Fiscal Outlook: Federal Fiscal Outlook
GAO's federal budget simulations provide a broad context for considering policy options. An understanding of fiscal exposures—programs that may expose the government to future spending—can also inform these considerations.
The federal government’s debt as a share of gross domestic product (GDP) A commonly used measure of domestic national income. GDP is the value of all goods and services produced within the United States in a given year and is conceptually equivalent to incomes earned in production. It is a rough indicator of the economic earnings base from which the government draws its revenues. is growing unsustainably because revenue and spending are fundamentally imbalanced over the long term, according to GAO’s simulations of the federal fiscal outlook.
The increases in spending shown in the simulations are driven by an aging population, rising health care costs, and net interest Net interest is equal to the amount of interest paid minus the amount received. The government pays and collects interest in various ways. For example, net interest outlays are dominated by the interest paid to holders of federal debt securities that the Department of the Treasury issues to the public. on the federal debt.
Further, the growing gap between revenue and spending will limit the federal government’s flexibility to address future challenges.
These key points are illustrated by GAO’s long-term simulations projecting federal deficitsThe amount by which the government's spending exceeds its revenues for a given period, usually a fiscal year. and debt under different sets of policy assumptions. A simulation is a hypothetical—a "what if?" Since 1992, GAO has periodically run two simulations of the federal budget that illustrate the potential implications of different policy choices.
GAO also publishes simulations of long-term fiscal trends for the state and local government sector, which likewise faces long-term fiscal pressures.
What do GAO's simulations show about the budget's sustainability?
Both simulations show debt as a share of GDP growing continuously, though the timing and pace of growth varies depending on the assumptions used. In 2016, the debt-to-GDP ratio was 77 percent, exceeding the historical average (44 percent since 1946). The simulations show debt held by the public surpassing its historical high (106 percent in 1946) within 15 to 25 years (see figure). By 2090, the last year of GAO’s 75-year projection period, the debt-to-GDP ratio reaches 280 percent in the Baseline Extended simulationThe Baseline Extended simulation begins with a baseline using Congressional Budget Office (CBO) estimates and generally assumes current law continues into the future, such as the expiration of tax provisions as scheduled. or 468 percent according to the Alternative simulation The Alternative simulation changes some of the baseline assumptions to reflect historical trends rather than current law. For example, tax provisions that are scheduled to expire, such as the credit for construction of energy-efficient new homes, are extended to 2025, and revenue is held constant as a share of GDP thereafter..
Source: GAO. | GAO-17-237SP
Notes: Data are from GAO's 2017 simulations. See the Assumptions and Data tab for a description of the assumptions underlying this simulation. Both of GAO’s simulations assume that Social Security and Disability benefits are paid in full regardless of the current projections of revenues into the Old-Age, Survivors, and Disability Insurance trust funds.
Data: txt | pdf
The federal government will not be able to sustain these budget paths over the long term. Although U.S. government securities remain attractive investments throughout the world, growth in debt as a share of GDP can limit the government's flexibility to address emerging budget issues and as-yet unforeseen challenges, such as another economic downturn or a large-scale disaster.
Debt is growing because revenue and spending are fundamentally imbalanced. Even though in the near term the deficit is relatively small, both simulations show it growing over time. Indeed, under revenue assumptions for both simulations (both simulations hold debt constant in the long term at the 2026 share of GDP – 18.5 percent for the Baseline Extended and 18.4 percent for the Alternative) the gap between revenues and spending gets bigger over time. By 2090, the government’s deficit under the Baseline Extended simulation would be about 15 percent of GDP and about 26 percent of GDP under the Alternative simulation (see figure). This trajectory is not a prediction or a forecast; it is an illustration of the need to think about the budget path over both a near-term and a longer-term horizon.
Source: GAO. | GAO-17-237SP
Notes: Data are from GAO's 2017 Alternative simulation. The graphs on this website generally reflect the Alternative simulation. Additional details on the two simulations are provided in the Assumptions and Data tab and in this document.
Data: txt | pdf
What needs to be done to address the problem?
The government must act soon to change the long-term fiscal path or risk significant disruption to individuals and the economy. Policymakers will need to discuss the entire range of federal activities and spending—entitlement programs, other mandatory spending, discretionary spending, and revenue. Moving forward, the federal government will need to make tough choices in setting priorities and ensuring that spending leads to positive results.
One way to quantify the magnitude of the policy changes needed to address the increasing debt is by calculating the fiscal gap. The fiscal gap represents the difference—or gap—between revenue and programmatic spending (i.e., spending other than interest payments) that would need to be closed immediately and permanently in order to hold revenue constant as a share of GDP beginning in 2016 for the 75-year period. The size of the fiscal gap in 2016 under the Baseline Extended simulation is 3.2 percent of GDP; under the Alternative it is 6.1 percent of GDP.
Closing the gap requires spending reductions, increases in revenue, or, more likely, a combination of the two. To illustrate this, one can calculate what it would take to have debt held by the public as a share of GDP in 2090 equal to what it was at the beginning of 2016 under GAO’s Alternative simulation:
To close the gap solely by cutting spending would require reductions in programmatic spending of about 25 percent, and maintaining that lower level of spending, on average, each year over the next 75 years.
- To close the gap solely by raising revenues would require a revenue increase of about 33 percent, and maintaining that level of revenue, on average, each year over the next 75 years.
The figures below illustrate how the change required at the beginning of 2016 compares with the current size and composition of spending and revenues. If action is delayed, then more significant changes would be needed. GAO's Baseline simulation shows similar results, with somewhat smaller changes needed to close the fiscal gap.
Source: GAO and GAO analysis of Congressional Budget Office data.
Notes: The colored pie on the left represents noninterest spending in 2015 (the most recent year of data available). The red slice in the pie on the right side is the fiscal gap: the percent of spending that would need to be cut on average this year and every year over the projection period to prevent debt from growing. Data on percent reduction needed to close the fiscal gap are from GAO’s 2017 Alternative simulation. See the Assumptions and Data tab for a description of the assumptions underlying this simulation.
The "Other Discretionary" category includes spending on education, training, employment, and social services; transportation; international affairs; the administration of justice; natural resources and the environment; general science, space, and technology; agriculture; energy; and other discretionary spending.
The "Other Mandatory" category includes spending on unemployment compensation, federal civilian and military retirement, the earned income tax credit, the Supplemental Nutrition Assistance Program, and other mandatory programs, minus income from offsetting receipts.
The "Health Care" category includes spending on Medicare, Medicaid, and other health programs.
Data: txt | pdf
Source: GAO and GAO analysis of Congressional Budget Office data.
Notes: The colored pie on the left represents revenue in 2015 (the most recent year of data available). The green slice in the pie on the right side is the fiscal gap: the percent of extra revenue that would be needed on average this year and every year of the projection period to prevent debt from growing.
Data on percent increase needed to close the fiscal gap are from GAO's 2017 Alternative simulation. See the Assumptions and Data tab for a description of the assumptions underlying this simulation. For more information on GAO’s work on the federal tax gap, see our key issues page on this topic.
"Other Receipts" includes excise taxes, Federal Reserve earnings, Customs duties, estate and gift taxes, and miscellaneous fees and fines.
Data: txt | pdf
What are the key drivers of the growing debt?
Growth in deficits and debt is driven by the gap between revenue and spending; both sides of the equation are relevant. On the spending side, both the Alternative and Baseline Extended simulations show net interest on the federal debt and spending on health care increasing over time (see figures).
Source: GAO. | GAO-17-237SP
Note: Data are from GAO's 2017 Alternative simulation. The Alternative simulation assumes that health care cost containment mechanisms, including those enacted in the Patient Protection and Affordable Care Act, are not sustained over the long term (the Baseline Extended simulation assumes that they reduce health care costs over the long term).
Data: txt | pdf
Revenue as a percentage of GDP can vary for a number of reasons, including changes in tax policy and the economy. For illustrative purposes, both of GAO’s simulations hold revenue constant at its 2026 share of GDP over the rest of the 75-year period (18.5 percent in the Baseline Extended, and 18.4 percent in the Alternative); they show that revenue will cover a decreasing percentage of spending over time.
Net interest is a function of the amount of debt to be financed and the interest rate the Department of the Treasury must pay on that borrowing. This in turn adds to the debt. While interest rates are currently at historic lows, both of GAO’s simulations assume interest rates will rise in the future. This increase, combined with growing debt on which interest is charged, makes net interest the largest and fastest-growing spending category in the long term. As debt grows, so will interest spending, which will in turn further drive the growth in debt. The “miracle of compound interest”—meaning that interest will accrue both on the original amount of debt as well as interest charges on that debt—will worsen the outlook.
National health expenditures – comprising Medicare, Medicaid, exchange subsidies, and the Children’s Health Insurance Program (CHIP) – roughly double as a share of GDP over the course of the projection period. This is due primarily to two factors:
1. An aging population: Americans are living longer than in the past. The proportion of the population that is age 65 and older is growing, with particular speed in the near future while the baby boom generation is reaching retirement age (see figure). The oldest members of the baby boom generation are already eligible for both Social Security retirement benefits and Medicare. Additionally, increasing longevity will mean that the percentage of individuals who are over 85 and 95 years are rising. These trends will contribute to increasing health care and other costs.
2. Rising per capita health care costs: Health care costs have historically risen faster than GDP and are expected to continue to do so. The extent to which costs per person exceeds the growth rate of GDP per person—adjusted to separate out the effects of an aging population as described above—is referred to as excess cost growth. According to the Congressional Budget Office (CBO), annual excess cost growth averaged around 2 percent from 1975 to 2014 (the most recent year for which data are available). Going forward the Medicare Trustees and CBO assume that excess cost growth will not disappear, but will slow down due to the financial pressure health care spending is putting on the federal government and others. How and when this slowdown takes place, however, is highly uncertain. (For more information on health care cost growth and its effects on the long-term federal budget outlook, see GAO's report on this topic.) GAO’s sensitivity analyses, under the “Assumptions & Data” tab, show how varying assumptions about excess cost growth affects the long-term outlook.
What do GAO's simulations show for other categories of spending?
As with spending on health care costs, spending for Social Security benefits is also driven by demographics. As the number of beneficiaries increases with the retirement of the baby boom generation, spending on Social Security grows to represent a sizable share of spending, but then remains relatively constant as a share of GDP over the long term. GAO’s website on Social Security includes more information on the challenges facing the program.
Although Social Security spending and the drivers described above are not directly tied to changes in other categories of spending, they do have an impact on them. Spending on Social Security, health care, and net interest absorbs a growing share of GDP and of any given share of revenue. This in turn can "crowd out" or put pressure on spending on other mandatory programs (veterans' disability benefits, agriculture, student loans) and on discretionary spending (national defense, transportation, education).
What else is potentially driving federal spending?
The federal fiscal outlook also faces risks from other fiscal exposures—responsibilities, programs, and activities that may either legally commit the federal government or create the expectation for future federal spending. Examples are responses to natural disasters, pension guarantees, financial crises, and ensuring care for veterans. Some of these fiscal exposures have increased over the past decade due to external events, trends and the government’s response to them. Increased attention to fiscal exposures will be important both for understanding those risks and enhancing oversight of federal resources.
What data does GAO use in its simulations?
GAO uses data from the Congressional Budget Office (CBO) for most of its projections through 2026. In the long term, GAO uses data from the Medicare and Social Security Trustees (Trustees) and the Office of the Actuary at the Centers for Medicare & Medicaid Services (CMS Actuary).
- For the baseline simulation, GAO generally follows CBO’s baseline for the first 10 years and in the longer term the simulation is based on the Trustees' intermediate assumptions for Social Security and current law assumptions for Medicare.
- For the alternative simulation, in the longer term, the simulation is based on the Trustees' intermediate assumptions for Social Security and the CMS Actuary’s illustrative alternative assumptions for Medicare.
GAO also compares its simulations to CBO's long-term projections to identify any deviation due to differences in assumptions for Social Security and major health entitlements. The results are consistent with GAO's simulations based largely on the Trustees' projections.
This document provides more detailed information on the data and assumptions used in GAO’s simulations.
How do changes in selected assumptions affect GAO's simulations?
All simulations are dependent on the assumptions used. For example, the longer-term outlook depends heavily on assumptions about the growth in health care spending for each beneficiary. To show how GAO’s simulations would change with different assumptions, GAO conducts sensitivity analyses showing outcomes under different scenarios for four key factors. Under all of these sensitivity analyses, the federal fiscal path is unsustainable. This provides further evidence that policy changes will be necessary to change the path of the gap between revenue and spending over the long term.
- Excess cost growth for health care. Health care cost growth several decades into the future is uncertain. Given that health care spending is one of the key drivers of the growing debt, future growth of health care costs significantly affects the long-term fiscal outlook.
- Interest rates. As explained in the “Key Drivers” tab, interest on debt held by the public will be a driver of the growing debt. While interest rates are currently at historic lows, CBO and the Trustees project they will rise significantly in the future, causing interest payments on federal debt as a percentage of GDP to rise.
- Discretionary spending. Caps on discretionary spending enacted in the Budget Control Act of 2011 and revised by subsequent legislation would bring discretionary spending as a share of GDP to historic lows by 2021. It is uncertain whether Congress and the President will adjust these caps in the near term, as they did in the Bipartisan Budget Act of 2015, which modified discretionary appropriations limits for fiscal years 2016 and 2017. The outlook for discretionary spending is even more uncertain over the long term.
- Revenue. Revenue as a percentage of GDP can vary with changes in tax policy and the economy. For example, in December 2015 Congress and the President enacted legislation that temporarily extended, made permanent, and in some cases modified a number of expired or expiring tax provisions, such as the Research Tax Credit, the Child Tax Credit, and the Earned Income Tax Credit. Changes such as these affect the federal government’s fiscal outlook. It is impossible to know what policy changes will be enacted in the future. At the same time, revenues can rise and fall with changes in the economy. Revenue fell to its lowest share of GDP in the past 50 years in 2009 after the recession, but, as the economy has recovered, it has recently risen to slightly above the 50-year historical average.
The ability of policymakers to control or influence these factors varies. Congress and the President enact laws that directly control discretionary spending through the appropriations process; legislated changes in eligibility, benefits or other parts of program design will change mandatory spending. Net interest, however, is not directly controlled through legislation. As noted, it is a function of the amount of debt to be financed and the interest rate, which is determined largely through economic factors and monetary policy.
Does GAO provide more detailed data on its simulations?
Data for each set of simulations are available through the following links:
Risk Sources for Fiscal Exposures
Americans expect a range of services and benefits from government for many reasons. For example, they expect that when they retire, the government will keep its commitment to provide them with Social Security and Medicare benefits. Similarly, Americans may expect that when natural disasters strike, the government will provide funding for emergency response and long-term recovery as it has in the past. In budgeting, such expectations are referred to as fiscal exposures—responsibilities, programs, and activities that may legally commit or create these expectations for future federal spending based on current policy, past practices, or other factors. The magnitude of this future spending is difficult to quantify and is not fully reflected in the budget. Nonetheless, fiscal exposures are significant, pervasive and pose a high risk to the American taxpayer.
Fiscal exposures fall along a spectrum of the government's legal commitment. In addressing a fiscal exposure, it is important to consider the source of the exposure. Fiscal exposures can be organized by five sources of risk: demographics and health care, economic downturns, environment and disasters, government operations, and security. These categories are not mutually exclusive and are not comprehensive of the government’s full range of fiscal exposures.
GAO-17-237SP: Published: Jan 17, 2017. Publicly Released: Jan 17, 2017.
GAO-16-622: Published: Jul 7, 2016. Publicly Released: Jul 7, 2016.
GAO-14-28: Published: Oct 29, 2013. Publicly Released: Oct 29, 2013.
GAO-13-167SP: Published: Nov 29, 2012. Publicly Released: Jan 8, 2013.
GAO-11-626T: Published: May 4, 2011. Publicly Released: May 4, 2011.
GAO-08-206: Published: Dec 20, 2007. Publicly Released: Dec 20, 2007.
GAO-07-1155SP: Published: Sep 7, 2007. Publicly Released: Sep 7, 2007.
GAO-06-276: Published: Jan 31, 2006. Publicly Released: Jan 31, 2006.
GAO-05-1009SP: Published: Sep 1, 2005. Publicly Released: Sep 1, 2005.
GAO-05-193SP: Published: May 2, 2005. Publicly Released: May 2, 2005.