Health Insurance Program provides enhanced federal matching funds to assist states in providing coverage for low-income children. Federal payments range from 65 percent to 85 percent of program spending, depending on a state's average per capita income. States may use SCHIP funds to expand Medicaid, to develop or expand other insurance programs for children, or to provide services directly. In addition, states may subsidize the purchase of family coverage through employment-based insurance if that option costs less than covering only the children.

Unlike with the Medicaid program--which, as an entitlement, serves all those who are eligible and enroll, regardless of the federal cost--federal funding for SCHIP is limited in the aggregate and at the state level. Federal outlays for SCHIP are expected to be about $2 billion in 2000. States are developing programs that may ultimately enroll 2.5 million children on an average annual basis. Given the size and focus of the current program, the extent to which proposals to broaden SCHIP would reduce the total number of people without health insurance depends on both the amount of new federal funding and the additional flexibility to design and implement programs that are extended to the states.

Only 19 states used SCHIP funds in 1998, the first year of operation, and many states have spent less than the amounts allotted to them in the federal budget. Recognizing that some states would need time to develop their programs, the Balanced Budget Act of 1997 gave states three years to spend their allocations and required the Secretary of Health and Human Services to redistribute unspent funds in the fourth year to states that had spent their allocation. But federal restrictions on the states' use of funds, particularly the limitation on outreach activities, have been criticized as unduly hampering the early development of SCHIP programs.

Some analysts have also criticized SCHIP as too narrowly circumscribed to be effective in increasing the number of children with health insurance. One option would expand SCHIP to cover the parents of eligible children. Such a policy could increase insurance coverage among parents of children already in SCHIP and encourage other parents to enroll themselves and their children in the program. As with other proposals to expand eligibility for government insurance programs, some of the people enrolling in an expanded SCHIP would have had group or individual coverage without the expansion. Some employers would discontinue their offer of insurance unless SCHIP subsidized that coverage.

Extending Medicare to Younger Ages. Unlike Medicaid and SCHIP, which do not offer insurance to all low-income people, Medicare provides nearly universal coverage to people age 65 or older and to many disabled individuals. In 2000, Medicare outlays will total about $220 billion and will finance health services for 39 million people.

Options for expanding Medicare eligibility target older adults who are not yet 65. Those people have more difficulty obtaining insurance than do younger people, and their premiums are high because they use more health services. The Administration has proposed allowing displaced workers ages 55 to 61 to purchase Medicare coverage. A separate proposal would allow certain people ages 62 to 64 to enroll voluntarily in Medicare. Participants in either program would be able to claim up to 25 percent of the buy-in premiums as a credit on their federal income tax.

The cost and effectiveness of such buy-in proposals depend on specific design features. The program for displaced workers would be narrowly targeted. Workers (and their spouses) would be eligible if they lost health insurance because of a job loss. Other eligibility requirements would include receiving employment-based health insurance for 12 months before losing their job, being eligible for unemployment insurance, and exhausting their coverage under the Consolidated Omnibus Budget Reconciliation Act (COBRA), which requires employers to offer continuing insurance benefits to workers (and family members) after workers leave their job. Premiums would be set at relatively high levels, although the tax credits could defray those expenses. Consequently, CBO estimated last year that participation would be limited to about 90,000 people by 2010. Those most likely to enroll would be people whose medical expenditures were higher than average for their age. As a result, premiums would not fully cover program costs. However, the bulk of the program's costs would come from the forgone tax revenue due to the tax credit, which would amount to about $0.7 billion through 2010.

The proposed Medicare buy-in for people ages 62 to 64 is designed to attract greater enrollment. Enrollment would be limited to people who did not have employment-based insurance or Medicaid. They would have to enroll as soon as they became eligible, which includes turning age 62 or losing employment-based coverage.

People buying in to Medicare under those circumstances would pay premiums that would approximately cover their expected cost to the program over their lifetime. The premiums would be paid in two parts. Before the age of 65, enrollees would pay premiums that reflected the average expected cost of benefits if everyone ages 62 to 64 participated in the buy-in. Up to 25 percent of the premiums could be claimed as an income tax credit. However, as with the buy-in for displaced workers, the people most likely to enroll would have higher costs than average for their age. Thus, premiums before age 65 would not fully cover the program's costs during those years. To offset those costs, people who bought in to Medicare early would pay a premium surcharge (in addition to their regular Supplementary Medical Insurance premium) once they reached age 65. The surcharge could not be claimed as an income tax credit.

Using those specifications, CBO estimated last year that the buy-in for people ages 62 to 64 would increase program outlays by $46 billion between 2002 (when the program would have begun) and 2010. Premiums would total slightly more than that, and tax revenue would be reduced by nearly $8 billion. About 650,000 people would participate in 2002, rising to about 1.3 million people by 2010.

Many of the people who would buy in to Medicare before they were 65 would have been insured even without the program. Most of them would have purchased coverage in the individual market. But the buy-in would give some people who were working and covered by employment-based insurance an incentive to retire early. CBO assumed that about 1 percent of workers ages 62 to 64 would retire early and buy in to Medicare if that option became available.

A policy that encouraged early retirement even to that limited extent would certainly not help Medicare's long-term financing crisis. A buy-in policy could, however, be a first step in advancing the age of Medicare eligibility beyond 65. The early buy-in could be coupled with a gradual move to a later age of normal eligibility comparable with the increase in Social Security's normal retirement age.⁽¹⁶⁾ The modest program savings that would be realized over the next 10 years from such an approach would grow rapidly in later years as an increasing number of people were affected by the change.

Some employers would drop their health insurance for retirees because of the availability of the Medicare buy-in. The prevalence of employer-sponsored retiree coverage has been declining, and the buy-in proposal would accelerate that trend. Other policy proposals, such as adding a Medicare prescription drug benefit, could worsen that adverse consequence of a buy-in.

Providing Tax Incentives for the Purchase of Insurance

The tax system currently provides substantial subsidies for health-related expenses, including the purchase of health insurance. The federal government annually forgoes over $100 billion in tax revenues, according to some estimates, by excluding from income and payroll taxes the contributions that employers make for health benefits and by allowing deductions for certain other health expenses. Those tax expenditures have significantly lowered the cost of health care for millions of people, primarily benefiting the more than 150 million people with employment-based insurance. Existing tax incentives might be restructured, or new ones added, to encourage additional people to purchase health insurance.

Subsidies Under the Current Tax Code. The largest health-related federal tax subsidy is the exclusion of employers' payments for health insurance and other health expenses from a worker's taxable income. Other health expenses include benefits paid through cafeteria plans and flexible spending accounts, as well as employers' contributions for long-term care insurance. That income tax exclusion will account for almost $60 billion in federal tax expenditures this year and will also reduce state income tax revenues. Employers' contributions for health benefits are also excluded from payroll taxes, accounting for as much as $30 billion in forgone federal revenues.⁽¹⁷⁾

Self-employed taxpayers may deduct part of their health insurance payments from taxable income. That deduction is "above the line" and is available to people who use the standard deduction as well as to those who itemize. Under current law, a self-employed person may deduct 60 percent of health insurance costs this year. That deduction rises to 100 percent by 2003.

Taxpayers who itemize their deductions may also use the medical expense deduction, which is geared toward families who incur high medical expenses (relative to their income). That provision allows them to deduct unreimbursed medical expenses that exceed 7.5 percent of adjusted gross income. Medical expenses include health insurance payments paid by the taxpayer, out-of-pocket payments for medical care, and certain costs for transportation, lodging, and long-term care.

In addition, people who choose to purchase qualifying high-deductible health insurance and are not otherwise covered may establish tax-preferred medical savings accounts (MSAs). MSAs are personal savings accounts that can be used to pay deductibles, copayments, and other health expenses not covered by insurance.

The tax system heavily favors health insurance purchased through employers over coverage purchased in the individual market. People without access to employment-based healh insurance cannot take advantage of a substantial tax benefit, yet they often face higher premiums than people who are covered through their job. Moreover, tax incentives in the current system are regressive. Since tax savings depend on the taxpayer's marginal rate, people in the highest tax brackets, who are most able to afford coverage, receive the largest subsidies. People who have low income and little or no income tax liability receive little or no subsidy if they buy health insurance.

The tax exclusion is a particularly inefficient way to subsidize health benefits. Because all insurance costs are subsidized, the exclusion encourages people to purchase more insurance than they otherwise would. For example, someone facing a marginal tax rate of 30 percent (counting federal and state income taxes and payroll taxes) is encouraged to spend a dollar for health insurance that is worth only 70 cents to that person--the remaining 30 cents is paid by the government as forgone tax revenue. Such excessive insurance encourages people to use more health services than they would have used without a subsidy.

Options for Expanding Tax Subsidies. Expanding tax subsidies for the purchase of health insurance would reduce the cost of that coverage, thus providing an incentive for more people to enroll in a health plan. The current structure of tax incentives could be extended to more people through the broader use of deductions, exclusions, or tax credits. Alternatively, the tax system could be restructured to expand insurance coverage more efficiently than at present.

People who do not have access to employment-based health insurance must pay the full unsubsidized cost of any coverage they buy in the individual market. As a result, they are less likely to have health insurance than are people who can obtain coverage through an employer.

One option would allow those people to deduct their health insurance expenses from taxable income. Those who purchased health insurance without a tax deduction would continue to do so under such an option, although many of them would pay less for insurance (after taxes) than previously. An expanded tax deduction of this kind would be regressive--benefiting those with higher income more than those with lower income--and might provide the greater benefit for people who would have purchased insurance coverage anyway. This option would probably induce few uninsured people to purchase insurance because most of them have low or moderate income.

Another option would offer a tax credit to people purchasing insurance in the individual or group market. Such an option might provide a credit of up to $1,000 to offset health insurance premiums, for example. That approach would be less regressive than expanding a tax deduction, but people with no income tax liability would not benefit from a nonrefundable credit. A refundable tax credit would be more effective in giving people at those lower income levels an incentive to purchase insurance.

The amount of a tax credit would have to be fairly large--approaching the full cost of the premium--to induce a large proportion of the uninsured population to buy insurance. Many uninsured people have low income and may not be able to pay much toward their health insurance. Some may be counting on the services of public hospitals and other publicly supported providers, which often write off the costs of care or require only modest payments from their patients. Moreover, many people who might be induced to buy insurance because of a tax subsidy would have access only to the individual market, whose premiums are generally higher than those in the group market. To make coverage more affordable, some tax credit proposals would permit uninsured people to buy in to government-sponsored insurance programs, including Medicaid, Medicare, or the Federal Employees Health Benefits (FEHB) program.

Other, more sweeping proposals would alter the current tax treatment of health insurance benefits in the context of new tax credit. One approach would limit the amount of the tax exclusion, which would increase tax revenues and discourage the purchase of excessively generous insurance. Those additional revenues could be used to finance a refundable tax credit. Another approach would be to completely replace the current tax preferences for employment-based coverage with a tax credit for everyone purchasing insurance. The tax credit could be allowed whether the insurance was purchased through the individual market, employers, or a government-sponsored plan.

Any proposal to expand tax incentives for the purchase of health insurance would have to deal with a host of technicalissues that would determine the proposal's cost and effectiveness in increasing insurance coverage.⁽¹⁸⁾ Some of those issues include:

• Defining the eligible group,
• Relating the subsidy to family income or some measure of need,
• Timing the receipt of the subsidy to coincide with the payment of premiums, and
• Defining and enforcing new regulatory standards for qualified insurance plans.

A tax subsidy could be targeted toward people who do not have access to employment-based coverage, or it could be made available to a broader group. Making a subsidy available to all who purchase health insurance might be the easiest policy to administer, but a substantial amount of federal aid would go to people who would have been insured anyway. Narrowing the focus to those who do not have access to employer sponsored insurance might be more cost-effective, but it would be administratively more complex. Any coverage that might have been available to a person and possibly a spouse would have to be verified, possibly long after the fact. In addition, such an approach might encourage employers to drop their health plans. Requiring employers to continue to offer that coverage could be difficult to enforce.

Tax subsidies could readily be tied to a family's income. But low family income, by itself, might distribute those subsidies inefficiently. As the current tax deduction for health expenses recognizes, a more accurate indicator would reflect both income and the level of health costs. The subsidy might also be adjusted to reflect variations in the average cost of health care in different geographic locations or other factors. Such adjustments could help people in high-cost areas buy as much care as people who receive the same dollar amount of subsidy but who live in low-cost areas.

An often-voiced concern about tax subsidies is that they would provide cash to families only at the time of tax filing, not when the cash was needed to pay premiums throughout the year. The health insurance tax credit that was available during the early 1990s did not offer payment advances, for example, and participation was well below expectations. One way to implement payment advances would be to lower income tax withholding. But making such adjustments precisely could be difficult, and some people might face unexpectedly high tax bills the following year. In addition, some other method of making advances would be needed for people who are eligible for a tax subsidy but do not have earnings.

Standards would be needed to define how health insurance plans that qualify for a tax subsidy could operate. Such standards might define a minimum benefit package that all health plans would have to offer, limit cost-sharing requirements, and establish other regulations for the private insurance market. Those regulations might include rules for medical underwriting, requirements to make insurance coverage available and renewable, limits on the premiums that may be charged, and other issues. Such standards and regulations are typically intended to protect consumers, but they also impose costs on the insurance industry that are ultimately paid by consumers.

Expanding Private Coverage

Expanding government health insurance programs or increasing the generosity of tax preferences for health insurance could require substantial new federal expenditures. Alternatively, regulation of the private insurance market could be modified with the intention of increasing health insurance coverage. Regulatory approaches have the appeal of not requiring new government spending, but they generally would impose some additional cost on the insurance industry that would ultimately be paid by consumers.

Both the Congress and the states have passed legislation affecting the benefits, cost, and accessibility of private health insurance, but the states have primary responsibility for regulating insurance. All states have passed legislation mandating the inclusion of specified benefits in health plans, which may have increased the cost of insurance. Most states also require insurers to issue insurance to all groups who apply and to guarantee the renewal of that coverage, and states frequently regulate the premium that may be charged for health insurance. In addition, some states have passed legislation creating health insurance purchasing cooperatives to facilitate insurance coverage for employees in small firms.

Federal regulatory initiatives have been intended to ensure more continuous coverage for people who are usually insured and to increase the number of lower-cost options available in the small-group market. Additional proposals might be considered to improve the availability and portability of insurance coverage and to reduce the cost consumers pay for that coverage.

Improving Insurance Availability and Portability. The Health Insurance Portability and Accountability Act of 1996 (HIPAA) addressed concerns that workers had become locked into their current employment because they risked losing insurance coverage for some period of time if they changed jobs. That act expanded COBRA protections for workers who leave their job. It also required insurers to make insurance available to people who had prior group or employer-sponsored coverage, and it guaranteed renewal of that coverage. The law limited the use of exclusions for preexisting conditions, which exempt the plan from paying for expenses related to a medical condition that already existed when the enrollee joined the plan.

The insurance mandates in HIPAA were intended to make group health insurance more available to workers and to make it easier for workers to change jobs by making that coverage more portable. But the law also imposed costs on insurers that would increase premiums somewhat--by about $500 million annually by 2001, according to CBO's estimates. The impact on insurance enrollment is uncertain: the increase in cost would tend to reduce coverage, but the loosening of insurers' restrictions would increase enrollment by some groups of people.

Additional initiatives might be considered to improve the continuity of private insurance coverage. Some options would extend the period of time over which COBRA coverage is available or broaden the availability of that protection. For example, firms that dropped their retiree health benefits might be required to offer their early retirees who were enrolled in the health plan extended COBRA coverage--perhaps until those retirees reached age 65 and became eligible for Medicare. Such a requirement could discourage employers from dropping their retiree health plans, but it could also discourage employers from offering coverage in the first place. Expanding COBRA coverage in that way would raise the cost of health insurance for workers, and fewer employees would enroll.

Making Small-Group Insurance More Affordable. Employees in small firms typically face higher health insurance costs than those in larger firms and are therefore less likely to have health coverage. Small firms lack purchasing power, limiting their ability to bargain for lower rates from providers and insurers. They have fewer employees to pay the fixed costs of a health plan, including marketing and enrollment costs, so their average administrative expenses are high. And small firms generally purchase coverage that is subject to state benefit mandates and premium taxes, both of which increase average premiums. Larger firms that self-insure are exempted from those state insurance regulations by the Employee Retirement Income Security Act.

Concerns about the affordability of insurance coverage in the small-group market have prompted recent proposals to establish association health plans (AHPs) and HealthMarts. Those new entities are intended to provide small firms and their employees with some of the premium-lowering cost advantages enjoyed by larger firms, including lower administrative costs and enhanced purchasing power. AHPs and HealthMarts would also enable small firms to avoid some regulations that generally increase their insurance costs.

AHPs could be sponsored by trade, industry, or professional associations and could offer a full range of health plans, including a self-insured plan, to their member firms. Both self-insured and fully insured plans (offered by a licensed insurer) would be exempt from state-mandated coverage of benefits. An AHP would offer its plans only to members of its sponsoring association and could price its premiums to reflect the expected health care costs of its association members rather than the costs of the small-group market as a whole.

HealthMarts would be nonprofit organizations that offered health insurance products to all small firms within an approved geographic service area. A HealthMart would have to make all of the plans it offered available to any small employer within its service area. Health plans offered through HealthMarts would be exempt from most state benefit mandates. Like AHPs, HealthMarts could offer premiums reflecting the expected health care costs of potential enrollees in small firms in its designated geographic service area rather than the entire small-group market in the state. Unlike AHPs, HealthMarts could offer only fully insured plans from insurance issuers licensed in the state.

Insurance offered through AHPs and HealthMarts could significantly lower premiums for some small firms compared with coverage offered in the traditional (fully regulated) small-group market. Some of those premium savings would result from reduced administrative costs or increased market power through group purchasing. Those savings would most likely be modest, however. Other savings would result from exempting AHPs and HealthMarts from state-mandated coverage of benefits that may not be strongly demanded by employees of small firms. AHPs and HealthMarts would also attract firms with healthier-than-average employees, further lowering premiums.

The exemption from state-mandated benefits could foster that favorable selection of firms with healthier employees. AHPs and HealthMarts might design benefit packages that were relatively unattractive to firms whose employees had costly health care needs. Lower-priced plans with leaner benefits might appeal both to firms that currently offer no coverage to their employees and to firms with healthy employees that already offer insurance.

If firms with healthier-than-average employees switched from traditional coverage to AHPs and HealthMarts, premiums for some firms in the traditional market would rise. However, proposals generally include requirements that would limit the ability of AHPs and HealthMarts to attract healthier groups. AHPs would have to offer their plans to any small firm that qualified for membership in the sponsoring association. Similarly, HealthMarts would have to make their plans available to any small firm located in a HealthMart's designated geographic area. And both types of plans would be subject to limits on the premiums they could charge. Moreover, aggressive efforts by AHPs and HealthMarts to obtain favorable health risks would add to administrative costs, which could temper such efforts to attract healthier groups.

In a recent analysis, CBO estimated that introducing the new entities would increase the number of people insured through small firms by approximately 330,000T19.⁽¹⁹⁾ Many more people--about 4.6 million--would be attracted by lower premiums to the new plans, but most of them would otherwise have been insured through the small-group market. Some firms and workers in the traditional market would drop coverage because their premiums would increase, but most would continue their coverage and pay slightly higher premiums.
 

Long-Term Care for the Elderly

The demand for long-term care services is substantial and will probably accelerate with the aging of the baby boomers. Over $120 billion, or more than 10 percent of national health expenditures, was spent on nursing home and home health care in 1998. Federal and state governments account for the bulk of that spending, perhaps as much as two-thirds of the total. The rest is paid for privately. Uncounted in that total are services provided to people with chronic health conditions by relatives and friends who are unpaid.

The demand for both paid and unpaid services is likely to grow as more people live longer. Policies could be adopted that might ease some of the financial pressures now facing families as they deal with the long-term care needs of a friend or relative and that might provide incentives for younger people to better prepare for such needs.

Use of Long-Term Care Services by the Elderly

Long-term care comprises a variety of medical and social services for elderly and disabled people whose disabilities prevent them from living independently. Formal, or paid, long-term care services may be provided in the home or community or in institutions for people who can no longer remain in their home. Not all people who could use such services receive them, however, because formal services are expensive if paid for out of pocket, and they may be less desirable than informal help from family and friends. Indeed, the most important sources of assistance for disabled elderly people who remain in the community are live-in caregivers and networks of family helpers. Despite recent rapid growth in long-term care spending, most long-term care services are still provided informally and are not, therefore, represented in expenditure data.

This year, 7.5 million people age 65 or older (or about 21 percent of the elderly population) are expected to require assistance because of physical disabilities, cognitive impairments, or behavioral problems. Of those people, 1.5 million will be in nursing homes; 2.2 million will receive assistance while living in the community, although they probably would have qualified for admission to a nursing home; and the rest will be less severely disabled but may still use long-term care services on occasion.

Over the next 30 years or so, the elderly population will double. Similar increases are foreseen for the "old old" population--those who are 85 or older and more likely to have disabilities that make them dependent on others for assistance. If current rates of disability among the elderly continue, almost 8 million severely disabled elderly people are projected to be living in 2030, with a similar number having lesser disabilities.

Those estimates are quite speculative, however, because of the uncertainty that surrounds future rates of disability and longevity among the elderly. If, for example, the Census Bureau's projectionsof the 85-or-older population are too low, as some demographers believe, the proportion of the elderly population in need of intensive long-term care could be considerably larger. By contrast, reductions in age-specific disability rates would lessen that effect. Recent data suggest that the incidence of disability has been declining over the past two decades, and those declines may continue.

Current Financing of Long-Term Care for the Elderly

Medicaid and Medicare, the two largest health financing programs, were responsible for about half of nursing home and home care expenditures for the elderly in 1995 (see Table 1). Medicaid is jointly funded by the federal government and the states and serves certain people who have low income and few assets. A significant fraction of Medicaid spending is for nursing home residents who have "spent down" their assets and income as a result of incurring large medical expenses. Because of the spend-down provisions, Medicaid is effectively the insurer of last resort for middle-income people facing high expenses for long-term care.

Table 1. Expenditures by the Elderly for Nursing Home and Home Health Care, 1995 (In billions of dollars)
Source of Payment	Nursing Home		Home Health Care		Total Expenditures		Percentage Share
Medicare	8.4		14.3		22.7		25.0
Medicaid	24.2		4.3		28.5		31.4
Other Federal	0.7		1.7		2.4		2.6
Other State and Local	0.6		0.5		1.1		1.2
Private Insurance	0.4		0.3		0.7		0.8
Out of Pocket and Other Sources	30.0		5.5		35.5		39.1
All Sources	64.4		26.5		90.9		100.0
SOURCE: Office of the Assistant Secretary of Planning and Evaluation, Department of Health and Human Services, as cited in Richard Price, Long-Term Care for the Elderly: Themes of Financing Reform, CRS Report RL30062 (Congressional Research Service, January 15, 1999).

Medicare pays primarily for acute medical treatment, but a sizable component of Medicare spending is for home health care and skilled nursing facility services. Although those services were originally intended to meet the short-term postacute needs of Medicare patients, Medicare's home health benefit is increasingly important for patients requiring chronic care. (Table 1 does not distinguish between postacute and chronic care services.)

Recent spending declines in Medicare outlays for home health services are not likely to persist. Those outlays declined in 1998 and 1999 partly because of legislative changes in payment methods and partly because of administrative actions curbing fraud and abuse in Medicare. Over the next decade, however, Medicare spending for both home health and skilled nursing care will grow substantially, CBO projects.

Because the federal government finances more than half of Medicaid's spending and all of Medicare's (apart from premiums and cost sharing paid by beneficiaries), it is the primary payer for long-term care services for the elderly. By contrast, the role of private insurance in financing long-term care is modest, accounting for less than 1 percent of all spending on nursing home and home care for the elderly in 1995. That may not be surprising since Medicaid provides some protection against potentially catastrophic long-term care expenses.

Many long-term care services are also provided informally by adult children or friends, who are not paid for providing that help. Informal services might be valued at $50 billion to $100 billion annually if they were purchased in the market. Economic and demographic changes may curtail family caregiving in the future, however. The increased participation of women in the labor force is already reducing the potential pool of informal caregivers. Average family size has also been declining, which further reduces the chances that an adult child will be present to provide informal services if the need arises. Reliance on paid care would be likely to rise if those trends continued.

Options for Financing Long-Term Care

A variety of policy options have been proposed to increase federal support for long-term care. Some would address the needs of people who are now using long-term care services and the needs of their families. Those policies might expand eligibility for services covered by Medicaid and Medicare or provide tax subsidies for paid or unpaid care.

Other options would increase opportunities for people to prepare financially for the possibility that they might need such services once they were elderly. Such options would encourage people to purchase long-term care insurance or to save more over their lifetime.

Expanding Medicaid and Medicare. Federal and state governments already finance long-term care expenses through Medicaid and Medicare. Those programs could be expanded, and eligibility rules could be liberalized.

All state Medicaid programs cover nursing home care, and many states also offer home- and community-based care in an attempt to avoid or delay a person's entry into a nursing home. To be eligible for Medicaid coverage, nursing home patients may not have assets with a value greater than a relatively low limit (that excludes the value of a home). In addition, all of their income except a small allowance must be spent on nursing home costs. If the patient is married, the spouse living in the community may retain higher amounts of income and assets. Most states permit people with relatively high income to qualify for Medicaid if they meet the asset test and if medical expenses exceed their income.

One option would allow patients whose long-term care expenses are covered by Medicaid, and their spouses, to retain more of their income and assets. That change would allow people with lower medical costs to obtain Medicaid coverage and would impose a larger share of total long-term care costs on the program. Such an option would increase the number of chronic care patients enrolled in Medicaid, primarily by allowing them into the program earlier. The option might also reduce some financial burdens for a spouse living at home, although nursing homes would also have an incentive to raise their prices for personal services paid for by the family.

Medicare provides postacute care services rather than comprehensive long-term care services. For example, Medicare covers up to 100 days of nursing home care for each spell of illness but only after the patient has been hospitalized for at least three days to treat an acute illness. Only the first 20 days are fully reimbursed, however; the remaining days require substantial coinsurance. That benefit could be made comprehensive by dropping the hospitalization requirement, eliminating the 100-day limit on coverage, and reducing coinsurance rates. Coupled with already generous home health benefits, those policy changes would establish a true long-term care benefit in Medicare. But such a policy would substantially increase Medicare spending as nursing home residents shifted from Medicaid and state-only programs (in which states bear some or all of the costs) or from private payment.

Another option would create a new benefit for respite care in Medicare. Respite care would cover paid services provided for a brief time, perhaps a week or two, to care for a chronically ill Medicare patient living at home. Those services would enable an unpaid caregiver to take time away from the patient. Such a policy could help caregivers cope with the physically and emotionally draining experience of caring for a loved one and might encourage more family members and friends to help someone with chronic needs, possibly delaying the patient's move to institutional care.

Tax Subsidies for Long-Term Care Services. Taxpayers who itemize may deduct unreimbursed medical expenses that exceed 7.5 percent of adjusted gross income. Taxpayers or their dependents who have substantial expenses for long-term care can deduct a portion of those costs, but the tax benefit is modest for most people. The deduction could be converted into a tax credit, which would provide a greater benefit to low-income people than would the deduction.

Another option would allow people to open a tax-preferred savings account, similar to a medical savings account, that could be used to purchase long-term care services. Such an account might be funded with pretax dollars, and interest accrued on account balances might also be excluded from taxable income. Medical savings accounts have not proved popular, however, perhaps in part because of restrictions on their use. Also, the amount of additional savings that might be gained through such vehicles might be modest, since some of the money going into the account would be diverted from other savings.

Expanding Long-Term Care Insurance. Although private long-term care insurance does not currently pay for a significant fraction of services, interest in such insurance is growing. The number of policies that have been written increased from about 800,000 in December 1987 to nearly 5 million by the end of 1996. Most long-term care insurance is sold in the individual and group association markets rather than through employer-sponsored group plans.

Private long-term care insurance protects policyholders against potentially large financial losses associated with a debilitating condition. The market for that insurance is concentrated among people with above-average income, who may have significant assets that they wish to bequeath to their heirs. Low-income people may reasonably have little interest in long-term care insurance. They may have limited assets to protect, they may not feel private premiums are affordable, and they are likely to be eligible for Medicaid coverage if they need long-term care services as they age.

People with higher income might face greater financial losses if they needed long-term care services and had to spend down their assets to become eligible for Medicaid. But even for those people, the availability of Medicaid is a significant deterrent to purchasing private insurance. Purchasers may find that long-term care insurance would give them more treatment options should the need arise. In addition, some states allow nursing home patients who purchased private long-term care insurance to become eligible for Medicaid with substantially higher-than-usual levels of assets. Even with those incentives, however, many people who buy private long-term care insurance allow their policy to lapse; perhaps as many as half of the people with that insurance drop their policy within five years of purchasing it.

The Congress could create new tax incentives for purchasing long-term care insurance. Options include a tax credit or "above the line" deduction for premiums paid by the taxpayer. Another approach would be to exclude from taxable income money withdrawn from qualified retirement plans that is used to purchase long-term care insurance. Such options would favor high-income taxpayers, particularly those who were already buying private coverage. Lower-income taxpayers, who are unlikely to buy long-term care insurance now for the reasons discussed above, would be unlikely to purchase that coverage unless the new tax benefit subsidized most of the premium.

Some people who purchased private insurance because of a new tax subsidy would use fewer Medicaid-covered services as a result, yielding some program savings. Those savings would be unlikely to fully offset the revenue loss, however. The net impact on the budget would depend on the details of the policy option.

Without other changes, tax incentives might have little impact on the private insurance market. Long-term care insurance remains a little-known product to employers, and the administrative cost of selling that insurance through the individual market is high. The Administration's recent proposal to offer long-term care insurance to federal workers through the Office of Personnel Management might promote that type of coverage more generally. Since that coverage would be sold to a group, premiums might be lower than those for comparable coverage sold through the individual market. Unlike the treatment of health insurance in the FEHB program, however, employees would pay the entire premium cost under that proposal--there would be no employer contribution.

Private insurance is one way for people to finance the costs of long-term care services themselves. Other policies, discussed earlier in the section on Social Security, could be pursued to increase retirement income, enabling the elderly to better afford health care and other services that they may need.
 

Education

The federal government historically has played a small role in funding the U.S. education system. Federal funds represent only about 8 percent of the cost of public elementary and secondary education, for example. State and local tax revenues provide most of the funding for public schools; parents of students in private schools pay most of those costs.

The same is true for other types of education. Most of the cost of preschool is paid by parents, with limited support provided by government sources for children in poor families. And although the federal government provided about $20 billion this year to help students pay for their postsecondary education through grants, loan subsidies, and tax benefits, family contributions and state subsidies have always been far more significant sources of funding for colleges and universities.

Nonetheless, the success of the education system is critical to the future of the nation, and additional spending has been proposed at all levels. The broad goals of those proposals are to promote equal opportunity; enhance the skills, productivity, and income of future workers; and provide greater assurance that children will become adults who can function effectively in society. Specific proposals might be more or less effective in achieving those goals.

Prominent among education spending initiatives are the following strategies:

• Helping children become better prepared to learn when they enter school by expanding the availability of preschool programs, most notably Head Start for low-income children.
• Improving the effectiveness of elementary and secondary schools by hiring more teachers and improving their training, as well as making improvements in facilities and other infrastructure.
• Increasing support for investment in education beyond high school by expanding federal student aid programs, especially Pell grants.

Expanding Preschool Education

Adequate preparation is a critical factor for success in school. Some analysts believe that the greatest return from additional spending in education could be obtained by investing in early childhood education.

Although universal public schooling is available starting at age 5, many younger children attend preschool programs. Nearly 40 percent of 3-year-olds attend some type of center-based program, as do almost 60 percent of 4-year-olds. Even with existing federal efforts focusing on low-income children, however, preschool attendance rates remain much lower among children from lower-income families than among those from higher-income families. For instance, preschool enrollment was about 38 percent in 1996 among 3- and 4-year-olds in families with annual income below $20,000, compared with over 65 percent in families with income above $50,000.

Head Start is the primary federal preschool program serving poor children. It provides a comprehensive set of services, mostly to eligible 3- and 4-year-olds, that includes child development, education, health, nutrition, social, and other services. The program strives not only to improve the education outcomes of children but to achieve other goals as well, including improving health status and reducing aggressive and other antisocial behavior.

In 2000, the program will enroll an estimated 877,000 children, over 85 percent of whom are from families with annual income below $13,000. The average federal service grant per child is about $5,800, with funds going directly to the approximately 1,500 public and private nonprofit agencies that operate the Head Start centers. Local grant recipients provide roughly 15 percent of total program resources.

Federal funding for Head Start has grown rapidly in recent years, rising from about $1.2 billion in 1989 to about $5.3 billion in 2000 (including advance appropriations). Increases occurred with the rise in the number of 3- and 4-year-old participants, which nearly doubled, and with the introduction of the Early Head Start program. That program provides early intervention services to pregnant women and families with infants and toddlers.

The Effectiveness of Preschool Programs. Two mechanisms could explain how children's experiences at age 3 or 4 might improve their subsequent education outcome.⁽²⁰⁾ Preschool might improve children's ability to think and reason as they enter school, enabling them to learn more in the early grades and keeping them "on track" toward high school graduation. It might also help increase their motivation to learn. The success children have in early grades could lead to higher expectations and added support from their parents and teachers, increasing their drive to succeed.

The effectiveness of preschool programs remains unclear, however. Most analysts agree that early childhood education programs in general can have positive short-term effects on participants' cognitive and social development, but there is less evidence about the longer-term effects of the programs. Although cognitive gains may fade, other effects--such as lower placement rates into special education and lower retention in grade--seem to persist. Analyses of small-scale "model" preschool programs also find in those programs long-term reductions in crime, teenage childbearing, and use of social services.

The efficacy of Head Start programs over the long term is even less clear. The positive long-term effects of model preschool programs may not pertain to Head Start programs because their teachers are often less well trained. Likewise, most Head Start programs do not provide some of the services, such as in-home tutoring, that are usually part of the model programs. Although both types of programs generally show favorable effects on reducing the placement of students in special education programs and on reducing the retention of students in grade, the question of Head Start's effects on participants in the long term remains open. The General Accounting Office has concluded that the body of specific research on Head Start to date is inadequate for use in drawing conclusions about the impact of the national program.⁽²¹⁾

Expanding Head Start. A range of proposals have been made to increase federal support for preschool education. Some options would make services like those provided in Head Start available to more 3- and 4-year-olds. For instance, one proposal would make federally supported preschool available to all 4-year-olds. Other options would increase the services provided to children who are already enrolled, including expanding the length of the program from half-day to full-day. Still other options would focus funding on programs that provide services to parents and children at earlier ages.

A specific proposal would be to increase Head Start funding sufficiently to enroll all poor 3- and 4-year-olds. In 1998, about one-quarter of eligible 3-year-olds and about one-half of eligible 4-year-olds were enrolled in the program. Enrolling all children who live in families with income below the federal poverty threshold today could raise the program's annual price tag from about $5.3 billion to about $12 billion if the average federal service grant per Head Start enrollee remained unchanged.

The federal cost could be higher, however, for three reasons. First, although the existing programs often make use of underutilized facilities and volunteer staff to save costs, significant further expansions of the program would be likely to exhaust those opportunities. Providing more classrooms and training more teachers to meet the program's expanded requirements would require additional resources. Second, a larger program would need to attract new teachers away from other jobs and career paths by offering them higher salaries. Furthermore, to prevent dissatisfaction and turnover among current teachers, their salaries would probably have to be raised as well. Third, for the positive effects of the model preschool programs to carry over to Head Start, many Head Start teachers probably would need increased training, and the program would have to provide an expanded array of services to participants and their families.

Achieving 100 percent enrollment of low-income 3- and 4-year-olds would be very unlikely, however--thus reducing the cost of the option. First, many parents prefer home-based care, irrespective of the availability and cost of center-based care. Second, the half-day schedule of most Head Start centers conflicts with the schedules of some working parents. It might be difficult for those parents to find adequate child care for the remaining part of the day and arrange for the transfer of their children from one place to another. Finally, the location of some Head Start centers makes them inconvenient for some families with limited transportation options.

Improving Elementary and Secondary Education

The federal government will provide approximately $24 billion in aid to elementary and secondary schools in the 2000-2001 academic year to fund a range of activities. Some aid supports improved education for children who are poor or have disabilities; other aid finances education reform and school improvement initiatives.

The government's first major effort to aid public elementary and secondary education (the Title I program) began in the mid-1960s as part of the war on poverty. Experience since then has shown that increasing the quality of schools that poor children attend can go only a small way toward closing the gap between their academic achievement and that of their higher-income peers. Other factors, such as difficult home situations and detrimental neighborhood influences, can undermine the efforts of schools to increase achievement but are much more difficult to address through federal policies. Federal spending on disadvantaged children through state grants for Title I totals $7.9 billion in 2000, or about one-third of all federal spending on elementary and secondary education.

In 1975, the Individuals with Disabilities Education Act (IDEA) became law, requiring states and school districts to provide a free, appropriate public education to children with disabilities. Doing so is very expensive. By some estimates, the cost of educating a disabled child is two to two-and-a-half times the cost of educating a nondisabled child, although that figure probably varies widely among states and school districts.⁽²²⁾ In passing IDEA, the Congress authorized a federal contribution for each disabled child served of up to 40 percent of the national average per-pupil expenditure (APPE) for all students. At about $5 billion, however, current federal funding gives states only about 13 percent of the APPE. Providing states with the 40 percent amount would require an additional $11 billion a year, assuming that the number of children identified as disabled remained unchanged.

Since the early 1990s, federal education policies have focused on a very different way of improving education outcomes. Along with continuing to aid special populations of students, those policies have encouraged broad-based education reform and school improvement.

Proposals to increase the effectiveness of U.S. schools range from state-level, top-down strategies to grass-roots strategies that address local problems. An example of a top-down strategy is one that would require states receiving federal funds to develop standards for what children should know in various grades and help states develop assessments of students' performance in various subject areas. An example of a grass-roots strategy is one that would support local groups who want to start new public schools (called charter schools) that implement specific education strategies appropriate to local needs. Another example is one that would create vouchers by tying Title I funding to disadvantaged students; those attending underperforming schools could be given the option of attending another public or private school, with the Title I funds following the student to the new school.

Other recent proposals would strive to improve schools by expanding or improving the inputs into the education process. Some proposals would support the professional development of teachers in areas such as science and math or would improve the quality of teachers by funding mentoring programs that team up experienced and inexperienced teachers. Other proposals would support state and local efforts to improve school facilities, including constructing and renovating school buildings and bringing Internet access to classrooms.

The quantity and quality of teachers are critical determinants of a school's success. Public elementary and secondary schools today employ over 2.7 million teachers. About half of them have a master's degree, and the median teacher has more than 15 years of teaching experience. Their average salary is an estimated $42,000 a year, and the starting salary is about $27,000.

Increasing the number of teachers in the early grades, thereby reducing class size, could be a tangible way to improve education outcomes. The average class size in elementary schools is about 24 students per teacher. The Congress appropriated $1.2 billion for academic year 1999-2000 and $1.3 billion for 2000-2001 to help reduce class size to 18 students per teacher in grades 1 through 3, and proposals have been made to continue and increase that amount.

Because average class size differs across states and school districts, the best method of allocating federal funds is not obvious. Funds could be distributed to jurisdictions according to the amount they would need to reduce their class size to a target level, but that method would give the greatest reward to those jurisdictions that had made the least progress on their own in reducing class size. Giving all jurisdictions the same amount of funds per student would avoid that outcome, but it would not necessarily result in the average class size falling to the target unless some jurisdictions reduced their class size below the target.

The best research evidence on the effectiveness of smaller classes on student achievement is Tennessee's STAR project.⁽²³⁾ Children entering kindergarten were randomly assigned to small classes of 13 to 17 students and regular classes of 22 to 26 students. Through third grade, students in small classes outperformed those in regular classes on both standardized and curriculum-based tests. In fourth grade, all students went to regular classes. Nevertheless, at least through eighth grade, a decreasing but still significantly higher level of achievement persisted for students who had been in the small classes.

Reducing class size in grades 1 through 3 from the current estimate of about 22 students per teacher to 15 would require hiring approximately 250,000 additional teachers. Paying those additional teachers at current beginning compensation levels would cost about $9 billion per year.

The salaries of current and new teachers would probably have to be raised to meet the extra demand, however. Those higher salaries could add another $4 billion to $8 billion annually to the price of this option, assuming that salaries of all elementary teachers rose by 5 percent to 10 percent. Additional costs would be incurred to recruit and train teachers, to give salary increases in future years, and to build the added classrooms that would be needed to accommodate the larger number of classes.

The task of reducing class size would be made even harder by the impending retirement of a large share of current teachers. Nearly 50 percent of elementary and secondary school teachers today are age 45 or older. Finding replacements for those experienced teachers when they retire would add considerably to the difficulty of expanding the overall number of teachers.

Promoting Greater Investment in Higher Education

Enrollment rates in postsecondary schools have increased in recent years, as have the returns on a college education. However, the cost of postsecondary education has also grown, having outpaced the growth in family income for more than two decades.

The federal government has long promoted attendance at colleges and trade schools. Perhaps the most important goals of federal policies for higher education are to remove the financial barriers to attendance faced by low-income students and to keep college affordable for middle-income families.

To help achieve those goals, the Congress created several programs, including a federal student loan program in 1959, the Pell grant program in 1972, and tax credits for postsecondary education in 1997. Last year, the student loan program provided $31 billion in loans to about 5 million students at a federal cost of approximately $4 billion. The Pell grant program provided more than $7 billion in aid to nearly 4 million students with very low income. And for the 1998 tax year, about 8 million filers received an estimated $4 billion in education tax credits and deductions for interest on student loans.

About 80 percent of students from upper-income families now enroll in college or trade school immediately after high school graduation. Less than 50 percent of students from low-income families enroll, even with the availability of significant amounts of federal and other aid.

Federal student aid has been increased several times in recent years:

• The interest rate on nearly all federal student loans was reduced by 0.8 percentage points in 1998 through 2003.
• The maximum Pell grant was increased incrementally from $2,900 for academic year 1997-1998 to $3,300 for 2000-2001.
• Tax credits of up to $1,500 for tuition expenses and tax deductions for interest expenses on student loans were created.

The Effectiveness of Student Aid in Increasing College Attendance. The availability of student financial aid--from the original GI bill to the more recent federal grant and loan programs--has allowed many students to attend college or trade school who otherwise would not have and others to pursue their postsecondary education further. On the basis of a recent study of students' experiences in the 1980s, a $1,000 increase in grant aid to all high school graduates would increase the proportion attending college or trade school by 4 percentage points.⁽²⁴⁾ Similarly, according to another study, a $1,000 difference in tuition at public two-year colleges was associated with a 7 percentage-point difference in enrollment rates among 18- and 19-year-olds.⁽²⁵⁾ There was no disproportional growth in enrollment by low-income youth relative to higher-income youth, however, after the Pell grant program was established in the mid-1970s. Overall, it appears that young people are sensitive to the cost of continuing their education beyond high school but that problems in understanding and applying for financial aid may deter college attendance, particularly among youth whose parents did not attend college.

Although the size of the effect is difficult to estimate, federal aid does induce some students, particularly those from low-income families, who would not have attended college or trade school to enroll in postsecondary education. It also increases the length of time some lower-income students remain in school. However, the aid also subsidizes many students who would have attended school without it.

Increasing Pell Grants. One option to promote greater investment in education would target additional aid toward students with low income by expanding the maximum award in the Pell grant program. That award could be increased from its current appropriated level of $3,300 to the full authorized limit of $5,100 in 2001. Doing so would raise the cost of the Pell grant program from $7.9 billion to about $15 billion.

Most of the added funding would go to the estimated 3.8 million current Pell grant recipients, whose average award would rise from $2,070 to about $3,400. The higher limit would also raise the number of current students who are eligible for Pell grants, adding about 600,000 new recipients to the program. Finally, raising the maximum Pell grant would induce some young people to enroll who previously found college or trade school too expensive. An estimated 300,000 new students would be added in that way.

In addition, the more generous aid would increase the number of affordable choices available to some young people already attending school. Some students might transfer from a two-year college near their home to a state four-year college a greater distance away. Others might give up jobs to focus entirely on school.

Several other considerations would affect the desirability of increasing the federal grant. Pell grants are available to any low-income student who has graduated from high school or passed the General Education Development tests. Many students who enroll in college drop out before graduating, in part because some of them are probably not adequately prepared. Increasing the amount of financial aid that is available might not be productive without taking steps to better prepare students.

One way to motivate students to prepare for college is to make them aware of available aid early in their school career. Some analysts believe that middle-school students are generally unaware of the amount of federal aid that is available to them and might therefore underestimate their ability to go to college. Programs to make all seventh or eighth grade students more aware of college aid might improve their preparedness for, and enrollment in, college.

A final consideration is that a large part of the return on higher education today is a private benefit. College graduates with a bachelor's degree earn substantially more than people with only a high school diploma. Furthermore, attending college enriches students' lives in other ways that are long lasting and extend to their children. Because students enjoy most of the benefits, one can argue that they should bear most of the cost. Accordingly, the role of federal policy might be to ensure that students who want to attend school are not prevented from doing so by temporary financial constraints; that could be achieved by increasing the availability of education loans. Although financing their education with loans increases the amount of debt the students amass by the time they leave school, federal policies already exist to provide borrowers with options for repaying loans that make the burden more manageable. For example, borrowers may extend the repayment period beyond the usual 10 years or choose graduated payments that rise over time with expected increases in income.
 

1. See option 570-13-A, "Simplify and Limit Medicare's Cost-Sharing Requirements," in Congressional Budget Office, Budget Options (March 2000), p. 228.

2. 1999 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors and Disability Insurance Trust Funds (March 30, 1999), p. 187, and unpublished tables from www.ssa.gov, based on the trustees' intermediate assumptions.

3. Congressional Budget Office, The Budget and Economic Outlook: Fiscal Years 2001-2010 (January 2000), p. 23. Over 85 percent of the revenues credited to the Social Security trust funds are from payroll taxes levied on workers and their employers. Most of the rest is from interest received on trust fund balances and from a portion of the income taxes paid by Social Security beneficiaries whose adjusted gross income is above a specified amount.

4. 1999 Annual Report, pp. 62 and 122. The intermediate assumptions in the report are that in 2030, the life expectancy of men who reach age 65 will be 17.1 years and that of women will be 20.2 years. In 2000, the life expectancy of men age 65 is 15.8 years, and that of women is 19.2 years. In 1940, soon after the program began, the life expectancies of men and women were only 11.9 years and 13.4 years, respectively. ("Life expectancy," as used here, is the average number of years of life remaining for a person if that person experienced the death rates by age observed in, or assumed for, the selected year.)

5. Bureau of the Census, Poverty in the United States: 1998, Current Population Reports, Series P60-207 (September 1999), Table 2. Poverty rates are particularly high for elderly women who are widowed, divorced, or never married.

6. Social Security Administration, Annual Statistical Supplement, 1999, p. 197. In December 1998, 154,000 beneficiaries received an average monthly benefit of $534. Most of those beneficiaries were retired workers, whose average benefit was $556. The annual poverty threshold for an elderly person living alone in 1998 was $7,818, or $651 a month.

7. The Social Security Administration's Office of the Actuary estimates that eliminating the earnings test for workers age 62 or older would worsen the 75-year actuarial balance by a small amount. See the memorandum from Stephen C. Goss, Deputy Chief Actuary, to Harry C. Ballantyne, Chief Actuary, "Long-Range OASDI Financial Effects of Eliminating the OASDI Retirement Earnings Test," September 13, 1999.

8. See, for example, Leora Friedberg, The Labor Supply Effects of the Social Security Earnings Test, Working Paper No. 7200 (Cambridge, Mass.: National Bureau of Economic Research, June 1999). Friedberg estimated that eliminating the earnings test for workers ages 65 to 69 would increase the total number of hours worked by people in that age group by about 5 percent.

9. In December 1998, the average monthly benefits paid to retired workers age 63 was $677 (see Social Security Administration, Annual Statistical Supplement, p. 181). Including the subsequent cost-of-living adjustments they would have received, the annual amount of those benefits would now equal about $8,300. Thus, workers receiving average benefits and facing the $10,080 threshold could have a total income of about $18,000 without any reduction in their benefits.

10. See Congressional Budget Office, Raising the Earliest Eligibility Age for Social Security Benefits, CBO Paper (January 1999), for an analysis of the characteristics, circumstances, and financial resources of men and women who claimed Social Security retirement benefits at age 62 or 63 in the early 1990s. That paper found that the majority of those retired workers had pensions and other sources of income sufficient to keep them well above the poverty line even if they had not received Social Security. But a sizable minority of them had non-Social Security income below the poverty threshold and might well have had serious difficulty finding a job.

11. If the eligibility age was raised, more workers would probably apply for benefits under Social Security's Disability Insurance program instead. If they were successful, that program would incur additional costs.

12. Congressional Budget Office, Long-Term Budgetary Pressures and Policy Options (May 1998), Chapter 3.

13. 1999 Annual Report, p. 169, and unpublished tables from www.ssa.gov, based on the trustees' intermediate assumptions. The trustees project that the gap will remain at around 5 percent of taxable payroll until 2050 and will then gradually increase to 6.5 percent by 2075.

14. Paul Fronstin, Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 1999 Current Population Survey, Issue Brief 217 (Washington, D.C.: Employee Benefits Research Institute, 2000), pp. 3-4.

15. Medicare and Medicaid also subsidize the provision of services to people without insurance through disproportionate share payments to hospitals that serve poor populations.

16. See option 570-19-B, "Permit Early Buy-In to Medicare and Increase the Normal Age of Eligibility," in CBO, Budget Options, p. 236.

17. Detailed estimates of those tax expenditures are reported by John Sheils and Paul Hogan, "Cost of Tax-Exempt Health Benefits in 1998," Health Affairs, vol. 18, no. 2 (March/April 1999), pp. 176-181.

18. For a more complete discussion of those issues, see Jack A. Meyer and others, Tax Reform to Expand Health Coverage: Administrative Issues and Challenges (Menlo Park, Calif.: Henry J. Kaiser Family Foundation, 2000).

19. Congressional Budget Office, Increasing Small-Firm Health Insurance Coverage Through Association Health Plans and HealthMarts, CBO Paper (January 2000).

20. Deanna S. Gomby and others, "Long-Term Outcomes of Early Childhood Programs: Analysis and Recommendations," The Future of Children: Long-Term Outcomes of Early Childhood Programs, David and Lucile Packard Foundation, Los Altos, Calif., vol. 5, no. 3 (Winter 1995), p. 10.

21. General Accounting Office, Head Start: Research Provides Little Information on Impact of Current Program, GAO/HEHS-97-59 (April 1997), p. 2.

22. M.T. Moore and others, Patterns in Special Education Service Delivery and Cost (Washington, D.C.: Decision Resources Corporation, 1988).

23. E. Ward and others, Student/Teacher Achievement Ratio (STAR): Tennessee's K-3 Class-Size Study (Nashville, Tenn.: Tennessee State Department of Education, 1990).

24. Susan M. Dynarski, Does Aid Matter? Measuring the Effect of Student Aid on College Attendance and Completion, Working Paper No. 7422 (Cambridge, Mass.: National Bureau of Economic Research, November 1999).

25. Thomas J. Kane, Rising Public College Tuition and College Entry: How Well Do Public Subsidies Promote Access to College? Working Paper No. 5164 (Cambridge, Mass.: National Bureau of Economic Research, July 1995).