Social Security reform is dead; long live Social Security reform

Anyone who hasn't yet seen Ways and Means Chairman Thomas's comments on Social Security ought to check them out immediately.
Obviously, the main point of interest is that it is rather significant when, at this stage of the process, the Chairman of the Ways and Means Committee declares the Bush changes politically dead in the water.
Second prize, in terms of early blogosphere attention, has gone to his suggestion that women's retirement ages under the program should be later than men's because they live longer. While everyone knows this is a political non-starter, it raises a lot of interesting discussion points. Indeed, if I may be so self-serving, it raises the sorts of fundamental social insurance issues that I tried to explore, in greater depth than one would find in discussions more directed at the current political process, in my books on Social Security and Medicare.
The later retirement age might be viewed as "unfair to women" because it facially seems to treat them worse. But in fact, giving women the same retirement age as men, where both pay the same taxes for the coverage, means the longer-lived women get a benefit of greater value and thus, in effect, a transfer from men. An arm's length insurer that was permitted by law to take gender into account in pricing retirement annuities would be expected charge women more or give them nominally less, such as by deferring the retirement date. Indeed, adverse selection in insurance markets would pretty much require this outcome under fully competitive conditions.
Since I don't judge just distribution from the standpoint of market outcomes, this is not dispositive, but it is of interest. It raises the question of why we would mandate transfers from men to women relative to the market outcome.
If women live longer then men, they have greater lifetime needs, which all else equal might justify transferring resources to them. But if they live longer, then they also presumably can work longer, assuming that the longer life expectancy implied on average being more vigorous at age 65 or 70. Given that point, we might have no reason to transfer resources from men to women (keeping in mind, of course, that in two-spouse households the name on the Social Security check may not affect how the resources are actually shared). So perhaps the later retirement age for women makes sense.
But here's another idea Chairman Thomas might not like as much. Life expectancy is positively correlated with wealth and with lifetime earnings; that is, the richer tend to live longer. How about using that little factoid to make high-earners retire later? The one difference between this and the gender-related proposal is that wealth or lifetime earnings are to a degree within the individual's control, and thus potentially responsive to the incentive effects of such a rule, whereas gender (Christine Jorgensen cases aside) is not.
Plus, one could ask Chairman Thomas if this is the only gender difference he wants to take into account. For example, how about, as Ed McCaffery has proposed, lowering women's (or secondary earners') marginal tax rates on earnings because their work decisions are more tax-responsive? No? I thought not.
Still, it is good to hear someone raising the issue of the relationship between retirement age and life expectancy. One of my pet Social Security and Medicare proposals is to peg eligibility or retirement age to increases in officially computed life expectancy. Part of the idea is simply to create a different default rule than present law, so that the definition of what is a "benefit cut" changes and we don't have people automatically getting larger lifetime benefits, without Congress's expressly "increasing" their benefits, because of living longer.