consequences; the actuarial status of Social Security varies dramatically with changes in real wage growth of even 0.1 percent annually.
Shoven suggests an answer to the risk issue: since both defined-benefit and defined-contribution systems are risky, we should have both types of systems. I agree with this solution. As a general matter, society appears overly invested in defined-benefit savings instruments and insufficiently invested in defined-contribution instruments. A partial privatization of Social Security is one way to even out this balance.
On net, I come out near where Shoven does. Partial privatization of Social Security is a good idea if it contributes to saving more as a society and offsets some of the risk of a defined-benefit retirement system. But perhaps more important, I find the debate about Social Security privatization inadequate. The debate is being guided by myths -- society can be wealthier and no one need be hurt, privatization will raise labor supply substantially. These claims are less true than is suggested, or are completely false. Conducting a debate on this basis is dangerous. If we persist in promising people a free lunch, what will we tell them when the bill arrives?