One of the hottest documents circulating around Washington today is a highly technical, statistics-laden, 131-page paper by Hoover Institution economist Michael Boskin. First reported by Jim McTague in Barron's on June 16, it estimates that the taxation of pension assets, including Individual Retirement Accounts and 401(k) plans, will yield a $12 trillion (in today's dollars) windfall to the federal government between now and 2040.
Business Week followed up with a major story in its June 30 issue. It noted that if Boskin's numbers are correct, this unexpected revenue stream would make up the entire shortfall in Social Security and Medicare through 2040. This possibility led Rep. Jim Saxton, New Jersey Republican, to declare that gloom over the government's long-term fiscal imbalance is "exaggerated."
In truth, this news is not as much of a revelation as it might appear. I wrote about it 3 years ago based on research (cited by Boskin) by Dartmouth economist Jonathan Skinner. He calculated that, contrary to conventional wisdom, Individual Retirement Accounts actually made money for the federal government.
Skinner did not base his calculations on supply-side economics or anything like it. It was a simple matter of mathematics. Although the federal government loses revenue when people get a deduction for their contribution to an IRA or 401(k), and also loses revenue from nontaxation of annual returns, it gets all the money back plus interest when funds are withdrawn from such accounts and taxed at that time.
That is because the size of the accounts at withdrawal is much larger due to dividends, interest and capital gains, compounded over several decades. Even if one assumes that the tax loss from IRAs causes the deficit to rise, the government still comes out ahead because it borrows at the lowest interest rates possible.
If individuals just buy high-grade corporate bonds, the rate of return will be higher; if they buy corporate stock, the return will be much higher. Therefore, the government gets more than enough extra revenue to compensate for the revenue loss and the interest expense resulting from it.
All withdrawals from tax-deferred accounts except Roth IRAs are taxable and always have been. Moreover, if someone should die with a balance in their IRA (including Roth) or 401(k), their heirs must pay tax on it at ordinary income tax rates -- losing the benefit of lower tax rates on capital gains and dividends. This will be true even if the estate tax is abolished.
For this reason, financial planners suggest that tax-deferred saving may actually be a bad idea for some people. Normally, assets received from estates are tax-free to the recipient. But this is not true of receipts from tax-deferred accounts. Those assets are now taxed by the estate tax and also constitute taxable income to the heir.
Of course, the receipt of a large lump sum likely pushes most heirs into a higher tax bracket, so many will be taxed as if they are rich, if only for a single year. The point is that all IRAs and 401(k)s ultimately are taxed.
Based on standard mortality tables, historical rates of return, estimates of future contributions and current tax law, Boskin simply calculated the revenue from such withdrawals.
We can get a back-of-the envelope estimate of what Boskin did by looking at the total assets in tax-deferred accounts today, which amount to $10.5 trillion according to the Federal Reserve. These will generate about $3 trillion in tax revenue (in today's dollars) even if no one makes another contribution. This revenue stream will rise in future years by the same rate of return that individuals receive on their investments.
With new contributions of some $400 billion annually, $12 trillion in revenue between now and 2040 from taxing withdrawals actually seems like a rather conservative estimate.
Experts are already pouring over Boskin's numbers, and they may well be revised before he publishes his paper. The important thing is not the precise estimate, but the order of magnitude and the acknowledgement of a future revenue stream for the government that was previously unknown.
Actually, it was not unknown to the government's revenue estimators. It's just that they only calculate revenues out for 10 years, and much of the revenue Boskin projects will come later, after the baby boom generation retires.
In any case, there is no practical way of tapping that future revenue stream today. Moreover, the tax laws undoubtedly will change in future years to prevent the overall burden of taxation from rising, as Boskin assumes.
In conclusion, there is really no pot of gold out there that has suddenly been discovered that obviates the need to restructure Social Security and Medicare. Those programs are still actuarially unsound, and nothing in the Boskin paper changes that.