Michael Tanner’s Testimony on Social Security and Personal Accounts


Mr. Chairman, distinguished Members of the Committee:

I would like to thank the Committee for holding this hearing andfor giving me the opportunity to appear. In particular, Iappreciate that the Committee is moving beyond the standoff overwhether Social Security reform should or should not includepersonal accounts to consider how such accounts might be structuredin ways that can maximize consumer choice and control, whileensuring efficiency, low costs, and preserving an appropriatemeasure of worker protection.

Of course, along with my colleagues at the Cato Institute, Ibelieve that Social Security reform must allow younger workers tosave and invest some of their Social Security taxes throughpersonal accounts. I believe that such accounts can significantlycontribute to restoring Social Security to permanent sustainablesolvency. More importantly, I believe that personal accounts areessential to modernizing Social Security in keeping with suchfundamental American values as ownership, inheritability, andchoice.

In particular, regarding the subject of this hearing, economictheory holds that private capital investment should provide ahigher rate of return than a mature PAYGO Social Security system.If one accepts the Social Security Administration’s assumptionsabout future bond and stock returns, a balanced portfolio (50percent stocks, 30 percent corporate bonds, and 20 percentgovernment bonds) could be expected to yield a return of 4.9percent. Subtracting 25 basis points of administrative costsprovides a net yield of 4.65 percent. Shifting the mix slightly infavor of equities should raise the expected return to roughly 5percent. This clearly exceeds the return available from SocialSecurity, and also significantly exceeds the offset interest ratesuggested under the president’s reform proposal.

This is not to say that personal accounts can perform miracles.They cannot, by themselves, solve Social Security’s entire $12.8trillion funding shortfall. However, workers who choose thepersonal account option‐​and I note that personal accounts arevoluntary under all the major reform proposals‐​can expect toreceive more in retirement benefits than Social Security canactually pay them.

That said, how personal accounts are structured and theinvestment options available to workers can make a significantdifference in the success of any personal account proposal. Inshort, details matter.

Any retirement system has four important administrativefunctions: collection, transmission, record keeping, and moneymanagement. First, there must be a system to collect the retirementfunds from the worker. Next, the funds must be transmitted to anadministrator. The administrator is responsible for keeping recordsof each worker’s contribution to the retirement program and thebenefits that each worker will eventually receive. Finally, themoney has to be invested and managed between the time it isreceived and the time it is disbursed.

In designing an investment and administrative structure ofpersonal accounts, you should be guided by these basicconcerns:

  • Simplicity and Transparency. Workers shouldclearly understand where there money is going and what theiroptions are. Where personal account plans have encountereddifficulties, such as in Britain, it has been primarily do tooverly opaque or complex schemes.
  • Balancing Return and Risk. While marketreturns can be expected to exceed Social Security returns, marketsare not risk free. In particular, they offer increased volatil​i​ty​.In addition, many of the new investors brought into the marketthrough personal accounts will be inexperienced. A personal accountinvestment plan must offer these individuals some degree ofprotection without stifling consumer choice, over regulatingmarkets, or unduly restricting the potential for positivereturns.
  • Keep Administrative Costs Low. Whileregulation of account fees would be unwise, accounts should bedesigned in ways that minimize administrative costs. SSA estimatesthat accounts would cost 25 – 30 basis points to administer. Thisseems like an entirely reasonable target.
  • Limit government involvement in investmentdecisions. Decisions about the investment of theaccumulating retirement funds should be left to private markets andinsulated from government interference as much as possible.
  • Avoid increased employer burden. Every effortshould be made to avoid any new burden on employers, particularlysmall employers. Possible sources of additional employer burdenswould include higher tax payments, greater complexity in taxcalculations, more extensive record‐​keeping requirements, or arequirement to report information more frequently.

Cato’s Project on Social Security Choice has devotedconsiderable time and study to these issues. As a result of ourwork, we believe that the following structure provides theframework for meeting these concerns. I do not, by any means,assert that it is the only acceptable administrative structure. ButI believe that any workable administrative structure that dealswith these issues will contain many or most of the followingelements.

First, at least initially, the collection, record‐​keeping, andtransmission functions should be handled centrally. The U.S.Treasury Department through its existing tax collectioncapabilities is well suited for this roll. Therefore, I wouldpropose that the collection of payroll taxes, including individualaccount contributions, would continue to be handled by the employerin much the same way as today and sent to the Treasury as they aretoday. The only difference would be that the employer would tellTreasury how much of the total payment is from employees who havechosen the personal retirement account option.

One little known facet of the current Social Security system isthat although payroll taxes are collected and paid by employersthroughout the year, the federal government will not actually knowhow much money was paid on behalf of any particular worker in 2005until about September 2006! This makes little difference under thecurrent Social Security system, but can matter a great deal with aninvestment‐​based system. There will have to be some mechanism tohold the funds until the contribution is reconciled to theindividual’s name using the worker’s W‑2 form. Moreover, thisholding pool should not be dependent on market timing. The bestsolution would be for Treasury, which is already operating as acentralized collection to transfer all funds designated for accountinvestment to a private‐​sector custodian bank, which then investsthe total amount in a money market fund that is always priced atone dollar, a standard industry convention. The following year,when the contribution is reconciled to the individual worker thefund’s shares are distributed to each worker representing hiscontributions and interest credit, and electronically transferredto the worker’s personal account as specified.

Second, because a system of personal accounts would extendinvestment opportunities to millions of Americans who do not nowparticipate in private investment, and are therefore likely to lackeducation and experience in choosing investments, a consensus hasdeveloped among account proponents that initial investment optionsshould be limited both in number and in the amount of risk a workermay assume. Therefore, most personal account plans call forinvestment options to be initially limited to a small number ofbroadly diversified funds. This can be done in a number of ways,such as:

  • A small number of index funds each composed of a different typeof investments. For example, the Federal Thrift Savings Plan, whichPresident Bush has cited as an example, offers a fixed income fund,a common stock fund, a “small cap” stock fund, a government bondfund, and an international stock fund. The Chairman’s proposal, S540, is also built around this model.
  • A small number of balanced funds, each composed of a differentmix of stocks, government bonds, corporate bonds, and cash. Thisoption is included in HR 530 among others.
  • A lifecycle fund that automatically adjusts the mixture ofinvestments as a person ages. Younger workers would be more heavilyinvested in stocks, with the mix changing more heavily to fixedincome assets as the worker nears retirement. The president hasraised the possibility of this option, as has the Chairman. And Ibelieve it is included in S 857, sponsored by Sen. Sununu.

To make things even simpler for the unsophisticated or apatheticworker, there should be a default option that would require theworker to make no decisions whatsoever.

Management of funds should be handled by the private sector on acontract/​bid basis, similar to the way the TSP is currentlyhandled. Given the amount of investments involved, it may be worthconsidering breaking up the management into several pieces each ofwhich would be put up for bid.

Finally. while it makes sense to limit investment optionsinitially, at some point a wider range of choices should be madeavailable. In part, this is a simple matter of increasing consumerchoice. One of the most important reason for having personalaccounts at all is to give workers more choice and control over howthey save for their retirement. Clearly, this should be extended asmuch as possible to the array of available investments.

There is also a larger economic reason for eventually expandingthe range of investments. The flow of investments to differentsectors of the economy provides important signals to the economy asa whole. Creating a large flow of investments that are essentially“homogenized” would deprive the market of these vital signals.

Therefore, once a worker has accumulated some “trigger” level offunds, the worker should be free to participate in a much largerrange of investment options, closely approximating the optionscurrently available under traditional 401(k) plans. Management offunds at this level would again be through the private sector, withentry open to any company offering qualified funds. ERISSA offers areasonable framework for determining participation andregulation.

This three tier structure — a central collection point andholding pool, a limited set of initial investment options, and aeventual expansion to a wider array of investments‐​can provideworkers with the greatest amount of security while maximizingconsumer choice and control. It would keep administrative costs andthe burden on employers as low as possible. And it would minimizegovernment interference with investment decisions.

Works should be able to switch investments, between and withininvestment tiers on an annual basis. While an “open season,“similar to that of the FEHBP health plan, is perhaps the mostsimple approach, that could lead to an excess of market “churning“over a limited period. An alternative approach would be to allow aworker to switch investments within a designated period centered onhis or her birthday.

There is one other additional point to keep in mind, at allthree tiers, the accounts would be the property of the worker. Thisownership is one of the perhaps the most important reasons forreforming Social Security, and it is vital that it be maintained aspart of any administrative structure.

Let me conclude by saying that I believe that Social securityreform is not an option, but a necessity. The program will beginrunning a deficit in just 12 years and faces unfunded obligationsof roughly $12.8 trillion. The need for reform presents us with anopportunity to create a new and better retirement program for allAmericans, a program that gives workers ownership of theirretirement funds, more choice and control over their money, and theopportunity to build a nest egg of real inheritable wealth.Therefore, any successful Social Security reform should includepersonal accounts.

This makes the work of this Committee all the more important:getting the design and structure of the accounts right. I believethat the structure I have set out today takes us in that direction.I look forward to the Committee’s questions.

Thank you.

Michael D. Tanner

Subcommittee on Securities and Investment
Committee on Banking, Housing and Urban Affairs
United States Senate