Sometimes an academic study offers much less information than it claims. And sometimes what appears to be an academic study is not even a study at all. Professor Austan Goolsbee of the University of Chicago, who has done respected work on other financial issues, has written an article[1] that claims to prove that the system of Social Security personal retirement accounts proposed by Plan II of the President's Commission to Strengthen Social Security[2] would result in an unprecedented "windfall" for financial firms. However, a closer examination of this article shows that it is riddled with errors, unjustified assumptions, and sensational but meaningless numbers.

A Fee Level Far Too High

Prof. Goolsbee rejects the 0.3 percent annual administrative fee that the Social Security Administration and other reputable experts use to estimate the costs of the personal retirement accounts that would be established under the Commission's Plan II. Instead, Goolsbee claims a 0.8 percent annual fee is more appropriate. This is simply wrong.

While that level would be appropriate for a system of individually managed accounts with a high level of personal services, this is not what the President's Commission to Strengthen Social Security would create. Instead, the Commission called for a simple, centralized investment plan modeled after the Thrift Saving Plan that is offered as part of the retirement program for federal employees.

But Prof. Goolsbee argues that a 0.3 percent fee would only cover "individual accounts where there is little or no customer service and where individuals are not allowed to make choices about their investment mix." But this is closer to what the Commission plan would create than the accounts Goolsbee envisions in his estimate. Moreover, the Social Security Administration and Congressional Budget Office have concluded that these accounts could offer their holders some investment choices at the 0.3 percent fee level.

A 1999 study by professional fund managers at State Street Corporation[3] led by Bill Shipman estimated that a plan similar to the one proposed by the Commission would have administrative fees around 0.3 percent. The State Street study is especially important because it estimated the costs of answering phone calls from account holders, developing the necessary computer equipment and programs, providing annual statements, and shifting between investment options.

In short, Prof. Goolsbee's administrative cost estimates might be appropriate for one type of personal account structure, but that structure is vastly different from the one that he claims to be studying. The President's Commission proposed a system of simple, low service accounts, for which a 0.3 percent annual administrative fee is more realistic. 

Based on a Mistake

Prof. Goolsbee claims that the Commission's Plan II requires account holders to convert their accounts from a low-cost centralized system to high-cost private management when they reach $5,000 in assets. He uses this to justify his assumption of higher fees.

But Plan II requires no such thing. Instead, it would create a low-cost centralized system in which workers could remain throughout their careers. And while the plan recommends examining the possibility of moving to private managers, that discussion is put off into the future. And even then, a system based on Plan II would allow such a move if it proved feasible but would not require workers to convert their accounts.

Even if a plan allowed workers to move their personal retirement accounts to a private manager-at a higher cost-only a few workers would take that opportunity. First, only a few of potential funds managers could offer higher returns, after subtracting the higher administrative fees, than the centralized system could provide. Workers who saw their earnings fall would promptly return to the centralized system. Second, when retirement account holders are faced with choices and are uncertain about what to do, they usually end up doing nothing. Because workers would have to affirmatively decide to use to a private funds manager and would not be moved to one automatically, most would do nothing and remain in the cheaper centralized system.

Ignores Benefits and Improved Retirement Security

Prof.Goolsbee's calculations ignore the benefits that would come from a system of personal retirement accounts. The first question that should be asked when someone talks about fees is "What do I get for that money?"

For a younger worker, today's Social Security offers minimal or even negative returns. A 20-year-old male can expect a return on his Social Security retirement taxes of -0.85 percent, while a 20-year-old female can expect a return of 1.91 percent. The difference is largely attributable to differences in life expectancy. The female can expect to live until age 84.3, while the male can expect to live to age 77.3.

On the other hand, Prof. Goolsbee admits that returns for personal retirement accounts would range from 4.6 percent under a 0.3 percent administrative fee to 4.1 percent under his 0.8 percent annual fee. Any positive return would certainly be an improvement for a 20-year-old male, and even the 4.1 percent return is more than twice the return that a 20-year-old female could expect from today's Social Security. In addition, both of these individuals would receive retirement benefits based in part on assets that they own, rather than depending solely on government promises.

In short, the better return is worth either administrative fee. This is not a case where the funds managers get fees for doing nothing; their activities will greatly increase individual worker's retirement assets.

Goolsbee Estimates Revenues, Not Profits

Even if Goolsbee's $940 billion estimate were correct, which it is not, it is still not an estimate of profit. Instead, Prof. Goolsbee estimates gross revenues. From that number must be subtracted the costs of keeping records on 140 million or more accounts, answering questions from account owners, providing regular statements, processing individual worker's investment changes, doing research into investment options, making trades, etc. Altogether, this is not a cheap proposition, and, significantly, the federal government, not the financial industry, would undertake many of these administrative functions. Only a small portion of the total fees would be left for the financial industry's profits.

A 1999 study by investment professionals at State Street Trust[4] assumed that, nationally, a PRA system would annually generate 175 million to 350 million calls from account owners, including 26 million to 52 million calls that must be handled by humans instead of an automated phone system, 14 million investment fund transfers, and over 140 million annual account statements.

Administrative fees would pay for all of these services. Calling the fees a "windfall gain" is both irresponsible and just plain wrong.

Sensational Numbers, Essentially Meaningless

The executive summary of Prof. Goolsbee's article is filled with sensational numbers, such as an assertion that administrative expenses for personal accounts would amount to over 25 percent of Social Security's anticipated deficit over the next 75 years. Prof. Goolsbee also makes claims about the impact of his estimated gross fee on financial firms and the costs workers' personal accounts would face from the administrative fees. But under close examination, all of these numbers mean little or nothing.

Comparing Prof. Goolsbee's inflated administrative cost estimate with Social Security's estimated deficit implies that this money could be used to partly fix the current program. But the money represented in his administrative fee estimates only results from having personal retirement accounts. Because these accounts grow at least twice as fast as today's Social Security, they produce earnings that could be used to pay administrative fees. Without the earnings from personal accounts, the money that would be spent on fees simply does not exist and would not be available to fix Social Security.

Similarly, saying that the administrative fees are more than eight times higher than financial firms' revenue loss between 2000 and 2002 sounds good until one remembers that he is comparing vastly different time periods. In other words, all that Prof. Goolsbee's statement means is that his estimate of total administrative fees over 75 years is eight times higher than the industry's losses over three years. One hopes that any industry would earn vastly more over 75 years than it loses over any three-year period.

Finally, stating that any set fee level reduces a personal account without considering what would be gained misses the point. Because administering an investment account is not free, every account has a cost. Even today's Social Security spends 0.6 percent of its outlays (about $250 million in 2003) to administer the program. A more realistic estimate is that fees will reduce the size of an individual's account by 8 percent rather than Prof. Goolsbee's estimate of 20 percent. However, even that 20 percent is a cheap price to pay if the account earns a return that is twice that of today's Social Security.

Not an Academic Study

A sure sign that Prof. Goolsbee's article is primarily intended to influence the election debate over Social Security is the intemperate language and exaggerated conclusions contained in the executive summary. Stating that the revenues that he claims would go to financial managers would be "the largest windfall gain in American financial history" without noting that this figure would be gross revenue and not profits is irresponsible. No real academic study would make this type of claim.

While Prof. Goolsbee has produced respected research on fees associated with 529 education investment accounts, this article is not up to those standards. It is filled with faulty assumptions and sensational numbers that are designed to confuse the debate over Social Security's future, rather then to educate the public.

Social Security personal retirement accounts are the only way to resolve Social Security's coming financial problems without massive tax increases or substantial benefit cuts. These accounts will have administrative fees, but the fees will be much lower than Prof. Goolsbee claims. And, most importantly, workers will receive real value in return.

David C. John is Research Fellow in Social Security and Financial Institutions in the Thomas A. Roe Institute for Economic Policy Studies at The Heritage Foundation.