You are here

Program Spending Historically Low Outside Social Security and Medicare, Projected to Fall Further

February 24, 2016

Republican leaders in Congress are expected to unveil budget proposals in coming weeks that would reduce projected deficits solely by cutting programs and to justify that approach by claiming that the size and reach of the federal government are growing dramatically.  The data do not support this claim.[1] 

Program Spending as Percent of Gross Domestic Product Under Current Policies
2016 2026
Non-interest spending 18.4% 19.8% 20.1%
Minus Social Security 14.0% 14.9% 14.3%
Minus Social Security
and Medicare
12.1% 11.7% 10.4%

Note: Program spending includes all federal expenditures other than net interest on the debt.

Source: Office of Management and Budget for 1976-2015; CBPP analysis of Congressional Budget Office for 2016 and 2026.

To be sure, total federal spending, including interest payments, rose considerably as a percent of the economy (gross domestic product or GDP) during the Great Recession and remained high in 2010 and 2011.  But it is much lower now than five years ago due to the economic recovery and the expiration of most of the 2009 Recovery Act.  In addition, policymakers have taken steps to reduce the deficit since 2010, primarily through program cuts required by the 2011 Budget Control Act (BCA). 

When only program spending (excluding interest payments on the debt) is considered:

  • Federal program spending outside Social Security and Medicare has fallen below its 40-year historical average and is projected to decline further under current policies.  As Table 1 and Figure 1 indicate, total spending on federal programs outside Social Security and Medicare will be 11.7 percent of GDP in 2016 — below the 40-year average of 12.1 percent — and is projected to decline over the next ten years, to 10.4 percent of GDP in 2026. 
  • The rise in Social Security and Medicare spending over time primarily reflects an aging population and rising health care costs.  Combined spending for these programs is projected to rise from 8.1 percent of GDP in 2016 to 9.8 percent by 2026,[2] compared to an average over the past 40 years of 6.3 percent.  The large majority of this growth reflects the aging of the population as the baby boom generation retires.  In addition, growing per-capita costs in the nation’s health care system (both public and private), stemming in part from new treatments and medications that improve health and save lives but increase costs, continue to put upward pressure on Medicare costs, though health spending growth has slowed recently.
  • Total program spending is expected to change little by the end of the decade, as the rise in Social Security and Medicare is nearly offset by the fall in other program spending.  Overall, program spending is expected to tick upwards, from 19.8 percent of GDP in 2016 to 20.1 percent in 2026.


Figure 1
Non-Interest Spending Outside Medicare and Social Security Set to Fall in Coming Decade


Total federal spending, including interest payments, reached 24.4 percent of GDP in 2009, during the depth of the recession.  It now equals 21.2 percent of GDP.  Over the coming decade, it is expected to rise, reaching 23.1 percent of GDP in 2026.  The vast majority of that 1.9 percentage-point increase from 2016 to 2026 reflects a projected increase in net interest payments, from 1.4 percent of GDP to 3.0 percent.  This increase in interest costs occurs primarily because Treasury interest rates are expected to rise gradually to more normal levels over the decade.

When Americans hear talk of the government growing in size and reach, they usually don’t think this means that more people will receive Social Security and Medicare because the population is growing older, or that Medicare will cost more because of factors like the aging of the baby boomers and improvements in medical technology.  Yet outside of demographic and health cost factors, the portrait of a rapidly growing federal government is at odds with reality.  Indeed, the rest of government — reflecting nearly three-fifths of program spending today, including health reform’s coverage expansions — is projected to shrink relative to the economy under current policies.

Modest Future Program Spending Rise Reflects Demographics and Health Costs

Federal program spending (i.e., spending outside of interest payments) stood at 19.8 percent of GDP in 2016 and is projected to stand at 20.1 percent of GDP in 2026.  While this would exceed the 18.4 percent average over the past 40 years (1976-2015), the increase is relatively modest, considering two key factors:

  • The aging of the population as the baby boomers continue retiring.  The share of the population aged 65 and over rose from 10.4 percent in 1975 to an estimated 14.8 percent in 2016 and will reach 18.6 percent in 2026.  Based on Congressional Budget Office (CBO) data, we calculate that the aging of the population alone will raise Social Security and Medicare spending by more than 1 percentage point of GDP over the next decade — accounting for about three-quarters of the total increase in Social Security and Medicare spending as a percent of GDP between 2016 and 2026.[3] 
  • Rising per-person costs throughout the U.S. health care system.  These costs are much higher today than in earlier decades and are expected to continue rising faster than GDP per person.  While health care cost growth continues to be a major budget challenge, it has slowed noticeably in recent years.[4]  In fact, overall costs for federal health spending — including the costs of health reform’s coverage expansions — are now significantly lower than CBO had earlier projected without those expansions.[5] 

The general spending picture changes substantially once Social Security and Medicare are removed.  At 11.7 percent of GDP, total program spending outside Social Security and Medicare is already below its 40-year average of 12.1 percent of GDP and is projected to fall to 10.4 percent of GDP in 2026.

This decline reflects a substantial drop in annually appropriated, or “discretionary,” programs.  Discretionary spending is expected to fall from 6.5 percent of GDP in 2016 to 5.1 percent of GDP in 2026.  By 2018, these programs will reach their lowest level on record as a percent of GDP, with data going back to 1962.

Discretionary funding is limited by the BCA caps, as further reduced starting in 2013 under “sequestration,” which was triggered by Congress’ failure to enact the additional deficit reduction that the BCA mandated.  Policymakers have provided partial relief from sequestration each year since then, most recently in the fall 2015 budget agreement.  Even so, the BCA caps are pushing discretionary spending to historically low levels.  Despite the recent budget agreement, for instance, discretionary spending will decline as a percent of the economy in both 2016 and 2017 before hitting its historic low in 2018.

Spending on entitlement programs outside Social Security and Medicare is projected to be effectively flat over the next decade at around 5 percent of GDP.  These programs include military and civilian retirement, veterans’ compensation, refundable tax credits such as the Earned Income Tax Credit, and key safety net programs such as SNAP (food stamps).[6]  Importantly, they also include Medicaid and health care subsidies provided under the Affordable Care Act. 

Rising Interest Payments Will Add to Federal Spending

Total federal spending, which also includes interest payments on the debt, will rise over the next decade under current policies from 21.2 percent of GDP in 2016 to 23.1 percent by 2026.  This primarily reflects a projected increase in net interest payments from 1.4 percent of GDP in 2016 to 3.0 percent in 2026. 

About four-fifths of the projected increase over the decade in interest as a percent of GDP comes exclusively from CBO’s assumption that Treasury interest rates will rise to more normal levels as the economy recovers more fully from the recession.  The remaining one-fifth comes from higher debt, reflecting the mismatch between revenues and spending over the coming decade.  It is worth remembering that higher deficits and debt can result as much from tax cuts and revenue shortfalls as from program growth.

Interest is not like federal spending on budget programs.  While policymakers can choose to increase or decrease other spending, they cannot choose to pay more or less interest on the debt.  Nor can they choose the interest rates Treasury must pay.  They can affect interest costs only indirectly, as a byproduct of their decisions changing the level of revenues and program spending.  For these reasons, this analysis focuses on program spending and excludes interest payments.

Nation Still Faces Long-Term Budget Challenge

The bottom line is that under existing policies, federal program spending is expected to grow very little over the next decade as a share of the economy, and the increase is due entirely to the effects of an aging population and rising per-person health costs on Social Security and Medicare.  Outside these two programs, program spending is already historically low as a percent of GDP and will continue declining over the coming decade, despite the impact of rising per-person health care costs on Medicaid, veterans’ and military health programs, and health reform subsidies to help people afford marketplace coverage.

To be sure, total federal program spending will continue growing slowly as a percent of GDP in subsequent decades as the population continues aging and health care costs continue rising.  CBPP’s long-term projections issued in September 2015 show total program spending rising by about 1 percent of GDP from 2026 into the mid-2030s, when the baby-boom retirement is expected to peak, and then leveling off.  Revenues are also projected to rise modestly as a percent of GDP (due to bracket creep) but not by enough to cover program expenditures, so the resulting primary deficits (that is, deficits excluding interest costs) will likely push up the ratio of debt to GDP.  This is the heart of the nation’s long-term fiscal challenge — debt that rises faster than GDP.[7] 

The nation will need to address these issues.[8]  In doing so, policymakers should establish spending and revenue levels somewhat higher than historical 40-year averages to better reflect the circumstances the country faces in future decades (see box).  But the portrait of a rapidly growing federal government is not accurate.

Historical Spending and Revenue Levels Not Adequate Guides for Future

This paper uses historical averages to help provide context for the analysis.  But this approach does not imply that historical levels of spending — and revenue — should serve as a guide for the future, as some commentators urge.  Neither is adequate to meet national needs in coming years. 

As this paper shows, the aging of the population and continuing increases in medical costs will necessitate higher levels of federal spending than over the previous 40 years, even when programs outside Social Security and Medicare are shrinking relative to the economy.  This, in turn, means that higher revenue will be needed to avoid drastic cuts in important programs.

An aging population and higher medical costs already influence spending levels and will do so beyond the ten-year period described in this paper.  In 2035, for instance, the elderly will constitute 21 percent of the U.S. population, up from 15 percent in 2016 and 19 percent in 2026.  And per-person health care costs are rising faster than per-person GDP, partly due to improvements in medical technologies.  This increase in the number of elderly cannot be reversed, and medicine should not be limited to the technologies and treatments of 40 years ago, or even of today.  In combination, these factors alone account for spending that needs to be several percent higher as a share of the economy than in the past.

In addition, interest payments on government debt are expected to rise in coming years, apart from the impact of any further deficits, as today’s exceptionally low interest rates return to more typical levels.

It is worth noting that historical levels of revenues were insufficient to match expenditures even when the population was younger and health care costs were lower.

  • Between 1976 and 2015, federal revenues averaged about 17.4 percent of GDP.  Revenues at this level would not have balanced the budget in any of the last 40 years.
  • The only balanced budgets over this period occurred from 1998 through 2001, when revenues ranged between 19 and 20 percent of GDP, above the 40-year average.
  • As a result of this mismatch between revenues and funding needs, deficits averaged 3.2 percent of GDP between 1976 and 2015.

Meeting national needs in coming decades, while also placing the budget on a sustainable path over the long term, will require spending and revenue at levels above their historical averages.  When considering how to address the nation’s long-term fiscal challenges, policymakers should not use historical averages as a benchmark to guide their efforts.  Rather, they should establish meaningful principles to ensure, for instance, that deficit-reduction policies are designed in ways that do not increase poverty or inequality, shortchange investments important for future economic growth, reduce health care coverage or quality, or impede economic recovery. 


Technical Note

This analysis uses CBPP budget projections over the next ten years (2017-2026), which are largely based on CBO’s January 2016 baseline projections.  As in previous years, CBPP adjusts the CBO projections in a few respects.  Specifically, we assume that:

  • Spending on Overseas Contingency Operations (OCO) will decline starting in 2018, reflecting lower levels of direct U.S. involvement in overseas conflicts such as Iraq and Afghanistan.  This adjustment uses a path from CBO’s August 2015 budget report that assumes a phasedown in OCO funding and applies that path starting with 2018.  (The Bipartisan Budget Act of 2015, which freezes total funding for overseas military operations in 2016 and 2017 at the 2015 level, will have expired by then.)  By contrast, the CBO baseline mechanically assumes that OCO funding will remain at today’s levels, adjusted for inflation, for each of the next ten years.
  • Emergency and disaster spending will reflect its average level for the last five years (2012-2016).  The CBO baseline generally uses the current-year funding level as the basis of its projections for this spending, regardless of whether that funding level is high or low by historical standards.  (Funding for 2016 is well below average.)
  • Funding for appropriated (or “discretionary”) programs will grow with inflation and population after the Budget Control Act caps on appropriated funding expire in 2021.  This approach better recognizes changes in need than the CBO approach, which adjusts these programs only for inflation after 2021.

As Table 2 indicates, the resulting CBPP baseline produces deficit estimates very close to those of CBO’s baseline.  Over the decade ahead, CBPP projects total deficits that are 2 percent below CBO’s.  The deficit projections for 2026 are virtually identical.

Effect on Projected Deficits of CBPP Adjustments (dollars in billions)
  2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 Total
Deficits under CBO baseline 561 572 738 810 893 1,044 1,077 1,089 1,226 1,366 9,378
Phase down overseas operations -1 -13 -29 -41 -48 -52 -54 -56 -57 -58 -409
Use historical emergency and disaster spending 2 6 8 9 10 11 12 12 12 13 95
Grow appropriations w/population           6 15 25 36 48 131
Debt service on adjustments 0 0 -1 -2 -3 -4 -5 -6 -7 -8 -36
Resulting deficits under CBPP baseline 563 566 716 776 852 1,005 1,045 1,064 1,211 1,361 9,159
MEMORANDUM (under CBPP baseline, as a percentage of GDP)
Deficits 2.9% 2.8% 3.4% 3.6% 3.8% 4.3% 4.3% 4.2% 4.6% 4.9%  
Debt held by the public 75.7% 75.7% 76.6% 77.5% 78.4% 79.7% 80.9% 82.1% 83.5% 85.3%  

Note: Components may not add to totals due to rounding.


End Notes

[1] The figures used in this paper are based on CBPP projections, which in most respects reflect the same assumptions as the CBO baseline.  See the technical note at the end of the paper for the differences between the two baselines.  

[2] Figures for Social Security and Medicare spending in this analysis do not include administrative costs, which total less than one-tenth of 1 percent of GDP.

[3] According to CBO, the aging of the population accounts for about 40 percent of Medicare’s growth relative to GDP over the next decade, with the remainder reflecting higher per capita health care costs.  See “The Economic and Budget Outlook:  2016 to 2026,” CBO, January 2016, p. 72,  Only population aging affects projected Social Security spending.  See “The 2015 Long-Term Budget Outlook,” CBO, June 2015, p. 22,  The estimate in the text combines these CBO Medicare and Social Security findings.

[4] Health care cost growth slowed to historically low annual rates during the recession and recovery.  According to the latest annual National Health Expenditure projections from the Office of the Actuary at the Centers for Medicare and Medicaid Services, health spending growth will tick up slightly in the next few years but will remain well below pre-recession growth rates.  See Paul N. Van de Water, “National Health Spending Growth Remains on Slow Track,” Center on Budget and Policy Priorities, December 2, 2015,

[5] Richard Kogan, Paul N. Van de Water, and Cecile Murray, “CBPP’s Projections Show Long-Term Budget Outlook Has Improved Significantly Since 2010 But Remains Challenging,” Center on Budget and Policy Priorities, September 14, 2015,

[6] For instance, SNAP spending rose considerably during the Great Recession, leveled off in 2012 and 2013 as a percent of GDP, and started declining in 2014.  CBO projects SNAP will decline further as a percent of GDP over the next decade, returning to its 1995 level by 2020.  For a full analysis of these trends, see Dottie Rosenbaum and Brynne Keith-Jennings, “SNAP Costs and Caseloads Declining, Trends Expected to Continue,” Center on Budget and Policy Priorities, updated February 10, 2016,

[7] Kogan, Van de Water, and Murray.

[8] As discussed in Kogan, Van de Water, and Murray, for instance, assuring long-run solvency for the Social Security and Medicare Hospital Insurance trust funds would substantially improve the long-run budget picture.