U.S. fiscal policy entered uncharted waters in responding to the COVID-19 pandemic. In July 2021, the Congressional Budget Office (CBO) projected that the federal deficit would be 13.4 percent of GDP for fiscal year 2021. This followed on the 14.9 percent deficit in 2020, a peacetime record. This bold fiscal response to the pandemic, the largest among industrial nations, helped to avoid what could have been a much sharper-than-observed decline in economic activity during the COVID-19 crisis. It also raised the debt-to-GDP ratio by nearly 30 percent in just two years. From a value of roughly 35 percent in 2007, on the eve of the global financial crisis, this ratio reached a value of just over 100 percent at the end of fiscal year 2021.

Most of the rise in the debt-to-GDP ratio in the last decade and a half can be attributed to the fiscal response to two extraordinary events: the financial crisis and the COVID-19 pandemic. The current structure of U.S. fiscal policy, however, suggests that further debt accumulation will be the norm, and that rather than returning to its historical average level, the debt-to-GDP ratio is likely to increase substantially in coming decades. The latest long-term budget projections from the CBO (2021) show federal revenues falling short of expenditures over the 30 year forecast horizon. The projection based on current law remaining in force suggests an average annual deficit of 9.7 percent of GDP between 2031 and 2051, a dramatic shift relative to the average of 3.3 percent over the last fifty years. The result is a projected debt-to-GDP ratio of 202 percent in 2051, nearly double the level of 2021. The CBO (2020) projected that to stabilize the debt-to-GDP ratio at 100 percent in 2050, the U.S. would need to raise revenues by 2.9 percent of GDP in every year beginning in 2025, or reduce spending by an analogous amount. If the fiscal adjustment did not begin until 2030, the required annual adjustment would be larger: 3.6 percent of GDP per year.

The rapid rise in the debt-to-GDP ratio in the last fifteen years has coincided with a period of falling real interest rates. This is an important consideration when evaluating debt burdens. Current forecasts by market participants, reflected in long-term interest rates, suggest that low rates may be a persistent feature of the U.S. and the global economy. Furman and Summers (2020) point out that the federal government’s net interest payments as a share of GDP have declined in the last 15 years, from 1.7 percent in FY 2006 to 1.4 percent in FY 2021. CBO (2021) reports that the average value over the last fifty years was 2 percent. Going forward, however, the CBO forecasts a substantial increase: 2.4 percent in 2031, 5.2 percent in 2041, and 8.6 percent in 2051. The CBO baseline calls for rising average interest rates on the federal debt. That average interest rate was 4.9 percent in 2007, is 1.4 percent in 2021, and is projected to exceed 3 percent in 2034 and 4 percent in 2043.

There is great uncertainty in any forecast that spans a period of three decades. There are a number of risks, however, that make higher interest payments as a share of GDP, and higher debt-to-GDP ratios, a serious possibility. These include a rise in real interest rates from current levels and the enactment of new federal spending programs that are only partially funded. The U.S. political system has struggled in the last two decades to reign in deficit spending, even during times of robust economic growth. Low real interest rates today reduce the burden of higher debt-to-GDP ratios, but they do not provide a warrant for a fiscal policy that involves projections continuing and accelerating increases in both debt levels and budget deficits in future decades.

The standard macroeconomic analysis of government borrowing suggests that higher debt-to-GDP ratios translate into higher real interest rates, greater interest payments to foreign investors, reduced business investment, and lower consumer investment in durable goods. Gamber and Seliski (2019), after reviewing past research, conclude that a 1 percentage point increase in the debt-to-GDP ratio is associated with between a 2 and 3 basis point increase in the interest rate on 10-year Treasury bonds.

Social Security, more precisely Old Age Survivors and Disability Insurance (OASDI), is an important federal program that, like the overall fiscal balance, displays an imbalance between future outlays and future taxes. This program plays a critical role in the retirement security of many U.S. retirees. Social Security outlays are projected to increase in coming decades, in part reflecting the retirement of the Baby Boom cohort, which is taking place now and will continue for roughly another decade. OASDI is funded by payroll taxes, and the OASDI trust fund has accumulated an excess of taxes over outlays during past decades, when the ratio of workers to program beneficiaries was larger than it is today or is projected to be in the future. In 1980 and 2000, there were more than three workers for every retiree drawing benefits from Social Security. As a result of the decline in U.S. birth rates that began decades ago, that ratio fell to 2.7 in 2020 and is projected to decline further to 2.2 by 2040. The Social Security Administration (2021) currently forecasts that the OASDI trust fund will be exhausted in 2034. At that point, absent any other fiscal action, the payroll tax income accruing to the OASDI system will cover only 78 percent of program costs. In the highly unlikely case that current policy remains unchanged and Congress does not take action to bring revenues and outlays into closer balance, the payments received by beneficiaries in 2035 would fall to 78 percent of their level in 2034, before trust fund exhaustion. Even though most analysts see the risk of across-the-board benefit cuts as very low, the uncertainty created by the prospect of such cuts, and the challenges of addressing the long-term Social Security deficit when there are large non-Social Security deficits, impose costs on current and potential beneficiaries.

Many Americans who have not yet retired doubt that they will receive Social Security benefits at current levels. Parker, Morin, and Horowitz (2019) report that 42 percent of those who are not retired expect their Social Security benefits to fall below current levels, and another 42 percent do not expect to receive any benefits at all. Among those over 50, 48 percent expect reduced benefits, and 28 percent expect to receive nothing. More than half of those between the ages of 30 and 49 do not expect to receive any benefits. These survey results suggest that a lack of understanding of Social Security’s financing contributes to undue pessimism. Even when the Social Security trust fund hits zero, the system will still be able to pay benefits, although only at a fraction of the currently-promised levels.

The projected shortfall of future Social Security revenues is the result of political action, not the absence of policy options for addressing this issue. There are a number of actions that could restore balance to the Social Security program. As with the fiscal policy adjustments that would stabilize the debt-to GDP ratio, however, the size of the disruption for workers and beneficiaries rises as the window between policy change and trust fund exhaustion narrows. The set of potential policy actions includes raising the payroll tax rate on workers, which is currently 12.4 percent, equally divided between employers and employees; increasing the limit on individual earnings that are subject to the payroll tax, currently $142,800, either by selecting a higher cap or by allowing an exempt range and re-introducing the tax on earnings above another level; increasing the normal retirement age, currently on a trajectory to reach 67 in 2027; drawing on general federal revenues to fund any shortfall in the OASDI program; and reducing benefit payouts.

Benefit reductions, whether in the form of a higher retirement age or a cut in the monthly payouts to some beneficiaries, are politically challenging. Some proposals for benefit reduction limit the cutbacks to subsets of the beneficiary population, or phase in the cutbacks gradually. For example, replacing the current full inflation indexation of benefits with partial indexation would disproportionately affect the oldest beneficiaries, for whom the cumulation of modest annual reductions in real benefits would be the greatest. Changing the benefit formula by making the payout structure more progressive would reduce the payouts of those with high lifetime earnings relative to those at lower levels. A combination of policy actions, for example higher payroll taxes and increased benefit progressivity, could spread the burden of restoring sustainability across many current or future beneficiaries.

For a significant subset of the elderly population, Social Security income represents a large share of total income. This underscores the importance of placing the OASDI program on a firm fiscal foundation. Dushi and Trenkamp (2021) use both survey and administrative data to study the share of household income, defined inclusive of withdrawals from defined contribution retirement accounts, that comes from Social Security. In 2015, for men over the age of 65, 37.3 percent received more than half, 18.6 percent received more than three quarters, and 12.1 percent received more than 90 percent of their income from Social Security. The analogous figures for women are 42.0, 23.3, and 15.1 percent. Across-the-board cutbacks in benefits, the default action if trust fund exhaustion is not addressed, would place heavy burdens on the subset of beneficiaries who are highly reliant on this program for retirement support.

There is growing evidence that the uncertainty surrounding the future financing of Social Security exacts a toll on current workers. In a survey of individuals aged 25 to 59, Luttmer and Samwick (2018) found that the average respondent was prepared to forego six percent of their expected benefits under current law if they could eliminate the uncertainty about the program’s future. On average, respondents expected to receive only 59 percent of their current-law benefits. Shoven, Slavov, and Watson (2021) calculate how much a current 45-year-old worker would be prepared to pay to know, with fifteen years of lead time, that benefits were going to be cut, or the retirement age was going to be raised. Illustrating their findings by focusing on women who are currently in the labor force, they conclude that a low-income worker would be prepared to pay just over $5000, a middle-income worker about $12,000, and a high-income worker nearly $15,000 to resolve this uncertainty.

The case of Social Security, where taking action sooner rather than later will reduce uncertainty for beneficiaries and for workers and permit smaller adjustments to taxes and to program parameters to achieve long-term sustainability, carries lessons for the broader U.S. fiscal picture. The higher the debt-to-GDP ratio when a program of fiscal adjustment begins, the larger the changes in taxes or spending must be to achieve a given long-term target. Even if real interest rates remain low for decades to come, and the debt capacity of the U.S. is substantially higher than in past eras, there will come a point at which the path of rising budget deficits like that associated with current forecasts is not sustainable. Recognizing that and taking steps to bring long-term spending and revenue streams into closer alignment sooner rather than later, potentially by considering reforms that could be enacted well before they take effect, will reduce the total cost of making such adjustments.