So I Was Off by a Few Trillion

Michael Boskin's miscalculation on tax revenues muddles the economic debate

Did you ever make a mistake that was so large and so public that it made you want to crawl under your bed? If so, you can imagine how Michael Boskin feels these days.

Earlier this summer, Boskin, a highly respected Stanford University economist who was head of the Council of Economic Advisers under George H.W. Bush, electronically circulated a startling paper to fellow economists. In it, he argued that the future tax payments on withdrawals from tax-deferred retirement accounts, such as 401(k)s and employer pension plans, were being drastically undercounted. That meant federal budget revenues could potentially be in for a huge, unforeseen windfall. The underestimate over the next 40 years -- almost $12 trillion, in today's dollars -- was so big, Boskin wrote at the time, that it was "larger than the sum of the 75-year actuarial deficits in Social Security and Medicare plus the national debt."

With such sweeping claims coming from such a reputable economist, Boskin's paper quickly drew widespread attention. The June 30, 2003, issue of BusinessWeek, for one, reported that "Boskin's paper has the potential to shake up the budget debate."

But as the paper received increasing scrutiny, it sparked a huge controversy and came under sharp criticism from fellow economists who argued that the savings would be nowhere near what Boskin predicted. He took another look at his numbers and concluded that he had made a serious mistake: A key term had been left out of a complicated calculation of the effect of tax deferral on future savings and business investment, possibly wiping out most of the estimated $12 trillion in savings. In mid-July, Boskin sent an e-mail notifying recipients of the paper about the error. "I apologize for any inconvenience this may have caused you," he wrote.

Boskin is now redoing his calculations, both to fix the mistake and to take the criticisms into account. But the debate is far from over. Boskin still argues that some of the revenues from deferred taxes are not being counted in long-term budget projections, but he's not willing to hazard a guess about the magnitude until he has had a chance to redo -- and double-check -- his figures.

His critics won't even grant him that much. They argue that even if deferred taxes are undercounted, the sum is too small to affect the enormous fiscal gap projected between long-term revenues and outlays. "The projections of revenues for tax-deferred accounts don't alter the fiscal picture in any significant way," says William G. Gale of the Urban-Brookings Tax Policy Center. Gale, along with Alan Auerbach of the University of California at Berkeley and Peter R. Orszag of the Urban-Brookings Center, wrote a stinging rebuttal to Boskin's paper after it was reported in the media. Other economists have also challenged Boskin on a host of issues, such as overstating the tax rate on withdrawals, using the wrong interest-rate assumptions to value future taxes, and underestimating the deferred taxes already counted in long-term budget projections.

If there is a moral to this tale of the missing trillions, it is that making fiscal projections 40 years ahead is a very complicated and perilous business, even for the best economists. Even when the calculations aren't in dispute, the conclusions can depend as much on political assumptions as on economic reasoning.

Consider one of the thorniest problems under dispute: figuring out how much of the deferred taxes are already counted in existing revenue projections. Sound simple? Hardly. For example, the Congressional Budget Office's long-term projections assume, based on historical averages, that no more than 19% of gross domestic product would go to federal revenues. However, the projections don't specify which taxes would generate the future revenues. So to determine how much of the total represents deferred taxes, economists have to make assumptions about what people will put into retirement accounts, what they'll take out, and what the tax laws will look like in 40 years. All that calls for plenty of difficult assuming.

Boskin says he has received sympathetic calls from fellow economists, telling of similar mistakes -- caught, however, before being published. For the man who projected the $12 trillion windfall that wasn't, that's slim consolation.

By Michael J. Mandel in New York

    Before it's here, it's on the Bloomberg Terminal.