Matt Yglesias endorses a take no prisoners approach to privatization:
Some people want to know why I seem to think Social Security should be a “litmus test” issue for Democrats. On one level, it’s a complicated question requiring a long answer. On another level, it’s a simple question that’s easily answered. First, Social Security is not some little obscure thing. It’s dull, yes, but it’s also the largest federal program. If anything’s worth fighting over, it’s the largest program out there. Second, privatization is bad policy. I won’t go into that now because I’ve written on it extensively. Third, compromise is bad politics.
I’ve made the case before that privatization is in fact bad policy, largely because social insurance, which shifts risk from risk-averse individuals to the risk-neutral government, creates real value. But that argument was couched in somewhat abstract terms: risk premiums, certainty equivalents, binomial distributions, and laws of large numbers (I’m an economist, that’s how I think). I recently came across a series in the LA Times by Peter G. Gosselin that puts a human face on those very risks which Social Security offers some protection against. See
The series, and the risks and travails it documents, also brings to mind another argument against privatization that I have not seen widely promoted: by reducing the costs of risky but on average value-creating economic ventures — i.e., entrepreneurship and ownership — Social Security encourages those very activities. Thus it’s not at all clear, as its advocates often claim, that privatization would increase overall entrepreneurship. (For a related argument, see this post by PGL.)