Professor N. Gregory Mankiw says that extra government borrowing required by personal accounts "is offset by a reduction in the government's liability to pay future Social Security benefits" ("Personal Dispute," March 21). This may be true in a balance-sheet sense, but it is not the end of the story. The government's liability on Treasury bonds differs from its liability on Social Security. If you increase the government's liability on Treasury bonds and reduce its liability to pay future Social Security benefits, you are doing three things. First, the Treasury has never issued a bond with a longer maturity than 40 years, whereas Social Security liabilities reach beyond 40 years. So you are shortening the term of the government's liability. Many a bankrupt businessman can testify to the dangers of such an action. Second, you are issuing an obligation that cannot be modified (short of default) for an obligation that can be modified (by reducing benefits). The Treasury has not defaulted on a bond since the Gold Clause cases in the 1930s. Default nowadays could have disastrous consequences. Third, Social Security obligations right now are "held" by ordinary Americans, who are required to accept them and cannot sell them. Actual Treasury bonds are issued to buyers who have to be persuaded to buy them and who can sell them if they choose. It is not at all clear the bond markets will calmly absorb vast new issues of Treasury debt at reasonable interest rates. These three factors make Mankiw's balance-sheet point almost irrelevant.
Of the myriad objections that Mankiw chose not to deal with, I would like to call attention to just two. Social Security currently provides annual inflation adjustments. Private accounts likely would not, especially if one were required at retirement to purchase a fixed annuity. Even if inflation averaged only 2.5 percent per year (unlikely, given the huge deficits in our future), someone retiring at 66 would see a 22 percent erosion in their purchasing power ten years later. Second, the current system provides participants with a disability benefit. Participants under Mankiw's plan would be wise to purchase disability insurance on their own, because there's no way private accounts would be able to provide the necessary income for someone to live with dignity if he became disabled. David Gawenda Dane County Treasurer Madison, Wisconsin Using the Harvard retirement system as a model for the privatization of part of Social Security is an effective rhetorical device, but it is a red herring. There are several problems with this model. First, it represents an addition to Social Security and not a substitute, as the Bush plan proposes. Second, it is not a social insurance plan, while Social Security is just that--and Social Security does not permit opting out; thus, it is not subject to the standard self-selection problem. Third, Harvard professors are not in the average- or median-income category, so it is in their interest not to have the progressivity of the Social Security system apply to their retirement accounts. ("Let them eat cake," one might as well say.) Fourth, the Harvard retirement system presumably does not have to deal with the disability aspect of Social Security nor the legacy aspect described in Jonathan Chait's article ("Blocking Move," March 21). And, finally, Harvard professors are probably slightly more sophisticated than the average Social Security recipient and are better able to deal with the risk involved in defined contribution plans.
Emeritus Professor of Economics
Stony Brook University
Stony Brook, New York N.
GREGORY MANKIW RESPONDS: In the brief space I have been allotted, let me respond to three key points raised in these letters. First, David Gawenda and Egon Neuberger are concerned about disability benefits. This issue is important but not germane. The personal accounts that President Bush has proposed would fund retirement benefits under Social Security. They would not affect benefits for the disabled in any way. Second, Neuberger mentions the progressivity of the Social Security system. It is worth pointing out that the "progressive indexing" proposed by Robert Pozen and recently endorsed by Bush would allow the real benefits for the poor to continue to rise, while asking for the largest sacrifice from more affluent recipients. Third, Jack Harllee suggests that implicit debt and explicit debt are different because defaulting on explicit debt is more difficult and onerous than cutting future Social Security benefits. He is right. Indeed, if no action is taken, benefits will be cut automatically across the board when the trust fund runs out. But doesn't this observation belie the claim of many Democrats that the current pay-as-you-go Social Security system provides a low-risk income for future retirees?