print-only banner
The White House Skip Main Navigation
 Home > Government > Council of Economic Advisers

Council of Economic Advisors


Katherine Baicker
Member, Council of Economic Advisers

Link to Full PDF Document Full PDF Document (41K)

The United States faces important fiscal challenges in the coming years, both on the spending side and on the revenue side of the federal balance sheet. Before elaborating on the challenges raised by federal entitlement programs and the complexities of the tax code, however, I would first like to put the current federal budget in longer-run perspective. I will begin by presenting some basic facts about the evolution of fiscal spending and revenues over the postwar period. We will see that for more than forty years federal spending as a share of GDP has remained roughly stable, but that the composition of the federal budget has changed in important ways. While that stability seems likely to continue for a few years yet, if no action is taken, about a decade from now the share of national output spent by the federal government will begin to rise inexorably, with potentially severe adverse consequences for the economy.


Let’s begin our retrospective of federal budgetary trends with the expenditure side. I will focus on the years since 1962, a period which excludes the post-World War II transition and for which detailed data happen to be easily available.

The ratio of government outlays to the gross domestic product (GDP) is perhaps the best summary measure of the share of national resources being absorbed by federal programs. The mix of revenue sources and the size of deficits then tell us how that spending is being allocated across people and across generations, as well as how efficiently it is being financed. Despite the many changes in the economy and in the mix of government programs that have occurred since 1962, in most years federal outlays remained close to the long-run average value of about 20.4% of the gross domestic product. Since 1962, outlays have ranged from a low of 17.2% of GDP in 1965 to a high of 23.5% in 1983, a figure that reflected both an increase in military spending under President Reagan and the effects of the 1981-82 recession, which pushed GDP below its potential and increased the need for financial assistance to individuals and communities. Outlays averaged 20.7% of GDP during the 1990s, just above the long-run average, and they have averaged 19.4% of GDP in the last 10 years.

The stability of the federal government’s spending as a share of the economy over more than four decades masks some important underlying trends in the composition of expenditures, however, and those trends will have important implications for the size of federal outlays in the future. Most notably, since 1962 the share of GDP and of the government’s budget devoted to entitlement spending has risen steadily as the share of GDP devoted to defense spending has fallen. The biggest entitlement programs are Social Security and Medicare, which provide income and health coverage primarily to the retired, and Medicaid, a program under which the federal and state governments cooperate to provide medical care to lower-income people. (In fact, the majority of Medicaid dollars go to people over age 65 as well.) In 1962, entitlement spending, which at the time consisted primarily of Social Security benefits, absorbed 2.5% of the nation’s GDP and about 13% of the federal budget. Medicare and Medicaid were introduced in the 1960s, and over time both the generosity of the three major entitlement programs and the sizes of the eligible populations have increased. In 2005, these three programs accounted for 8.0% of GDP and made up just over 40% of the federal budget – and that does not include the substantial contribution to Medicaid made by the states.

This tripling since 1962 of the GDP share of the three major entitlement programs is mirrored by the decline in defense spending over the same period. In 1962, even before the bulk of the buildup for the Vietnam War, U.S. defense spending amounted to 9.3% of GDP. Expenditures on defense as a share of national output fell after Vietnam, rose somewhat under President Reagan, but then declined again after the fall of the Berlin wall. Defense spending hit its postwar low of 3.0% of GDP in the fiscal years 1999, 2000, and 2001. Under President Bush, spending on defense was increased following the 9/11 attacks, reaching about 3.9% of GDP in 2005, with additional funds going for homeland security.

All the other spending that the federal government does—from highways to education to national parks to interest on the national debt—takes a slightly smaller share of GDP today (about 7.9% in 2005) than it has historically (8.8% on average since 1962). Excluding interest payments on the national debt, the comparison is 5.1% of GDP in 2005 for spending outside of defense and major entitlement programs versus an historical average of 5.7% of GDP.

To summarize, then, over the past four decades federal spending has been fairly stable around its long-run average of 20.4% of GDP. However, that overall stability masks an increase in federal spending on three large entitlement programs of about five-and-a-half percent of GDP and a commensurate decline in defense spending.

The revenue or collections side of the federal budget likewise shows overall stability and some interesting underlying trends. Total federal revenues have averaged 18.2% of GDP since 1962, with modest variation around that mean. Comparing average revenues of 18.2% of GDP with the long-run spending average of 20.4% of GDP, we see that in the average year the federal government has run a deficit, inclusive of interest payments on the national debt, of about 2.2% of GDP. Despite those deficits, the ratio of federal debt held by the public to GDP has fallen over the past four decades, from about 44% in 1962 to about 37% in 2004. (For comparison, the current debt-to-GDP ratio is about 57% in the Euro area and about 78% in Japan, according to the OECD.) The U.S. debt-to-GDP ratio has fallen since 1962 despite sustained deficits because strong U.S. economic growth has allowed increases in GDP to outstrip increases in the national debt.

Although federal revenues have been generally stable as a share of national output, as on the spending side, the composition of revenues has changed. Payroll taxes, primarily Social Security contributions, have risen from about 3% of GDP to just over 6% of GDP today, reflecting the need to finance rising benefits outlays. Offsetting that increase were declines in the share of revenues collected as excise taxes and from corporate income taxes (the latter are also highly cyclical). Most of the increase in payroll taxes, as a share of GDP, as well as the offsetting changes in other revenue sources, had occurred by the early 1980s.

The largest single source of federal revenues is the personal income tax. Here, the most important trends of the past four decades relate to the structure of the tax. In particular, over the years, marginal income tax rates—the tax paid on an extra dollar of income—have declined substantially. Economists generally agree that low marginal rates improve incentives for work, entrepreneurship, and other productive activities, and thus promote economic growth. Low tax rates also reduce the incentives to do work “off the books” or to look for economically unproductive ways to shelter income from taxation. At the same time, the tax base has shifted from being primarily a tax on income towards a hybrid income-consumption tax. Because of continuing growth in the economy and in taxable income, other changes and reforms in the tax code, and various forms of “bracket creep” (the fact that higher prices and incomes tend to push people into higher tax brackets), the long-run decline in marginal rates has not resulted in a corresponding downward trend in personal income tax collections. Indeed, revenues from the income tax have been fairly stable since 1962 at about 8.2% of GDP and about 45% of federal revenues.

To summarize, tax revenues, like federal expenditures, have remained fairly stable as a share of GDP over the past four decades, at about 18.2% of GDP. Payroll taxes have risen with Social Security benefits, offsetting a decline in excise taxes and corporate income tax collections, while personal income tax collections have been relatively stable.


With this background, let’s return to the current and prospective fiscal situation. In fiscal year 2005, the federal government ran a deficit of just under 2.6% of GDP on spending of 19.8% of GDP and revenues of 17.3% of GDP. This deficit is higher than the long-run average deficit of 2.2% of GDP but a significant reduction relative to the previous year. The improvement in fiscal year 2005 was due largely to a 14.6% increase in federal revenues from the previous year, the product of a rapidly improving economy.

Bringing down federal deficits over the next few years would be a positive and very important step. But even greater challenges lay ahead. I have emphasized the remarkable stability over more than four decades in the share of national output spent by the federal government. However, without major policy changes, there will soon be a major break in the generally stable fiscal situation that the United States has enjoyed for most of the postwar period. The opening shot in this transition will happen just a few years from now, in 2008, when the first baby boomers reach age 62, the age of early retirement for Social Security. As you know, the retirement of the baby boomers, and the ongoing graying of America, will contribute to an increasing ratio of retirees to active workers. But the underlying problem is more than a “pig-in-the-python” effect resulting from the retirement of a single large cohort – it is not a temporary or one-time event. Rather, even as the baby boomers begin to depart from the scene, the combination of lower fertility rates and longer life expectancies will imply a permanent decline in the number of active workers available to support each retiree. That ratio, which was 16 workers per retiree in 1950, and which is currently about 3.3 workers per retiree, will decline more or less permanently to about 2 workers per retiree by the middle of the century. It is no consolation that the demographic ratios will be even more unfavorable in Japan and some European countries.

To put the problem in terms of the share-of-GDP metric I have emphasized today, Social Security benefits payments are projected to rise from 4.2% of GDP in 2005 to 6.0% of GDP forty years from now, in 2045. (This share-of-GDP perspective cuts through a confusing debate surrounding the role of the Social Security Trust Fund: for the government and, more importantly, for society as a whole, the real burden of future Social Security benefits is better represented by the projected ratio of those benefits to GDP than by calculations that involve the state of the Social Security Trust Fund.)

Even more dramatically, Medicare spending is expected to rise from 2.3% of GDP currently to 7.0% of GDP in 2045, while federal spending on Medicaid (that is, excluding state contributions) will expand from 1.5% to 2.6% of national output. Much of the increase in Medicare spending is driven by per-beneficiary costs rising faster than inflation, even controlling for demographic trends.

In sum, the costs to the federal government of the three most important entitlement programs are expected almost to double from 8.0% of GDP today to about 15.6% of GDP in 2045. In 2005, all other spending programs of the federal government, excluding interest payments on the national debt, amounted to 9.0% of GDP. Thus, if the major entitlement programs grow as expected, and other programs remain constant as a share of GDP, in 2045 the federal budget excluding interest on the debt will consume 24.6% of the GDP, compared to 17.0% today, with continuing increases beyond that date. Adding back in interest on the national debt could make the share of GDP absorbed by the federal budget much larger still. Put another way, if total federal spending excluding interest were held constant, the rise in entitlement spending would just about fully crowd out all other federal spending.

An alternative metric is the so-called fiscal imbalance, the difference between the present value of promised benefits and the present value of dedicated resources (with an infinite horizon). According to the Trustees of the Social Security and Medicare programs, the fiscal imbalance of Social Security is $11.1 trillion and that of Medicare is $70.5 trillion. For comparison, aggregate household wealth in the United States is currently about $50 trillion.

The implications of these trends are dire. If the major entitlement programs grow as forecast, our children will be forced to choose between massive tax increases, near-elimination of all government programs outside of entitlements (including defense and essential services), or some combination. Of course, considerable uncertainty surrounds any projection forty years in the future; nevertheless, the basic trends and their likely fiscal consequences are difficult to deny.

Because of these rising entitlement obligations, ensuring long-term fiscal stability requires much more than addressing current spending and deficits, important as that is. Also necessary is finding a means of controlling the costs of the major entitlement programs, without compromising their essential functions.

On Social Security, as you know, the President has called on Congress to enact reforms that achieve permanent solvency without changing benefits for people born in 1950 or earlier, that maintain or enhance the progressivity of the system, that do not increase the payroll tax rate, and that allow voluntary personal accounts. Specifically, the President has urged Congress to consider a system of progressive indexing. With progressive indexing, the initial benefits of lower-income workers would continue to rise at the same rate as wages, as under current law, but the initial benefits of the highest-earning workers would be tied instead to increases in prices. This measure would constitute a major step toward permanent solvency, while at the same time increasing progressivity, protecting lower-income workers, and ensuring that all future retirees receive benefits that are at least as great in real terms as those being paid today.

Because payroll taxes are currently sufficient to cover benefits, some temptation exists to leave this problem for a later generation to solve. But the long-term financial deficit of Social Security is growing rapidly (indeed, the annual increase in Social Security’s financial shortfall significantly exceeds the official government deficit), and delay only makes the problem worse.

Medicare and Medicaid present even graver challenges. I take as given America’s commitment to assist seniors, children, the disabled, and the poor in obtaining a decent level of health care. The primacy of this commitment implies that, although every effort must be made to reduce waste and control costs, ultimately the financial problems of Medicare and Medicaid cannot be solved by denying beneficiaries an adequate level of coverage or by making such deep cuts in provider payments that most doctors and hospitals decline to participate in the programs. Instead, controlling the costs of the government-financed programs will require that we address the costs of health care more broadly, including in the private market. Indeed, federal tax expenditures on private health care (primarily through the exclusion of employer-provided insurance) total more than $150 billion per year (more than 10 percent of federal income tax revenue!).

There is ample evidence that these resources are not being allocated efficiently. On the public side, we see dramatic variation in the spending on Medicare beneficiaries in different parts of the country (even controlling for demographic characteristics), without commensurate differences in health outcomes – or even in beneficiary satisfaction. We see duplicated tests and questionable use of expensive procedures with limited potential health benefits, while at the same time we see underuse of cost-effective procedures generally accepted as “best practices.” On the private side, we see dramatic increases in health insurance premiums (with employer health insurance premiums having grown more than 60% in the last 5 years, at the same time that there has been little increase in real wages) – accompanied by the same overuse of tests and procedures.

There are a number of ways in which health care spending – both public and private – could be made more effective. These potential changes would not only improve the federal fiscal situation (as well as that of the states), but would also create a more cost-effective health-care system – which would have broad social and economic benefits, over and above the positive effects on government budgets. The most important aspect to all of these changes is to generate incentives for consumption of high-value care. Under our current financing structure, most consumers of health care need never evaluate whether the benefits of that care are greater than the costs. While this may make sense for the catastrophic expenditures against which almost everyone would want to insure, such insulation from real costs in decisions about routine care drive up health care expenditures both for private consumers and for public programs.

One of the main reasons that private health care consumers do not adequately evaluate the costs and benefits of the health care that they consume is the bias towards employer provided health insurance built into the tax code. The tax code subsidizes employer provided health insurance not only relative to other forms of compensation, but also relative to individually-purchased insurance and relative to directly-purchased health care. This leads to first-dollar insurance coverage of routine, predictable expenditures, rather than paying for those expenditures out of pocket and using insurance for unexpected and catastrophic expenditures. Research has found that after switching to higher deductible policies, people use substantially less health care, with little ill impact on health, by becoming more careful consumers of their own health care.

The President’s Tax Reform Panel has come up with an interesting recommendation to cap the employer exclusion of health insurance premiums. This cap would increase Treasury revenues (which could be used to correct other inefficiencies in the tax system while maintaining budget neutrality), but would perhaps more importantly limit the bias in the tax code towards overly-generous employer health insurance relative to other forms of compensation. This same tax preference would be extended to non-group insurance purchases, which would help reduce job lock and enhance the market for portable insurance policies. Another possibility would be going one step further and extending the same (capped) preference to other forms of health care spending (such as out-of-pocket expenditures for people with catastrophic insurance coverage), which would reduce the bias towards insurance coverage of routine expenditures.

In the long run this could make health care consumption more efficient in ways that would benefit both public and private consumers. Increased sensitivity to the cost of care would ensure that health resources were allocated to uses with higher value. Furthermore, this consumer-driven health care is likely to increase the competitiveness of health care markets, lowering prices and improving quality. Both Medicare and Medicaid recipients (and federal and state budgets) could benefit from these improvements. Those currently uninsured would gain access to better-functioning insurance markets (especially if insurance premiums were subsidized for those with low-income), which would not only improve their health but would cut back on federal, state, and private spending on uncompensated care. Employees could see more of their compensation in the form of wages, rather than health insurance, and overall performance of the economy could be improved by more efficient allocation of resources.

One pathway to this improved efficiency is through expansions of Health Savings Accounts, or HSAs. HSAs were created as part of the Medicare Modernization Act of 2003. HSAs allow individuals or families to pay for health care with before-tax funds, as long as their health insurance policy includes a sufficiently high deductible ($1,000 for an individual, $2,000 for a family) and catastrophic coverage (a limit on out-of-pocket expenditures). HSAs, which are increasing rapidly in popularity, have many advantages: the funds in an HSA may be rolled over indefinitely, the account is portable between employers, and the greater affordability of high-deductible policies make it easier for small businesses to offer coverage as well as strengthen the non-employer market for insurance.

One option is further expansions of HSAs, which would allow even more people to take advantage of this opportunity, such as adding more flexibility in the catastrophic plans that meet HSA requirements, allowing individuals to use HSA funds for health insurance premiums, and increasing the limits on HSA contributions to cover the full health expenditure risk to which individuals are exposed. These expansions should be accompanied by other measures to improve the quality of information available to patients and the removal of barriers to competition in the health care industry.


America faces daunting fiscal challenges, both in the near term and the long term. In the near term, we face the difficult task of controlling government spending and reducing deficits even as we must find the necessary resources to help victims and to rebuild after the hurricanes, to fight the global war on terror, and to meet our commitments to seniors and to the less advantaged. Economists have long pointed out that a dollar of government spending may cost society a dollar-fifty or more, with the extra fifty cents representing “excess burden”—the output that is lost because higher taxes impair incentives and distort economic decisions. In evaluating existing or proposed government programs, therefore, we should set a tough standard that requires program benefits to be substantial relative to program costs.

In the long term, we must confront the fact that entitlement programs, unless reformed, will take an ever-increasing share of the government budget and of national output. Absent reform, the projected increases in these programs would imply economy-killing tax increases, drastic reductions in non-entitlement spending, or both. The demographic and economic forces that underlie the trends in entitlement spending are powerful, and these issues will take years of concerted attention to address adequately. The time to begin is now.