Paul Krugman has a list of four:

1. The relevant life expectancy is life expectancy at or near retirement age. Falling infant mortality doesn't make a case for delaying Social Security -- and that's important, because gains have been much less striking at age 65 than at birth.

2. Gains in life expectancy have been very strongly correlated with income and class; those with lower incomes and lower status -- the very people who depend most on Social Security -- have seen very small gains in life expectancy:

3. The retirement age has already been increased: the Greenspan Commission of the early 80s set it in motion, so that it's now 66 and scheduled to rise to 67, essentially consuming all of the life expectancy gains of the bottom 50 percent.

4. The alleged wise men of DC don't know any of this. When Ryan Grim tried to ask Alan Simpson [of Simpson-Bowles] about it, Simpson replied by denying the facts, attacking the interviewer, and insulting the AARP.

I'll add a fifth: because we can afford it and everybody deserves to share in this nation's productivity gains.