Report | Budget Taxes and Public Investment

Unbalanced budgeting: Federal spending cap may endanger Social Security

Policy Memo #186

 Download PDF

Share this page:

  • Supplemental table: Cut to Social Security outlays under CAP Act, by state and congressional district [PDF] [Excel]

Congress is considering a range of policies to limit, or “cap,” overall federal spending, either as stand-alone legislation, or as part of a balanced budget amendment. Supporters argue that these efforts are necessary to reduce the deficit and to impose limits on current and future Congresses.

One such proposal is the Commitment to Prosperity (CAP Act),  sponsored by Senators Bob Corker (R-Tenn.) and Claire McCaskill (D-Mo.), along with 14 other senators. The CAP Act (S. 245) would create hard caps on government spending, to be phased in over 10 years (2013–2022). In 2022 and subsequent years, the Act would cap spending at  20.6% of gross domestic product, based on the bill’s formula and assumptions.

This policy memorandum shows the possible impact of overall spending caps on the elderly through reductions in Social Security payments and on communities through decreases in federal outlays. It analyzes the impact nationally and within states and congressional districts. The memorandum also briefly reviews the even more dire cuts to Social Security under a proposed balanced budget amendment.

The cap specified by Corker and McCaskill would likely lead to severe reductions in projected spending on Social Security, Medicare, Medicaid, and other government programs. Cuts of the magnitude proposed could not only significantly impact overall program spending but could also cut individual benefit levels, and harm employment growth.

Specifically, the spending caps as defined in the CAP Act, assuming levels are achieved through across-the-board proportionate decreases in spending, could lead to the following outcomes:

  • $882 billion in cuts to Social Security outlays over 2013-2021. In 2021, Social Security outlays would be cut 13.6% relative to what they would have been absent the act.
  • Significantly greater cuts over the long-term. By 2025 Social Security would be cut by between 21% and 36%, depending upon baseline assumptions on cost growth; by 2045, outlays would be cut by between 47% and 83.5%.
  • Deep cuts to benefits in response to lower spending levels. In 2017, we could expect the average Social Security  benefit to drop by 10.1%, an almost $2,000 cut in benefits for a scaled medium earner; by 2036, the year the Social Security Trust Fund is expected to expire, the benefit cut could grow to a 37.5% cut, or $9,300 for this earner. For comparison, a cut of that size is a much deeper benefit cut than the estimated 23% reduction in benefits that would occur if no policy changes were made to close Social Security’s long-range funding gap.
  • Unnecessary cuts. Reductions of the magnitude required to meet the annual spending caps would leave Social Security overfunded over the long-term, with Social Security tax contributions exceeding benefits.
  • Lost jobs. Employment levels could suffer as a result of cuts. In 2013, the shock to aggregate demand from these cuts could lead to around over 250,000 jobs lost. In 2014, these cuts could lead to around 470,000 jobs lost.
  • Widespread impact on local economies. States and congressional districts would lose spending power with cuts to outlays—somewhere between $17 million and $124 million in 2013 (depending on their size and their demographic composition), with older populations losing proportionally more. (See the supplemental tables posted online for a breakdown of cuts under the global spending cap by state and congressional district, and the appendix to this paper for cuts by gender, race, and ethnicity.)

Though less likely to pass Congress, the cap provisions in balanced budget amendments proposed as part of the debate over raising the federal debt limit could also pose a threat to Social Security. Specifically, a balanced budget cap as currently defined in the proposals could:

  • Cut Social Security outlays by $1.8 trillion from 2016 to 2020.
  • Cut the Social Security program 32.7% in 2016, the first year the amendment, growing to 35.8% in 2020.

About the CAP Act

The CAP Act sets a nominal cap on federal spending at 25% of GDP in 2013. This ratchets down to 23.63% of GDP by 2021, and 23.46% by 2022. When applied to GDP levels in 2022, the effective cap on outlays (i.e., the real cap when the difference between past and future spending is accounted for) is 20.6% of GDP.1 Proponents of the CAP Act are highlighting this 2022 capped level of spending—20.6% of GDP—because it represents average spending between 1970–2008. (OMB 2011).2

Under the CAP Act, projected spending would not be allowed to exceed permissible spending levels (as specified by the effective cap levels) in any given year. Policymakers could make spending cuts to help reach these levels so that automatic spending cuts (“sequestration”3) would not kick in. They could, for instance, choose to reduce all program spending by the same percent (i.e., imposing across-the-board cuts). Or they could cut spending on, for example, health care and education, while increasing spending in other areas. They could choose modest (or zero) reductions in some areas and more severe reductions in others. This lack of specificity in the CAP Act is also a perceived political strength of the act—policymakers today need not specify particular cuts.

Across-the-board cuts made to avoid sequestration would cut overall federal spending (excluding any spending done to pay our interest) by 13.6% in 2021, relative to projected spending under the laws and current policy trends in place.4 (See EPI Briefing Paper #315 “Why Spending Caps Are Poor Policy,” for a detailed discussion of the effects a CAP Act would have on spending.)

In 2021, the cut would be equivalent to 2.8% of GDP and amount to $656 billion in that year alone. From 2013 to 2021 the total cut to spending excluding spending on interest would be around $3.5 trillion.

Social Security trends

Barring legislative changes that would reduce benefits, spending in the Social Security program is and will continue to automatically increase as the population ages and baby boomers retire. Social Security outlays are expected to grow from 4.8% of GDP today to 6.2% of GDP in 2035 before settling at or below 6% through 2085. This is a relatively modest increase given the projected increase in retirees and the decline in the birth rate in the generations following the baby boomers (CBO 2010).5

As the baby boomers retire, increased costs to the program will in part be paid for by drawing down on the Social Security Trust Fund, the invested assets of which are expected to be worth $2.7 trillion by the end of 2011 (Office of the Chief Actuary 2011a). We have this surplus in part due to planning by policymakers for the retirement of the baby boom generation. The 1983 Social Security reforms that increased the retirement age, the payroll tax, and the payroll tax cap all helped build the trust fund we have today. Furthermore, because Social Security is prohibited by law from borrowing, it is already subject to an effective cap of its own. The program can only spend what it has saved, plus interest and what it takes in as revenue in a given year.

CAP Act could cut Social Security spending and jobs in the short-term

This analysis assesses the potential impact of the CAP Act on Social Security under two scenarios. The first assumes that Congress cuts all programs  by an equal percent, to achieve the capped level of permissible spending in each year. (Policymakers could choose to spare Social Security, but would then have to make more severe cuts in other areas. Alternatively, if other programs were maintained, the cuts to Social security would be greater. This analysis assumes neither of these options are chosen.)

The second scenario assumes that Congress does not come to an agreement for spending cuts, and therefore reductions are achieved through the CAP Act’s proposed enforcement mechanism.

It should be noted that Congress could cut Social Security more than estimated under these scenarios to make room for other priorities.

Cuts made across the board by Congress

If outlays were cut on an across-the-board basis in order to comply with the spending cap, Social Security would not fare well. As shown in Table 1, in 2017, five years after the cuts go into effect, Social Security would be reduced by 10.1%, a cut of nearly $101 billion relative to baseline outlay levels. By 2021, that cut would grow to $172 billion, or 13.6%. Total cuts from 2013 to 2021 would equal $882 billion.6

As a result, Social Security recipients would likely suffer from severe cuts to benefits. Because outlays in 2017 would be 10.1% lower than they would otherwise be, benefit levels could presumably fall by 10.1% as well. In 2017, the average annual benefit for a scaled medium earner7 who retired at the normal retirement age would be $19,255 (SSA 2010).8 If cut by 10.1%, this benefit level would fall to $17,309; a loss of just under $2,000 in one year. This calculation assumes an across-the-board cut to all beneficiaries, including retired workers and their spouses, survivors, and disabled workers.

The pain from these cuts would be spread across the country. States and congressional districts, particularly those with older populations, would lose spending power as beneficiaries receive less. For example, congressional districts would lose anywhere from $17 million to $124 million in 2013. The supplemental table posted online to accompany this report shows how across-the-board cuts needed to comply with CAP Act spending levels in 2013–2017 would affect beneficiaries in every state and congressional district, including the District of Columbia, American Samoa, Guam, the Northern Mariana Islands, Puerto Rico and the Virgin Islands. Additionally, the appendix includes tables showing the breakdown of cuts over 2013–2017 by gender and race and ethnicity (SSA 2011c).

If policymakers chose to exempt certain areas of the budget from CAP Act-mandated reductions, cuts to Social Security would be even greater. For instance, if security spending (including spending by the Department of Defense) were exempted from an across-the-board cut, Social Security would be cut by $131 billion, or 13.2%, in 2017. By 2021, that cut would grow to $220 billion or 0.9% of GDP, and it would be a 17.3% cut relative to baseline levels.

Cuts to Social Security outlays, as with cuts to other types of government spending, would result in job losses. The immediate shock to aggregate demand from a $23 billion cut to Social Security outlays would result in just over 250,000 jobs lost in 2013. In 2014, cuts of $45 billion would result in nearly 470,000 jobs lost.9 Our estimates use a standard macroeconomic model that is consistent with that used by private- and public-sector forecasters.

Cuts made through CAP Act enforcement mechanism

If Congress failed to enact cuts to reach permissible spending levels, cuts triggered by sequestration under the automatic method specified by the CAP Act would fall disproportionately on the fastest-growing categories of spending. The CAP Act divides primary outlays into three categories: mandatory spending (which includes Social Security), non-defense discretionary spending, and defense discretionary spending. Under this scheme, Social Security is treated as a part of the unified budget (which current law forbids). The budget cut percentage applied to each of the categories would be proportional to the category’s share of total expenditure growth between one year and the next. Because mandatory spending is expected to grow faster than discretionary spending, the discretionary spending categories would absorb less of the spending cut, leaving the mandatory programs to bear the brunt of cuts. Under this process, Social Security would see a $128 billion hit, or a 12.9% cut from baseline levels, by 2017. By 2021, Social Security outlays under the spending cap would be cut by $208 billion, a 16.4% cut from baseline levels.

As with across-the-board cuts by Congress, the cuts made under the CAP Act enforcement mechanism would have an employment impact. In 2013, a cut of $37 billion would negatively impact GDP by 0.3%, and result in around 407,000 jobs lost. In 2014, a cut of $72 billion would lead to job losses of just over 750,000.

Long-term effects of CAP Act on Social Security would be even more severe

The effects of the CAP Act on Social Security outlays, benefit levels, and jobs would grow significantly over the long term. Table 2 shows estimated cuts in program outlays under two long-term scenarios: our analysis as applied to CBO’s more conservative extended baseline scenario, which depicts what we would see with the laws currently in place, and our analysis as applied to CBO’s alternative fiscal scenario,10 which is their version of adjusted current policy. Although both CBO baselines are problematic in some ways,11 they allow us to provide both low-end and high-end estimates of the long-term potential impact of the CAP Act on Social Security outlays without having to endorse either baseline scenario. We assume in these cases that spending cuts are done in equal percentages across the board.

As a result of the CAP Act, if all spending were cut equally, by 2025 we could expect reductions to Social Security outlays of between 21.1% and 36.0%. By 2045, those cuts would grow to between 46.9% and 83.5%.

Benefit levels would also plummet, even more so than over the short-term. The Social Security Trust Fund is expected to be depleted in 2036. That year, the CAP Act would mandate a 37.5% cut to baseline Social Security outlays, under our conservative extended baseline scenario. This is a significantly larger cut than the program would experience if no policy changes were made between now and 2036. We can thus assume that since outlays in 2036 are 37.5% lower than they otherwise would be, benefit levels would be cut by the same amount. In 2036, the average annual benefit for a scaled medium earner who retires at the normal retirement age is projected to be $24,796, in constant 2010 dollars. If cut by 37.5%, this benefit would fall to $15,504, a cut of just over $9,000. As with the short-term outlays estimates, this calculation assumes an across-the-board cut to all beneficiaries, including retired workers and their spouses, survivors, and disabled workers.

Other proposed spending caps could also hurt beneficiaries and workers

Besides the CAP Act, both the Senate and the House have floated balanced budget amendments (BBA), which include spending caps, as a potential addition to a vote on the debt ceiling.

One Senate version, S.J.Res.10, states that the president must submit a balanced budget to Congress and that Congress must pass a balanced budget with limited exceptions. Introduced by Sen. Orrin Hatch (R-Utah), it also says that spending must be capped at 18% of prior-year GDP.12 Government spending has not been at 18% of prior-year GDP since 1957 (OMB 2011). Additionally, this BBA would require a supermajority vote in Congress to either raise the debt ceiling or pass any new tax increases.

The cap in this Senate balanced budget amendment would force even greater cuts to spending than the CAP Act. Holding spending to 18% of prior-year GDP would cut total primary outlays in 2016 (the year the BBA would go into effect were it ratified immediately) by $1.25 trillion, or 6.5% of GDP, relative to our baseline levels. In the first five years, from 2016 to 2020, this 18% spending cap would result in $7 trillion in cuts to total outlays relative to the adjusted current policy baseline.

The legislative text of this bill offers no specifics on how the 18% cap should or would be enforced, leaving the hard decisions to policymakers of the future.13 Assuming the cuts necessary to hit spending levels of 18% of GDP were distributed evenly across the board, outlays for Social Security would take a big hit, as shown in Table 3. From 2016 to 2020, Social Security would see $1.8 trillion in cuts. In the first year the amendment would take effect, the program would be cut 32.7%. That cut would grow along with Social Security outlays if the cut were administered proportionally; by 2020 program outlays would be cut 35.8%.

The House version of the BBA (H.J. Res. 1), spearheaded by Rep. Bob Goodlatte (R-Va.), has much in common with the Senate version, though it caps spending based on GDP of the current fiscal year, as opposed to prior to the fiscal year (as the Senate bill does). The House BBA would limit total federal spending to 18% of GDP and require a two-thirds supermajority vote in Congress to waive this limit. Because this legislation caps spending (excluding interest payments) at 18% of GDP for the current fiscal year, budgets would be drafted based on projected GDP. This would politicize the CBO’s economic projections and also force deep cuts if those projections were too high, for instance at the onset of a recession. This could force cuts at a time when cuts would only further harm our economic health.


A  CAP Act or a Balanced Budget Amendment would pose enormous risks to Social Security. Under the CAP Act, if spending reductions were made across the board or according to the CAP Act enforcement, current and future beneficiaries would be faced with significant cuts to their benefits.

Capping future spending at historical levels ignores the realistic costs of our future obligations, which include an aging population, the growing cost of providing health care, and increased security spending. Caps also constrain our ability to plan for such things as demographic changes.

From a policy perspective, Social Security has been the nation’s most effective anti-poverty program, and cutting it would deepen poverty—particularly among the elderly and disabled. Social Security recipients have been contributing to the system for their entire working lives, and should receive promised benefits.


  1. The CAP Act sets a nominal cap on spending based on what is called “lookback GDP,” which is the average GDP of the first three of the four preceding years. “Lookback GDP” is used in the Corker-McCaskill bill to allow for cyclical changes to be smoothed out over a period of time, to leave a buffer since GDP numbers are not final until three full months after a year has concluded, and to allow for a permissible spending level to be available well before the start of the fiscal year so that policymakers have enough information to deliberate their spending decisions.
  2. When the nominal cap is applied to the lookback GDP and then imposed upon the fiscal year ahead, it creates an effective cap that is lower than the nominal cap. For example, under the CAP Act, the nominal cap would be 25% of GDP in 2013, but the effective cap on spending in 2013 would actually be 22.5% of GDP. This effective cap is slated under the CAP Act to be ratcheted down, eventually reaching an effective cap of 20.6% of GDP by 2022 (this of course would likely change as they base this number off of GDP estimates through 2022). In order to arrive at their estimate of an effective cap of 20.6% of GDP, the CAP Act authors assume that GDP in 2022 will have grown along with the average growth rate of GDP over the previous 10 years. The spending cap would remain at 20.6% of GDP from 2022 on.
  3. Sequestration refers to the cancellation of budgetary resources through an enforcement mechanism.
  4. Analysis in this paper is done relative to an adjusted current policy baseline. This adjusted baseline is based on projected spending as written into existing law, called a “current law baseline,” and makes adjustments that reflect the more realistic policy assumptions that we can make. Our adjusted current policy baseline starts with the CBO March current law baseline for outlays, then adjusts these outlays by adopting the Overseas Contingency Operation (OCO) path in the President’s Budget, as well as a permanent extension of the “doc fix” (maintaining Medicare payment rates for physicians at 2011 levels).  It then adds in the extra interest costs of permanently indexing the Alternative Minimum Tax (AMT) and extending all the Bush tax cuts (which push up spending by nearly $800 billion through 2021).
  5. All analysis was done for the nine-year time period of 2013-2021. Insufficient data were available on the budget picture through 2022 to accurately portray the effect of the CAP Act in this year.
  6. Although the demographic changes are real and will require Social Security spending increases, this does not mean that the system is in crisis and that benefit cuts are necessary. Most of the cost of the Baby Boomer retirement will be covered by savings in the trust fund, and an increase in revenue equal to 0.7% of GDP would be sufficient to put the program in long-term balance. The relatively modest shortfall could be closed simply by eliminating the cap on covered earnings, for example (currently, earnings above $106,800 are subject to Medicare but not Social Security tax).
  7. EPI estimates total Social Security cuts of $882 billion between 2013–2021, which is less than estimates made by the Center on Budget and Policy Priorities (CBPP). CBPP estimates $904 billion in cuts over same period (Park, Ruffing, and Van de Water 2011). The discrepancy is due to differing baseline assumptions, particularly when it comes to the interest effects of projected spending levels.
  8. The Social Security Administration uses hypothetical earnings histories to project a range of benefit levels. Thus, SSA’s “scaled medium earner” is a hypothetical worker with earnings throughout his or her career equal to 100% of the average wage of workers the same age. However, because actual workers have more uneven work histories, including years spent outside the labor force, 56% will have average indexed monthly earnings below those of the hypothetical medium earner and will therefore receive lower benefits than those projected for the scaled medium earner (Clingman and Burkhalter 2011). In addition, most workers retire before either age 65 or the normal retirement age (which is currently 66), and will therefore receive reduced benefits compared to Social Security projections.
  9. This is in constant 2010 dollars.
  10. This analysis employs Congressional Budget Office assumptions of a fiscal multiplier of 1.5 for direct government spending. From there, we calculate that a 1% drop in GDP corresponds with 1.2 million jobs lost.
  11. CBO’s alternative fiscal scenario makes a number of assumptions. These include a “doc fix”, that is, that Medicare physicians payments do not continue to grow with the SGR mechanism, as well as AMT relief and the extension of the Bush tax cuts. Additionally, it assumes that Medicaid and insurance exchange price controls are not implemented beyond 2020, and that other noninterest spending (excluding Medicare, Medicaid, insurance exchange subsidies, CHIP, Social Security, and some refundable tax credits) would be frozen at 2010 levels as a share of GDP (excluding stimulus spending) beyond 2013. Finally this scenario assumes that revenues remain constant as a share of GDP beyond 2020.
  12. Analyzing the long-term effects of the CAP Act on Social Security requires a number of assumptions, the most significant of which is determining which baseline to use. CBO’s alternative fiscal scenario (offered in the Long Term Budget Outlook), is a somewhat plausible baseline to use in our analysis, particularly because Corker and McCaskill use assumptions from that baseline in calculating the effects of the CAP Act from 2013-2022, and have estimated savings relative to that baseline. That said, the baseline cannot be viewed as a realistic budgetary picture for the future. The main problem with using this baseline is the revenue estimates baked into long-term outlook. The baseline assumes revenue levels will remain at 19.3% of GDP from 2020 to 2084, which is the end of the long-run scenario. This is not only unrealistic, but has the effect of increasing interest costs due to the ever-widening gap this scenario assumes between spending and revenue levels. Deficits thus become unrealistically large, averaging 19% of GDP over 2030–2050, and skewing our results. However, when attempting to calculate the effect of the CAP on Social Security outlays in the long run, a baseline must be chosen. This analysis thus provides numbers based on both of CBO’s long-term baseline scenarios, the extended baseline scenario, which is equivalent to current law, and the alternative fiscal scenario, which is CBO’s version of adjusted current policy. In providing both sets of numbers, we provide what we feel are both low-end and high-end estimates of the potential impact of the CAP Act on Social Security outlays, without having to endorse either baseline scenario.
  13. A balanced budget refers to a budget where outlays do not exceed revenues. This particular measure stipulates that for Congress to pass a budget with a deficit, it would need the support of two-thirds of both houses, even during economic downturns. In times of declared war, it would still take a majority vote in Congress to pass a budget running a deficit.
  14. Note, the text states that “total outlays for any fiscal year shall not exceed 18 percent of the gross domestic product of the United States for the calendar year ending before the beginning of such fiscal year.” Though vague, what this actually means is that the cut for any given fiscal year would be calculated off of adjusted calendar year GDP from the previous year. When that cut is actually measured against GDP for the immediate prior year, it results in an average effective spending cap that is actually lower than 18% of GDP. For the first five years that such a policy would be in effect, the average cut would actually be 16.6%. See “An 18% spending cap is not just bad policy, it’s simply not feasible” (Fieldhouse 2011) for additional explanation on this.


Bivens, Josh. 2011. “Abandoning What Works (and most other things, too).” Economic Policy Institute, Briefing Paper #304, Washington, D.C.: EPI;

Clingman, Michael and Kyle Burkhalter. 2010. “Scaled Factors for Hypothetical Earnings Examples Under The 2010 Trustees Report Assumptions.” Social Security Administration Actuarial Note, No. 2010.3.

Congressional Budget Office (CBO). 2011. “Budget and Economic Outlook: Fiscal Years 2011 through 2021.” Washington, D.C.: CBO;

Congressional Budget Office (CBO). 2011b. “An Analysis of the President’s Budgetary Proposals for Fiscal Year 2012.” Washington, D.C.: CBO;

Congressional Budget Office (CBO). 2010. Long-Term Budget Outlook, Washington, D.C.: CBO.

Fieldhouse, Andrew. 2011. “An 18% spending cap is not just bad policy, it’s simply not feasible,” Economic Policy Institute commentary, March 31;

Goss, Stephen C. 2010. “The Future Financial Status of the Social Security Program.” Social Security Bulletin 70(3).

Office of Management and Budget. 2011. Historical Tables, Table 1.1 and Table 1.2;

Office of Senator Bob Corker. 2011. “Commitment to American Prosperity (CAP) Act.” Washington, D.C.: United States Senate, February 2.

Park, Edwin, Kathy Ruffing, and Paul Van de Water. 2011. “Proposed Cap on Federal Spending Would Force Deep Cuts in Medicare, Medicaid, and Social Security.” Washington, D.C.: Center on Budget and Policy Priorities;

Social Security Administration (SSA) Office of the Chief Actuary. 2011a. “The 2011 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds,” (The 2011 OASDI Trustees Report);

Social Security Administration (SSA) Office of the Chief Actuary. 2011b. Single-Year Tables Consistent with 2011 OASDI Trustees Report;

Social Security Administration (SSA) Office of the Chief Actuary, 2011c. “Annual Statistical Supplement to the Social Security Bulletin, 2010,”

Social Security Administration (SSA) Office of the Chief Actuary. 2010. “2010 Trustees Report,“ Table VI.F10—Annual Scheduled Benefit Amounts a for Retired Workers With Various Pre-Retirement Earnings Patterns Based on Intermediate Assumptions, Calendar Years 1940-2085;

United States House of Representatives. 2011. Proposing an amendment to the Constitution of the United States relative to balancing the budget, H.J. Res. 10, 112th Cong., p. 2;

See related work on Budget Taxes and Public Investment | Retirement

See more work by Rebecca Thiess