It's a good thing we didn't pass Social Security reform three years ago, isn't it! If we had, all those seniors on the brink of retirement would have lost half their savings in the big stockmarket slide of 2007-08. That just goes to show that we can't make people risk a dignified retirement on the Wall Street Casino.
First, in the later stages of the Social Security reform debate, Senator Jim DeMint proposed a plan by which Social Security private accounts would be created which would consist only of Treasury bonds. I supported this plan, because the move to semi-privatized Social Security accounts holding private stocks and bonds, though it is the correct way to handle publicly-mandated retirement finance, involves fiscal and institutional complexities which I thought should be handled gradually. DeMint's plan was a step towards a fuller reform that channeled mandatory retirement savings into private assets. If private accounts consisted solely of Treasury bonds, no one would have taken any losses on them during the past year and a half.
However, even if a more ambitious version of Social Security reform had been adopted, there was never any question (a) of forcing people who are near retirement to move to the new plan (Bush set the limit at 55), or (b) of putting older people's money all in stocks. A sensible private investor will invest mainly in stocks and riskier assets when young, then shift towards bonds and safer assets as she grows older. Social Security private accounts would do this automatically, erring on the side of caution. Young people might indeed have lots a good deal of money in their private Social Security accounts, but they have time to recover, and older people would have lost little. Also, with Social Security private accounts in place, the stockmarket slide might not have occurred, or not have been as deep. However, that's so complex that I can't certainly rule out the possibility that it might have been deeper. It is simpler to think what would have happened now if we had adopted the DeMint plan in 2005, and started creating accounts with only Treasury bonds in them.
If the government sets up private Social Security accounts which consist only of Treasury bonds, the fiscal effect is, to a first approximation, nothing. The government is still taking your money and making promises in return. None of the money is going to private assets, and the government doesn't have to borrow any other money to make up for revenue flows that have been diverted into private assets. The only difference is that the promise takes a different, more definite, legal form, and you possess a highly restricted form of ownership over assets (Treasury bonds) attached to a definite income stream. That income stream is still reinvested into your account, so you don't see it, but it is, in some sense, legally yours.
Now, suppose the crisis of 2008 hits. All of a sudden, there is massive demand for Treasury bonds from all over the world as investors "flee to safety." Meanwhile, stocks and housing prices are plunging, to a level which, though it's hard to know what they'll do in the immediate future, it's a pretty good bet that stocks at least (and homes too if immigration reform passes) are a good long-run bargain.
By now, people would have accumulated significant quantities of Treasury bonds in their Social Security personal accounts. So, the government could enact a simple reform. First, let people opt to transfer the management of their Social Security private accounts to licensed financial managers. Second, allow them, through these managers, to sell some of the T-bonds in their accounts and buy stocks instead. Of course, some people would be too spooked to invest anything in the stockmarket right now, but clever young people would know that, since the volatility shouldn't matter to them, it's worth the risk. Older people who have other retirement assets and aren't heavily reliant on Social Security, or who for whatever reason simply have high risk tolerance, would also make the switch.
Social Security private accountholders who took this option still wouldn't see any of the money. People could control how their money was reinvested, but it would all be reinvested. These are still mandatory retirement savings. But as they bought stocks, the money would go into private hands. Stock values would rise, encouraging business investment and/or, through wealth effects, private consumption. Note that the total stock of government liabilities would not increase in this scenario. They'd just change hands, from Social Security private accountholders willing to take on risk at a time when long-term returns look especially good, to other investors "fleeing to quality." Yet the effect would be "stimulus," that is, the use of the government's slack borrowing capacity to offset a sudden wave of financial disintermediation.
Why can't we do that now? Because although all those entitled to future Social Security checks are implicitly owners of government obligations, and although some of us might be happy to sell those government obligations and take on the risky assets whose falling prices are causing so many problems, the types of government obligations we implicitly own are of a such a poorly-defined and dysfunctional form that it's impossible to render them liquid. Therefore, the government can use its slack borrowing capacity to stimulate the economy only through tax cuts, which are distortionary, or through government spending, which is even more distortionary. These distortionary policies, in turn, tend to dampen confidence still more. Meanwhile, the government has to issue still more obligations, making more promises, because it cannot put its present stock of promises outstanding to work. Thanks to soulless Congressional Democrats who, not content with the politics of personal destruction, resorted to the politics of destruction of democratic deliberation, and refused even to talk with Bush and the Republicans (who had a powerful electoral mandate for Bush's agenda) about Social Security reform. As a result, we are still burdened with dysfunctional institutions of public finance, and handicapped in our ability to cope with the crisis.
UPDATE: It occurs to me that even as the Democrats were introducing an artificial panic about the not-really-very-risky-at-all prospects of putting Social Security money of people not near retirement into accounts partly of private stocks and bonds, Fannie Mae, Freddie Mac, the mortgage-interest tax deduction, and federal homeownership promotion policies generally, were actively encouraging the much riskier practice of highly leveraged homebuying. If people are so risk-averse that they can't tolerate the limited long-term risks involved in long-term investment of retirement savings in diverse portfolios of high-quality financial assets, then they shouldn't be buying homes on credit and exposing themselves to large undiversified risks in local real estate markets. If homebuying on credit does make sense then Social Security private accounts should be a piece of cake.
Of course, Social Security private accounts would probably be, not more, but less risky than the current Social Security program, only it involves a different kind of risk. Today's Social Security involves political risk: twenty, thirty, or forty years from now, the nonretired majority might decide they're sick of paying ever-higher taxes to subsidize seniors, and pull the plug. Market risk has several advantages over political risk. One is that market risk motivates people to engage in activities such as scrutinizing firms and gathering information about investment opportunities, which have positive social value since they encourage the allocation of capital to its best uses. Political risk, by contrast, motivates rent-seeking by organizations like the AARP, as different groups of society scramble to get politicians on their side. Another advantage of market risk is that you can opt for different degrees of it by the way you structure your portfolio, putting more in growth stocks if you don't mind some risk, or moving into blue-chip stocks, AAA-rated bonds, or Treasuries if you want to reduce or minimize risk. Political markets can offer no such menu of risk/return choices. Finally, financial markets generate vast amounts of data from which you can derive at least some estimate, albeit subject to a good deal of revision and error, of the mathematical expectation distribution of the income streams you'll get from your assets. With political risk, no such data is available.