Sometimes you just have to hand things off to the pros. Via Brad DeLong, here's Peter Orszag's House testimony (PDF) utterly demolishing the Bush privatization plan on all fronts. The man is truly a weapon of mass destruction. There's one subtle point, however, that's worth highlighting.

At a basic level, the Bush plan works by providing workers with a loan at a 3 percent real interest rate, which then gets paid back on retirement. The White House has quibbled with this characterization, but for accounting purposes, that's how it works. Now notice—at every single point in time, from now until eternity, the federal government will have a number of outstanding loans to workers who haven't retired. So the program will always be running deficits. As Orszag says, "the Administration’s account proposal would raise public debt by more than 30 percent of GDP over the very long term."

By contrast, under the current system, the Trust Fund will run out in 2042 or 2052 or whenever, and after that, barring any other changes, we will have to cut benefits slightly in order to make the system's expenditures match revenue. Benefits will still be very generous, and much higher in real terms than they are today (people my age can expect a median of $19,000 in annual benefits), and more importantly, Social Security will never run deficits again. So by the time I'm 60, we would no longer have to worry about Social Security's solvency.

Under the Bush plan, by contrast, there will always be outstanding loans, and the government will need to continue borrowing money long after I've reached 60. My grand-kids and great-grand-kids will get saddled with eternal debt. The only way the system can stop running deficits is by eliminating guaranteed benefits altogether. The program no longer splashes red ink across the budget when (and only when) full privatization is achieved.

Orszag also notes four important scenarios in which the "loan" to workers will not ever be repaid back to the federal government, thereby further weakening the system. First, if a worker dies before retirement. Normally a worker contributes payroll taxes to the system, and if he or she dies before retiring, those revenues are used to pay other retirees. But under the Bush plan, those revenues would be diverted from the system by being passed on in the form of a bequest, to a family member, say. Now that's an appealing feature of the plan except for two things: 1) the current system already provides for survivors of workers who die before retirement, and 2) If this was really something we wanted to do, it would be much, much cheaper for the government simply to buy life insurance for everyone. But doing it the Bush way means there's less money to pay current retirees—and since one-seventh of workers die before retirement, this creates a huge hole.

As well, the Bush plan will create bigger-than-expected deficits if: a) someone works, say, ten years, receives the loan for his private account, but then never has t to pay it back in the form of a reduced benefit because he didn't work long enough to receive a SS benefit in the first place; b) lots of seniors are left in poverty by the Bush plan and Congress needs to manually increase the guaranteed benefit to alleviate this; or c) real interest rates on T-bills go above 3 percent. Note how c) works—the government assumes it can borrow money at 3 percent interest, so that it can then loan the money to workers at a fixed 3 percent interest. But if interest rates go above 3 percent—a likely scenario with the massive deficits Republicans have created—then the government loses money on the whole venture. (Obviously if real interest rates on gov't bonds go below 3 percent, the opposite happens. So which is more likely?)