revised July 20, 2005

by Jason Furman and Robert Greenstein

PDF of full report

Related Statement by Robert Greenstein On The McCrery-DeMint Social Security Proposal

View Additional Social Security Reports

If you cannot access the files through the links, right-click on the underlined text, click "Save Link As," download to your directory, and open the document in Adobe Acrobat Reader.

On July 15, Rep. Jim McCrery and a number of other Republican House members introduced legislation to establish private accounts in Social Security.  The legislation embodies the plan that Rep. McCrery, Rep. Clay Shaw, and other members unveiled several weeks ago and is very similar to the plan that Senator Jim DeMint announced on June 23.  Under the DeMint and McCrery plans, Social Security’s current annual surpluses would be shifted to private accounts rather than used to purchase Treasury bonds for the Social Security Trust Fund.  House Ways and Means Committee Chairman Bill Thomas reportedly plans to use the McCrery plan as the private-accounts component of Social Security legislation he will bring before his committee after the August recess.[1]  This analysis examines both the DeMint and McCrery versions of the plan.

Under both versions, the shifting of revenues to private accounts would end when the Social Security surpluses disappeared.  By Senator DeMint’s own admission, the plan would do nothing to restore solvency to Social Security.  Its purpose instead is to serve as a foot in the door for more extensive private accounts in the future.

For those who are under age 55 today, the DeMint proposal would establish “carve-out” private accounts that would start at 2.2 percent of a worker’s taxable earnings in 2006 and phase down to 0.2 percent of taxable earnings by 2016.  The McCrery plan has identical accounts, although it uses a somewhat different mechanism to fund the accounts, as described below.

Analyses of the DeMint and McCrery plans issued by the Social Security actuaries show that the plans would make Social Security’s financing problems worse were it not for the inclusion in the plans of an assumption that large general revenue transfers would be made from the rest of the budget to Social Security.  The actuaries find that the DeMint plan relies on $422 billion (in net present value, or $1 trillion in today’s dollars) in assumed general revenue transfers, despite the fact that the federal budget is in deficit as far as the eye can see and has no surplus revenue to transfer.  The McCrery plan relies on general revenue transfers of $610 billion (in net present value).  The actuaries also find that the both plans would add to deficits and the federal debt (i.e., the “debt held by the public”) every year for the next 75 years and beyond.

The following analysis is based on the descriptions of the plans that their sponsors have provided and the analyses of the plans that the Social Security actuaries have issued. 

1.   By shifting substantial sums (i.e., the Social Security cash surpluses) from the Social Security trust fund to private accounts, the plans would worsen Social Security’s solvency problems both over the short run and over a longer horizon.  These solvency problems are masked under these plans through budget gimmicks, including the assumption that large revenue transfers will be made from the rest of the budget.  The House version also “double counts” the Social Security surplus revenues. 

i.      Administrative Costs.  Social Security would have to cover the administrative costs associated with the private accounts, which would be substantial.

ii.      New inheritance benefit.  The framers of the plans have said that the accounts would be “inheritable.”  Any such inheritances would, however, reduce the amounts repaid to Social Security through the reductions in the Social Security benefits of workers who had elected the accounts.  That would worsen Social Security solvency, because it would result in Social Security not being compensated in full for the funds shifted from it to the private accounts.

DeMint Sees Plan as Gateway to Free-Lunch Approach that Avoids Hard Choices

Senator DeMint recently revealed that one of the goals of his plan is to create future political pressure for what can be described as a free-lunch approach to Social Security.  In an address in mid-July to two groups that support private accounts, DeMint said of his proposal: “After people get their first statement that there is money in a Social Security account that they own, then we’re going to have the political pressure we need to continue to grow those accounts.”  He also declared:  “I think it’s a mistake to talk about benefit cuts or tax increases until we establish personal accounts, because once these personal accounts are established, I am convinced that we could see our way very quickly on how solvency could be created through just funding these personal accounts.”  He spoke of getting to solvency “without ever going back and cutting benefits, without even raising a tax (emphasis added).*   Indeed, as described below, in 2003, Senator DeMint proposed a free-lunch plan with precisely these characteristics. 

Senator DeMint’s remarks highlight an important aspect of the DeMint and McCrery plans.  They are designed in a way that would open the door to free-lunch plans that rely upon large-scale borrowing and budget gimmickry to restore solvency on paper, without instituting reductions in Social Security benefits or increases in Social Security revenues.  The new DeMint and McCrery plans contain the two core ingredients of such a free-lunch approach: borrowing and general-revenue transfers.

The Social Security actuaries’ analyses show that the transfers in Senator DeMint’s current plan would add three years to Social Security solvency, compared to how long Social Security otherwise would remain solvent under that plan.  The transfers in the McCrery plan would add five years to solvency.  Simply continuing the DeMint/McCrery accounts beyond 2016, enlarging the size of the accounts, and borrowing and transferring funds on a large scale could fully restore solvency, on paper.

To get a sense of what such a free lunch plan could look like, one need look no further than the plan that DeMint introduced in 2003, as a member of the House.  His 2003 plan included large private accounts, and it restored solvency.  The Social Security actuaries’ analysis of that plan showed, however, that it restored solvency by transferring $8 trillion (in net present value) to Social Security over 75 years.  By way of comparison, Social Security’s total 75-year deficit is $4 trillion.  Senator DeMint’s 2003 plan was able to restore solvency and establish private accounts, without either benefit reductions or revenue increases, because it called for general revenue transfers equal in size to twice the Social Security shortfall.

As a result of the very large amounts of borrowing involved, the DeMint plan of 2003 would cause massive increases in the national debt.  Based on the actuaries’ analysis  of the plan, it would by 2050 cause an increase in the debt held by the public equal to 80 percent of the Gross Domestic Product.  The increased interest payments on the debt as a result of the plan would by 2050 equal 4.4 of GDP, a sum that in today’s economy would amount to $540 billion a year in added interest payments.

*Keith Koffler, “DeMint Presses For Accounts As Means To Achieve Solvency,” CongressDailyAM, July 14, 2005.

2.  The plans would substantially increase budget deficits as well as the debt that the federal government owes to outside creditors (i.e., the debt “held by the public”).

3.  The plans would be highly inefficient, as they would require the hiring of thousands of new federal employees and cause the government to incur large new administrative costs to administer millions of mostly modest or small accounts. 

The added administrative costs would amount to at least $25 billion over the first 10 years the plans were in effect and possibly much more than that.

Thousands, and possibly tens of thousands, of new federal employees would be needed to administer the accounts. 

Even though contributions to the accounts would stop being made after only 11 years — because the Social Security surpluses would have run out — the added administrative costs would continue to be incurred every year through at least 2080, since millions of small accounts would remain and would have to be administered until the account-holders died.  For example, a 21-year old individual who began working in 2016 and earned $25,000 that year would be able to contribute a total of $50 that year — about $1 per week — to an account.  No further contributions to the account would be made after 2016, since the surpluses would be gone.  But this worker’s $50 account would have to be tracked, the worker’s investment choices implemented, and regular statements provided to the workers every year for the next half century.  The administrative costs almost certainly would exceed the size of the account several times over.

4.  The plans would reduce traditional Social Security benefits, potentially leaving workers worse off.


Plans Would Not Stop a “Raid” on Social Security or Promote Fiscal Discipline

Supporters of the DeMint and McCrery proposals argue that the Social Security surplus is currently being “raided” by the rest of the budget because the surplus is being used to help cover the deficit.  The claim that Social Security is being raided and that its finances are being weakened as a result is, however, incorrect.

The Social Security trust fund is fully compensated for all of the surplus funds that the rest of the budget borrows.  This compensation is provided in the form of Treasury bonds, which are universally regarded as one of the world’s safest and most secure investments.  Moreover, if the rest of the budget were in surplus rather than deficit, the Treasury would still borrow the Social Security surplus (using it in that case to pay down the federal debt) and give the Social Security Trust Fund the exact same Treasury bonds that it now receives.

There is, of course, a major problem in the non-Social Security portion of the budget.  The deficit outside of Social Security is quite large, principally because of unusually low levels of federal revenue but also due to some spending increases in the last several years.  The DeMint and McCrery proposals do not do anything to reduce non-Social Security expenditures or increase non-Social Security tax revenues.  As a result, they would not alter the need of the federal government to borrow funds to cover its deficits.  If the Treasury borrowed less from Social Security, it would simply have to borrow more in private credit markets.

A specific proposal to reduce the non-Social Security deficit would be welcome.  The DeMint-McCrery approach, however, does not constitute such a proposal.  Some supporters of the DeMint-McCrery approach contend that larger reported budget deficits would lead to spending restraint.  But the DeMint-McCrery approach represents a dubious and highly implausible way of achieving such a goal for three reasons.

First, neither the Administration nor Congressional private-accounts proponents have said that legislation to establish private accounts should include budget cuts or tax increases to offset the increased deficits that the private accounts would engender. 

Second, it is highly implausible that spending reductions of the magnitude required to avoid an increase in deficits under the DeMint and McCrery plans would be enacted.  It is not realistic to expect that federal programs would be cut by $800 billion over the next 10 years (the amount that, with the accompanying savings in interest payments, would offset the cost of the accounts), especially when defense, Medicare, and Social Security are the three largest expenditure areas in the budget and account for the majority of federal expenditures.  In 2004, these three areas accounted for 58 percent of expenditures other than interest payments on the debt.

Third, many of the plans’ supporters claim (incorrectly, as the actuaries’ analyses of the DeMint and McCrery plans show) that the plans would not increase the deficit or that the increases in the deficit that would result would have no effect on the economy.  It is hard to imagine that policymakers would take the politically tough steps to address the resulting increase in budget deficits when they continue to deny that such deficit increases would matter for the economy.  (For a further discussion of this issue, see a companion analysis:  Jason Furman and Robert Greenstein, “Would the DeMint Social Security Plan Cause Policymakers to Become Fiscally Responsible?,” Center on Budget and Policy Priorities, June 24, 2005.)

Indeed, the DeMint and McCrery plans are much more likely to weaken fiscal discipline than to strengthen it, especially since they would open the door to a “free lunch” approach to Social Security, under which solvency would be restored on paper through large-scale borrowing and general-revenue transfers without hard choices being made.  As the box on page 4 explains, Senator DeMint himself has indicated this is one of the goals of his plan.


A Deliberately Slippery Slope to Permanent Private Accounts

For the reasons examined here, it would make little sense to establish accounts into which people could contribute for only 11 years (from 2006 through 2016), especially since administrative costs would continue to be incurred for another 60 years or so.  Yet this anomaly may be one of the reasons the plan’s proponents are advancing it.  Supporters of the plans do little to hide the fact that they view it as a foot in the door.  Their stated goal is to get private accounts in place.  Once in place, supporters of such accounts would argue that it made no sense to halt contributions to the accounts and that the shift of Social Security payroll tax revenues to private accounts needed to be continued even after the Social Security surplus was gone.

Senator DeMint, for example, has said “After people get their first statement that there is money in a Social Security account that they own, then we’re going to have the political pressure we need to continue to grow those accounts.”[11]  DeMint also has declared: “Our whole intention is to come back and add reform on top of this – at least [my intention] – is for larger accounts, a permanent solution.”[12]  At the press conference announcing the plan, Representative Sam Johnson (R-TX) said, “The bill is an achievable first step towards private accounts.”  Other supporters of the DeMint and House plans have made similar statements.


Whether one supports private accounts or not, the DeMint and McCrery proposals fail on the merits.  Their design, structure, and costs make little sense from a policy standpoint.  To the degree that these plans include unpaid-for general revenue transfers, they mask the problems they create and open the door for “free lunch” solvency plans that achieve solvency on paper without making any improvements in the long-term fiscal situation.  The DeMint and McCrery proposals represent a distraction from three important debates:  the debate over whether permanent private accounts funded by payroll tax contributions ought to be established; the debate over what measures should be instituted to restore solvency to Social Security; and the debate over what should be done to reduce the troubling budget deficit outside Social Security both in the near term and especially in the long term.

End Notes:

[1] See, for example, Robert Novak, “The GOP’s Pain-Free Path,” The Washington Post, July 18, 2005.

[2] Specifically, Table 1a of the actuaries’ analysis of the DeMint plan shows that with the transfers, the trust fund would have $799 billion at the end of 2038.  The same table shows that the cumulative transfers would total $1,107 billion in that year.  Without the transfers the trust fund would have been $308 billion in deficit.  See Social Security actuaries, “Estimated Financial Effects of a Proposal to Finance Individual Accounts with the OASDI Cash-Flow Surplus – INFORMATION,” June 23, 2005.

[3] The McCrery bill states its purpose as, “To amend the Social Security Act and the Internal Revenue Code of 1986 to stop the Congress from spending Social Security’s tax revenue surpluses on other Government programs by dedicating those surpluses to personal account.”  The operative legislative language of the bill, however, is at variance with this statement of purpose.  The statement of purpose would not be binding; the operative language of the legislation would be.

[4] Larry Lipman, “Shaw Co-sponsors Social Security Plan Stepping Towards Private Accounts,” Palm Beach Post, June 23, 2005

[5] The $1 trillion in transfers (in today’s dollars) would be made between 2033 and 2036.  (In net present value terms, the transfers would amount to $422.5 billion.)  See the actuaries’ analysis of the DeMint plan, Table 1a.

[6] This is shown in the furthest right column of Table 1a of the actuaries’ analysis of the DeMint plan, which shows “theoretical Social Security” with the plan’s general transfers.

[7] The actuaries project that Social Security currently has a deficit of 1.92 percent of taxable payroll over 75 years.  The DeMint plan claims to reduce the deficit to 1.77 percent of payroll.  But this improvement is more than accounted for by the general revenue transfers.  Without the transfers, the plan would enlarge Social Security’s 75-year deficit to 1.96 percent of taxable payroll.

[8] These deficit figures are for calendar year, rather than fiscal year, 2006.  The use of the Social Security surplus to fund private accounts would begin on January 1, 2006, so calendar year 2006 would represent the first 12-month period during which the plan would be in effect.  These deficit figures are based on the deficit estimates in OMB’s recently released Mid-Session Review and also reflect the assumption that relief from the Alternative Minimum Tax will be continued in 2006.

[9] Estimates are from Table 1b.c and are adjusted to current dollars, as in conventional in 10-year budget estimates.

[10] If, in analyzing the projected returns under these accounts, no adjustment is made to reflect the higher risk involved in investing in the stock market, then anywhere from 50 percent to 90 percent of the accounts of workers with diversified portfolios would be found to be “taxed back” at retirement.  The percentage would be closer to the upper end of this range for workers who are now older and closer to the lower end of the range for younger workers.

[11] Keith Koffler, “DeMint Presses for Accounts As A Means to Achieve Solvency.” CongressDailyAM, July 14, 2005.

[12] Keith Koffler, “DeMint Unveils Bill To Tap Surplus; Grassley Noncommittal,” CongressDailyPM, June 23, 2005.