May 15, 2000

Voluntary Individual Accounts for Social Security: What Are the Costs?
by Peter R. Orszag

View PDF version of report

Proposals for Social Security reform sometimes include voluntary individual accounts, under which individuals can choose to participate fully in the existing Social Security system or to divert some of their payroll contributions into their own individual accounts. For example, the legislation proposed by Senators Moynihan and Kerrey in 1998 (S. 1792) included such voluntary individual accounts. The individual accounts in the Moynihan-Kerrey legislation, along with several other proposals, are "carve out" accounts — they finance individual account contributions by carving out part of existing Social Security revenue and diverting it to these accounts.

The purpose of this paper is to analyze the costs of such carve out voluntary individual accounts. It may initially appear that voluntary accounts must be beneficial, since no one is forced to contribute to them. The reality is more complicated: Voluntary individual accounts involve difficult tradeoffs, and experience with such accounts in other countries has proven troubling.

The paper emphasizes five principal points:

The rest of this paper explores these points in more detail.


I.  Voluntary Individual Accounts, By Themselves, Exacerbate Social Security's Financing Problems

The intermediate estimates from the most recent report of the Social Security Trustees suggest that over the next 75 years, the Social Security system faces an actuarial imbalance of 1.89 percent of taxable payroll. Those estimates also suggest that the Social Security Trust Fund will be exhausted in 2037, at which point tax revenue would be sufficient to pay for about 70 percent of current-law benefits.

To improve Social Security's financial condition, three basic options exist: increase revenue, reduce benefits, or raise the returns on the Social Security Trust Fund. Individual accounts, in and of themselves, do nothing along these three dimensions and therefore do not improve Social Security's financial condition.

Indeed, carve-out individual accounts actually make Social Security's financial condition more adverse. The reason is that they divert revenue away from Social Security, creating a larger actuarial gap. Consider voluntary individual accounts that allowed individuals to divert 2 percentage points of their Social Security contributions into individual accounts. If all workers chose to divert their contribution, the actuarial imbalance within Social Security would rise from 1.89 percent of taxable payroll to 3.89 percent in the absence of other changes. Furthermore, the Trust Fund would be exhausted starting in 2023, rather than 2037.(1)

Only if individual accounts are linked in some way to benefit reductions within the traditional Social Security system, which is likely to be part of any individual accounts proposal (although proponents often do not highlight that feature), could they help to put the system on a sounder footing. For example, Professor Martin Feldstein of Harvard University has proposed that each dollar taken out of an individual account automatically generate a reduction of slightly less than a dollar in current-law Social Security benefits. A linkage of this type is necessary for individual accounts to have the potential to improve Social Security's long-term financial condition.


II. Voluntary Individual Accounts Undermine Progressivity and the Social Compact Behind Social Security

The current Social Security system is progressive: It provides a higher replacement rate to lower-income workers than to higher-income workers. For example, for workers currently in their 30s, Social Security is projected to replace approximately 56 percent of previous wages for a worker with steady low earnings (who is assumed to earn 45 percent of the average wage), 42 percent of previous wages for an average earner, and 28 percent of previous wages for a worker with the maximum taxable earnings each year.(2) (It is worth noting that the progressivity of the Social Security system is somewhat reduced when factors such as the longer life expectancies of higher earners are taken into account.)

Voluntary individual accounts could undermine this progressivity. Higher-income taxpayers would generally have a stronger incentive to partially opt out of the Social Security system than lower-income taxpayers. Since under Social Security's progressive benefit structure, higher-income taxpayers generally earn a lower rate of return on their Social Security contributions than lower-income taxpayers do, the option to place a portion of existing contributions into an individual account is likely to be more attractive to higher-income workers.

The partial withdrawal of higher-income workers from the Social Security system would leave behind a pool of disproportionately lower-income workers. The partial withdrawal of higher-income workers from the system would consequently weaken the system's ability to accomplish redistribution toward such lower-income workers. For example, allowing higher-income workers to divert into individual accounts 2 percentage points of their current Social Security contributions would reduce the revenue available to the Social Security system by billions of dollars a year. It would therefore reduce the resources which Social Security could redistribute to lower-income workers.

Harvard economist David Cutler, a leading scholar on aging populations and their economic implications, has emphasized that contrary to first impressions, the voluntary nature of an individual account system could be harmful. He has written:

"We typically think that giving people choice is optimal since people can decide what is best for them. Thus, the economic bias is to believe that, if people want to opt out of social security, they should be allowed to do so. In the context of social security privatization, however, this is analysis is not right. Allowing people to opt out of social security to avoid adverse redistribution is not efficient; it just destroys what society was trying to accomplish...

An analogy may be helpful. Suppose that contributions to national defense are made voluntary. Probably, few people would choose to contribute; why pay when you can get the public good for free? Realizing this, we make payments for national defense mandatory. The same is true of redistribution. Redistribution is a public good just as much as national defense; no one wants to do it, but everyone benefits from it. As a result, making contributions to redistribution voluntary will be just as bad as making contributions to national defense voluntary. We need to make redistribution mandatory, or no one will pay for it."(3)

In addition to higher-income workers finding voluntary individual accounts more attractive than lower-income workers, younger workers are likely to find the option to make contributions to individual accounts more attractive than older workers.(4) For example, consider two workers earning $25,000 a year. One worker is aged 60 and intends to retire in five years. The other worker is aged 25 and intends to retire in 40 years. Both are given the option to put 2 percent of their wages into an individual account. If the older worker puts 2 percent ($500) of her wages into an individual account and earns 5 percent per year (after inflation) on the balance in the account, her account will accumulate to $638 (in inflation-adjusted dollars) upon retirement. However, if the younger worker puts 2 percent ($500) into an individual account and earns the same rate of return per year as the older worker, the $500 will accumulate to more than $3,500 upon retirement (because interest will compound for a much longer number of years). If both workers would receive $750 more in lifetime Social Security benefits if they did not opt to contribute the $500 to the individual account, the older worker would choose not to contribute to the account (since $638 is less than $750) while the younger worker would choose to do so (since $3,500 is more than $750).

Voluntary individual accounts thus could at least partially undermine the social compact that has supported Social Security throughout the past 65 years. Over time, the Social Security system has become increasingly comprehensive, and it now covers the vast majority of workers in the United States. Voluntary individual accounts would reverse this trend and attenuate the comprehensive nature of the system. They would tend to produce splits between higher-income and lower-income workers, and between younger and older workers.


III. Voluntary Individual Accounts Would Induce Misleading Comparisons of Rates of Return

Early beneficiaries under the Social Security system received extremely high rates of return because they received benefits disproportionate to their contributions. They contributed for only a limited number of years, since much of their working lives had passed before Social Security payroll contributions began to be collected. Indeed, the earliest beneficiaries under Social Security — those born in the 1870s — enjoyed real rates of return approaching 40 percent. To pay for those past extremely high rates of return, all subsequent generations must earn below-market rates of return on their Social Security contributions.

With this in mind, we can see why the type of rate-of-return comparisons that some proponents of individual accounts make are not valid. Imagine a simple pay-as-you-go Social Security system, under which one generation pays $1 while it is young and receives $1 while old. Generation A, which is old in period 1 when the system is implemented, receives $1 in period 1 without having made any contributions itself. That $1 is paid for by Generation B, which is young in period 1. In period 2, Generation B is old and receives $1, paid for by Generation C, which is young in period 2, and so on. The table below presents the operation of the system.

The Simplified Pay-As-You-Go System
























Assume further that the interest rate the market pays is 10 percent per period. Now consider the system from the perspective of Generation C during period 2:

After allowing Generation A to earn an extremely high rate of return, all subsequent generations must earn a rate of return on the pay-as-you-go system that is below the market interest rate. Allowing some members of Generation C to opt out of the pay-as-you-go system thus would appear to produce a higher rate of return, since such members would earn the 10 percent market rate of interest, rather than the zero percent real rate of return on the pay-as-you-go system. But would Generation C really receive a higher rate of return?

If all members of Generation C put $1 into individual accounts during period 2, that $1 could not be used to finance the benefits for Generation B. Yet Generation B's benefits still would have to be paid unless society is willing to allow Generation B to go without benefits.

Assume that Generation B's benefits are financed through government borrowing (or reduced debt reduction) and that the higher interest costs that result are paid for in each period by the older generation. With an interest rate of 10 percent, the interest payments on the $1 in extra debt (relative to a baseline without the individual accounts) would amount to 10 cents per period. Generation C would get $1.10 back from its individual accounts, but it would have to pay 10 cents in interest costs, so it would receive a net benefit of only $1 after paying the interest costs. As a result, the rate of return for Generation C would be zero: this generation would pay $1 and receive a net benefit of $1. Once the interest costs are accounted for, Generation C would earn the same rate of return as the rate of return the pay-as-you-go system provided (e.g., a zero rate of return).

For Generation C and each generation thereafter, the extra benefit from the individual-account system thus is more apparent than real: the extra benefit is exactly offset by the cost of the interest payments on the debt that financed Generation B's benefits. A simple comparison of rates of return of Social Security and individual accounts that ignores this aspect of Social Security financing consequently is misleading.

This basic problem in comparing rates of return and benefit levels between individual accounts and Social Security is illuminated by a recent, important set of papers by economists John Geanakoplos, Olivia Mitchell, and Stephen Zeldes.(5) Geanakoplos, Mitchell, and Zeldes demonstrate that when analytically valid comparisons are undertaken, the supposedly higher rates of return from individual accounts (as compared to Social Security) essentially disappear. They write: "A popular argument suggests that if Social Security were privatized, everyone could earn higher returns. We show that this is false...the net advantages of privatization and diversification are substantially less than popularly perceived."(6) Other economists, including quite conservative ones, have reached essentially the same analytic conclusions as Geanakoplos, Mitchell, and Zeldes. For example, Kevin Murphy and Finis Welch conclude that "many of the touted gains to privatization are more apparent than real, and any gains have more to do with the details of what is done (whether private or public) than with privatization per se."(7) A similar point is emphasized in a new working paper published by the National Bureau of Economic Research.(8)

Now consider a situation in which the individual accounts are voluntary, and not all members of society choose to contribute to them. For example, assume that half of Generation C decides to opt out of the system, and the other half decides to stay in the current system. Half of Generation B's benefits could then be financed by the contributions from the half of Generation C that remains in the Social Security system. The other half would be financed by borrowing. But if the additional borrowing costs were paid by all members of a generation, including those that did not opt out of the Social Security system, those who did opt out would gain at the expense of those who did not. In particular, those who opt out would put $1 into their individual accounts and receive $1.10 in the next period. They would have to pay some interest on the borrowing to pay for Generation B's benefits, but that interest would amount to only 5 cents, rather than 10 cents, because the borrowing would need to cover only half the cost of Generation B's benefits. Those who did opt out would thus receive, after interest, $1.05 ($1.10 minus the five cents they would pay in interest), and earn a rate of return of 5 percent. Those who did not opt out, however, would receive a negative rate of return: They would pay in $1 in period 2 and receive $1 from the Social Security system in period 3. But they also would have to pay 5 cents in interest costs. Those remaining in the Social Security system thus would receive only 95 cents (after interest) in exchange for the $1 they contributed. The likely consequence of this dynamic is that either everyone would opt out of the system or those who were least informed about the system's true implications would be left bearing a disproportionate burden.

Fundamentally, allowing some individuals to escape the financial obligations resulting from the decision made many years ago when Social Security started to cover the initial retirees under the Social Security system would be inequitable. As a society, we decided to provide benefits to those early retirees despite the fact that they had made little if any payroll contributions, which resulted in those retirees' receiving exceptionally high rates of return. As a society, we cannot collectively escape the obligation this decision created. It is inequitable to allow some individuals to escape it partially. We should not allow some members of society — those likely to be best informed and to possess more financial sophistication — to escape that responsibility partially while forcing others to bear the burden to an even larger degree.

Voluntary individual accounts thus would not only allow some members of society to escape their obligations in paying for redistribution of retirement income to less-fortunate members of their own generation but also would enable some members of society to escape their obligation to pay for the redistribution across generations that the Social Security system embodies.


IV. Voluntary Individual Accounts Would Involve Difficult Administrative Issues

Voluntary individual accounts would involve a variety of complicated and difficult administrative issues. Space constraints do not permit a full examination of all the issues. Instead, the following bullets provide an illustrative sample of some of the thorny administrative issues that would be involved:

This discussion of potential implementation issues is not intended to be exhaustive. Rather, it is intended to highlight a small number of such issues. Actual implementation of a voluntary approach to individual accounts would undoubtedly raise a broader array of difficult questions.


V. Experience From Other Countries With Voluntary Individual Accounts, Especially the United Kingdom, Is Not Encouraging

The United Kingdom has had a system of voluntary individual accounts since 1988. Its experience with that system has been particularly unhappy: administrative costs have proven to be surprisingly high, and a massive scandal — the so-called "misselling" scandal — erupted when individuals were given misleading advice about whether they should choose the voluntary individual accounts instead of other Social Security options.

Since it provides the only example of individual accounts among the G-7 countries and since it is broadly similar in culture and general outlook to the United States, the United Kingdom may offer trenchant lessons for the debate here.(11) About one-quarter of workers in the United Kingdom have opted out of the state-run Social Security system and into individual accounts. The government's payroll tax rebate finances contributions into individual accounts that are roughly equivalent to three percent of average annual earnings for American workers covered by the U.S. Social Security system. Roughly half of account holders contribute an additional amount on top of the government rebate. As a result, the contributions deposited in individual accounts in the United Kingdom are at least as large as those being considered for individual account plans in the United States.

The British experience with voluntary individual accounts has been anything but inspiring. In particular:

Given the United Kingdom's poor experience with voluntary individual accounts and the cultural and other similarities between the United Kingdom and the United States, it seems at least somewhat surprising that voluntary accounts could be given serious consideration here.



Voluntary individual accounts do nothing in and of themselves to improve Social Security's financial condition. To the extent that they divert current revenue away from Social Security, they could exacerbate the Social Security shortfall. Individual account contributions equal to 2 percent of taxable payroll could accelerate the year in which payroll tax revenue would no longer be sufficient to pay current-law benefits from 2015 to as early as 2005, and could increase the actuarial imbalance within Social Security from 1.89 percent of taxable payroll to 3.89 percent of taxable payroll.

Furthermore, voluntary individual accounts could attenuate the social compact behind Social Security, by allowing higher-income and younger workers to opt out of part of the system, leaving the rest of society with fewer resources to redistribute toward lower earners and a relatively larger burden to bear in honoring the commitments made to retirees and older workers under the Social Security system. Voluntary accounts also would involve a variety of difficult administrative issues.

Finally, experience from other countries that already have experimented with voluntary accounts is not encouraging. The experience in the United Kingdom should serve as a particularly forceful indicator of the potential problems associated with voluntary individual accounts. The United Kingdom has witnessed a scandal in which vulnerable members of society were given misleading advice regarding the benefits of individual accounts and also has suffered from high administrative costs under its voluntary individual account system.

Policy-makers considering a system of voluntary individual accounts in the United States should carefully examine the potential costs involved. The fact that the accounts are voluntary does not mean they are not potentially harmful.

End Notes:

1. See Peter Orszag, "Impact of 2 Percent Individual Accounts on Social Security Solvency," Center on Budget and Policy Priorities, May 12, 2000.

2. Ways and Means Committee, 1998 Green Book, page 4.

3. David Cutler, "Comment on Gustman and Steinmeier, 'Privatizing Social Security: Effects of a Voluntary System'", in Martin Feldstein, editor, Privatizing Social Security (University of Chicago Press: Chicago, 1998), page 358.

4. The voluntary individual account system in the United Kingdom includes a provision intended to offset this effect: older workers receive a larger contribution for opting out of the Social Security system than younger workers.

5. John Geanakoplos, Olivia S. Mitchell, and Stephen P. Zeldes, "Would a Privatized Social Security System Really Pay a Higher Rate of Return?" in R. Douglas Arnold, Michael J. Graetz, and Alicia H. Munnell, eds., Framing the Social Security Debate: Values, Politics, and Economics (Brookings Institution Press: Washington, 1998), and John Geanakoplos, Olivia Mitchell, and Stephen P. Zeldes, "Social Security Money's Worth," in Olivia S. Mitchell, Robert J. Myers, and Howard Young, Prospects for Social Security Reform (University of Pennsylvania Press: Philadelphia, 1999). A summary of these arguments can be found in Peter R. Orszag, "Individual Accounts and Social Security: Does Social Security Really Provide a Lower Rate of Return?" Center on Budget and Policy Priorities, March 1999.

6. Geanakoplos, Mitchell, and Zeldes, "Social Security Money's Worth," op.cit., pages 2-3.

7. Kevin Murphy and Finis Welch, "Perspectives on the Social Security Crisis and Proposed Solutions," American Economic Review, May 1998, page 142.

8. Hans-Werner Sinn, "Why a Funded Pension System is Useful and Why It is Not Useful," NBER Working Paper 7592, March 2000.

9. John D. Rea and Brian K. Reid, "Trends in the ownership cost of equity mutual funds," Investment Company Institute, Perspective, Volume 4, Number 3, November 1998. It is worth noting that the 1.49 percent per year figure is based on an average account balance of $15,000. Given a two percent contribution rate, it would take years, if not decades, for many workers to accumulate that large an account. In addition, the 1.49 percent per year figure excludes brokerage fees, which could add an additional 0.12 percent per year. See Estelle James, Gary Ferrier, James Smalhout, and Dimitri Vittas, "Mutual Funds and Institutional Investments: What is the Most Efficient Way to Set Up Individual Accounts in a Social Security System?" NBER Conference on Administrative Costs of Individual Accounts, December 4, 1998. On the other hand, it is worth noting that the Investment Company Institute figures indicate that the average cost (excluding brokerage fees) has declined significantly over time: It was 2.45 percent per year in 1981 and declined to 1.49 percent per year in 1997.

10. With a 1.5 percent annual charge on holdings in accounts, a dollar deposited in an individual account in the first year of a 40-year career will be subject to the 1.5 percent fee 40 times, while a dollar deposited in the final year before retirement will be subject to the fee once. On average, dollars in the account will be subject to the 1.5 percent annual charge roughly 20 times, suggesting that approximately 30 percent of the account will be consumed by these charges. The precise figure depends on wage growth and the rate of return, as well as the time profile of contributions and fees over the worker's career.

11. The G-7 countries include the United States, Germany, Japan, the United Kingdom, Italy, Canada, and France.

12. Mamta Murthi, J. Michael Orszag, and Peter R. Orszag, "The Charge Ratio on Individual Accounts: Lessons from the U.K. Experience," Birkbeck College Working Paper 99-2, March 1999 (University of London), available at For a summary, see Peter Orszag, "Administrative Costs in Individual Accounts In The United Kingdom," Center on Budget and Policy Priorities, March 1999, available at

13. See John L. Shuttleworth, "Operating costs of different forms of pension provision in the U.K.," Coopers & Lybrand, June 27, 1997, and John Chapman, "Pension plans made easy," Money Management, November 1998.