Promoting freedom, innovation, and growth

Connect with IPI

Receive news, research, and updates

July 13, 2005

Stop the Raid, Start the Accounts


The announcement of a new Social Security reform initiative to be introduced jointly in the House and the Senate revolutionizes the debate over Social Security. The proposal would stop the practice of raiding the annual Social Security surplus to finance other government spending and use that money instead to start personal retirement accounts.

The proposal focuses on starting the personal accounts alone, without any additional provisions for tax increases, future benefit cuts or delays in the retirement age. This wisely keeps the spotlight on the personal accounts and their many appealing features. Now we can have a debate about personal accounts and not about other reform ideas the public overwhelming opposes. Despite the opposition of the Left, the public will broadly support personal accounts when the issue is framed this way. That has already been shown by the many elections over the past three election cycles that have been won on personal accounts.

The accounts would not be financed by diverting payroll tax revenues, which eliminates the biggest focus of attack on personal accounts so far. The accounts would be financed, in part, by Federal bonds and, as we argue below, should also be financed further through general revenues. It does not matter where the money for the personal accounts comes from. What matters is that benefits from the accounts substitute for a proportional share of benefits coming from the old Social Security framework.

With that substitution, the accounts will start to reduce the long term Social Security deficits and help to achieve solvency. Indeed, payroll tax revenues paid into the program will remain the same, but the long term obligations of the program will be reduced as the accounts take over some of the responsibility for paying future benefits.

This benefit substitution formula, however, should not be too harsh. It should not take away the better deal that is available to workers through personal accounts because market investment returns are so much higher than what the Social Security transfer payment system promises (let alone what it can pay). If the substitution formula takes away a dollar in Social Security benefits for every dollar that comes out of the accounts, then workers will not gain from the accounts. The formula should be designed so that workers will be able to keep the net gain from higher market returns, and, therefore, get a much better deal through the accounts.

That would be accomplished through a proportional formula that substitutes personal account benefits for Social Security to the degree that the worker exercised the personal account option over his career. For example, if the worker exercised about half of the full account option over his career (calculated on a present discounted value basis) he would still receive half of the promised benefits from the current Social Security system with the other half replaced by the personal account benefits. 1 This should leave workers with a large net gain. As we understand it, the Members behind this reform plan have now agreed to this.

We will advance in this study a few features that should be included in this still not fully defined reform effort which, even with a generous substitution formula that allows workers to gain greatly, would, we believe, lead to full solvency for Social Security if the accounts were extended beyond the first 10 years. We believe, indeed, that the score of the proposal by the Chief Actuary of Social Security would confirm these results. We will also argue for strong provisions that definitely do “stop the raid” on the trust funds.

Stop the Raid
The current Federal government practice of raiding the Social Security trust funds each year to finance other government spending is despised by the American public. Definitively stopping that practice would consequently receive broad and deep public support.

To stop that practice completely, the legislation should provide for an appropriation each year to the accounts from general revenues equal to the cash surplus in Social Security, the surplus of taxes over revenues. That cash surplus now flows directly into general revenues, where it is spent on other programs. To prevent that, the surplus would be sent instead to the accounts.

The Federal government should then reduce its future spending by the amount of this cash surplus. Sponsors of the legislation should not be shy about including this policy goal as part of the reform, and stating that they plan to achieve such future spending restraint. But the legislation would not attempt to disable Congress from spending whatever it chooses in the future. If it wants to keep its spending growth going unchecked, then it is still free to do so. But it would not be able to use the Social Security surplus to do so, in accordance with the public’s wishes. It would have to find other financing for such spending, such as raising taxes or increasing the deficit and Federal borrowing. In the latter case, the Social Security surpluses would no longer mask the true Federal deficit and levels of spending and borrowing.

Hopefully, these more accurate revelations and strong pressure for the policy goal of reducing spending growth (now that the Social Security surpluses are no longer available for spending) will lead to reductions in future spending levels. This would mean that the transition to the accounts would be financed in part by such spending restraint, which would increase support for the reform from some quarters.

The media has already begun questioning whether the new Social Security reform proposal really does “stop the raid.” For members to say they are stopping the raid but to not actually do so in the legislation would be cynical and hypocritical. To ensure that the legislation wins the powerful public support for stopping the raid, the provisions of the legislation must actually stop the raid.

Some have suggested that general revenues should not be used for personal account reform. But since Social Security is a pay-as-you-go system with no savings for future benefits, the transition to personal accounts cannot succeed without general revenues to help finance the transition. As informed critics on the Left have rightly pointed out, trying to adopt personal accounts without general revenues for the transition would require immediate benefit reductions equal to the size of the accounts. All personal account reformers should know this.

Trust Fund Interest
There is more to the annual Social Security surplus than just the cash surplus of taxes over required expenditures. Each year the Social Security trust fund also receives interest on the Federal bonds it holds. This interest is paid in the form of additional Federal bonds newly issued to the trust funds each year. These bonds should be issued to the personal accounts instead, so that the total annual Social Security surplus will be used to start the accounts.

The total surplus including these bonds is currently projected by the Chief Actuary of Social Security as sufficient to finance an account over the first 10 years of the reform equal on average to 3.8 percent of wages. (See Table 1) With this projected financing, the legislation can and should specifically provide for accounts of this magnitude for 10 years. Congress would then come back and address extending the accounts past those first 10 years into the future, and specify the transition financing that would apply after the first 10 years.

Table 1

Putting the Trust Fund Surplus into Personal Accounts

Year Cash Surplus (excluding interest)
(billions of 2005 dollars)
Percent Payroll Surplus as% of Taxable Payroll Total Surplus (including interest)
(billions of 2005 dollars)
Percent Payroll Cash Plus Scheduled Trust Fund Interest
2006 84.8 1.7% 183.6 3.8%
2007 88.7 1.8% 194.7 3.9%
2008 90.2 1.8% 204.3 4.0%
2009 84 1.7% 206.3 4.0%
2010 80.2 1.5% 210.9 4.0%
2011 75.6 1.4% 215.1 4.0%
2012 65.3 1.2% 213.2 3.9%
2013 52.9 1.0% 209 3.7%
2014 38.5 0.7% 202.5 3.6%
2015 24.3 0.5% 196.3 3.4%
2016 8.2 0.2% 188 3.2%
Source: 2005 Social Security Trustees Report and authors’ calculations

Just offering all workers the same 3.8 percent would mean that higher income workers, who get a worse deal from Social Security today, would gain much more from the accounts than low income workers, which would be politically intractable.

Therefore, the specified account must be progressive to ensure that lower income workers gain as much as higher income workers. That would be achieved by specifying that the account would equal 5% of the first $10,000 of wages each year, and 2.5 percent of taxable wages above that, which would amount to an average account of 3.2 percent of taxable wages—slightly less than projected surpluses would allow, providing for a cushion in case surpluses are less than projected.

Despite what any critics or anyone else may say, using these bonds for the personal accounts does not involve any new debt for the Federal government. These bonds are going to be issued to the Social Security trust funds under the current system anyway. Those who say they do involve new debt are really saying the trust fund bonds are phony and we do not intend to make good on them. Let’s see if any members of Congress take that position.

Achieving Full Solvency
Adding the interest due to the trust fund into the personal accounts would have a much greater effect over the long run in reducing future Social Security spending obligations, with the accounts providing much higher benefits in their place. If Congress extends the 3.2 percent accounts permanently past the first 10 years, with payroll tax revenues still coming into Social Security in full, we believe the long term Social Security deficits would be eliminated entirely as a result, and permanent solvency for Social Security will be fully achieved. We believe a thorough score of the proposal by the Chief Actuary of Social Security would confirm these results.

Extending the accounts past the first 10 years would require additional transition financing for that period. That required financing would be much less than the unfunded liability of Social Security eliminated by the reform. There are several possible sources of such transition financing (which again would be included only in a second bill extending the 3.2 percent indefinitely).

Ideally, the restraint on total Federal spending proposed in the Ryan-Sununu legislation would be adopted and start to run even before the accounts are extended past the first 10 years. This modest spending restraint would hold Federal spending to grow just one percentage point less than GDP for eight years. Federal spending relative to GDP was restrained much more during the 8 years of the Clinton Administration.

This provision would reduce Federal spending relative to GDP by only 1.6 percentage points. CBO projects that under current law Federal spending will grow by 14 percent of GDP by 2050, from 20 percent today to 34 percent. So much greater spending restraint than this will have to be adopted to prevent that result. Those who want to argue that even the modest restraint of Ryan-Sununu is unrealistic need to start discussing their ideas for almost doubling Federal taxes relative to GDP over the next 40 years.

Another source of transition funding is the increased Federal revenues that would result due to increased investment resulting from the accounts and higher economic growth due to the reform. Much has been written regarding the substantial increase in economic growth that would be expected to result from personal accounts. 2 Part of the future transition can be financed by issuing additional Federal bonds, which would then be paid off after the transition is complete. Congress could also adopt a net tax increase to help finance the longer term transition if it desires, though we would not favor that approach.

Over the longer run, the goal would be to eventually adopt a full personal account option equivalent to the employee share of the Social Security payroll tax, as proposed in Ryan-Sununu and other proposals.

With the changes discussed above, using the total Social Security surplus to start a 3.2 percent account option would ultimately lead to full solvency for Social Security while providing higher (not lower) retirement benefits for workers through the personal accounts. Workers would also enjoy personal ownership and control over the accounts, and would be free to leave some or all of the account funds to their families. With such a plan providing a clearly better deal for workers, reformers could then successfully take their case to the people, and either win reform now, or make their opponents suffer at the polls.


  1. In this legislation, the full option would be the full original option of 6.4% proposed by Rep. Paul Ryan and Sen. John Sununu. Under this approach, a worker who exercises the 3.2% option proposed for starters in the new bill over his entire career would forego half of promised future Social Security retirement benefits in return for the account benefits. We would not want the worker to give up all retirement benefits under the current system with a 3.2% account alone, because that account is not large enough to replace all of those benefits.
  2. See, e.g., Martin Feldstein, “The Missing Piece in Policy Analysis: Social Security Reform”, American Economic Review, May 1996; Martin Feldstein, “Privatizing Social Security: The $10 Trillion Opportunity”, Cato Institute Social Security Paper no. 7, January 31, 1997; Martin Feldstein, “Toward A Reform of Social Security”, The Public Interest, Summer, 1975; Peter Ferrara and Michael Tanner, A New Deal for Social Security , Washington, DC: Cato Institute, 1998, Chapter 6.


  • TaxBytes-New

Copyright Institute for Policy Innovation 2017. All Rights Reserved Privacy Policy Contact IPI.

e-resources e-resources