In the comments on my post on retirement last week, Chris Koresko asks a good question, writing:

Would it make sense to link the size of a retiree’s Social Security payment to the number of kids he has in the workforce paying Social Security taxes?
After all, if not for SS it is likely that working people would be spending more of their income taking care of their parents; therefore, such a rule would help reduce the redistribution from large families to small ones.

I think it would make sense, for the reason he says. His proposal, if implemented, would reduce the distribution from large to small families. But then, if one’s goal is to reduce distribution [I don’t call it “redistribution” because the term “redistribution” presupposes that someone distributed in the first place] why not take the next logical step: make Social Security non-distributive. How would you do that? End it. Indeed, when you look at many government programs and you start thinking about reforming them to make them less inefficient or less distributive, you can usually reason your way to ending them in toto. I have no difficulty with that. But it does raise an issue: maybe the purpose of many of the supporters of the program is to do what many of the opponents don’t want it to do.