Arnold Kling: Guilty as Charged

Arnold Kling responds to my post on global warming and Social Security. I said:

Here's what I've noticed. Some of the same people who argue there's too much uncertainty about the climate 75 years in the future to justify drastic action now use the so-called crisis in Social Security funding 75 years from now (which is far more uncertain than climate change) to argue for drastic change today.

He says he's guilty as charged, but for good reason:

Social Security and Global Warming, by Arnold Kling: ...The reason we should try to fix Social Security now is that the cure can be painless now. The problem is that under conservative assumptions about productivity, promised future benefits exceed future revenues by an ever-growing amount. So cut promised future benefits, and then if productivity does well, you can restore benefits if you like. The point is, planning for the worst case scenario does not hurt anybody in the non-worst-case scenario.

In the case of global warming, taking drastic steps now to prevent the worst-case scenario has real costs, against benefits that are potentially nil--in fact likely nil, for a variety of reasons. So, contra Jane and Mark, it is possible to be rationally precautionary about Social Security and not so much about global warming--or precautionary in a different way about global warming, as I've been suggesting.

As an aside, I disagree with Mark's assessment of the relative likelihood of a Social Security crisis and a global warming crisis. ...

If I were to put numbers on it, my subjective probabilities would be that there is about a 5 percent chance of a significant shortfall in Social Security down the road, and less than a 1 percent chance of a climate catastrophe caused by carbon dioxide emissions. The climate catastrophe is potentially much, much worse, which makes it a legitimately larger worry. But the cost of acting now on Social Security is essentially zero, and the cost of acting now on carbon dioxide emissions is huge…

Arnold also says, in reference to a December 2004 post by Jane Galt about him:

Before I plead guilty, let me remind Mark that Jane Galt made this point over two years ago.

I appreciate the reminder. I didn't have a blog until a few months after Jane's post and had only just discovered econ blogs at that time so he's right, I missed that one.

I'm confused about something though. If you were to tell me that I will receive 1 million dollars on the day I turn 67 would that change my behavior today? Yes. I'm pretty sure it would. Would I feel better off today? Yes again.

So, go in the other direction. If you tell me I will have less money in the future because my benefits will be cut, maybe not by a million dollars but by enough to matter, will that change my behavior today? Yes it will, and I’ll feel worse off.

More importantly, is it reversible like Arnold says? Suppose I give up a vacation and save instead because I believe benefits will be cut in the future. I can't always reverse that later (I can't necessarily climb the same mountain or ski the same slopes when I'm older, and even if there's no physical constraint, the opportunities will differ with time). Or maybe the benefit cuts make it so I can't simultaneously send a child to college and save as much as I need to for retirement due to resource constraints. How does cutting future benefits, even probabilistically, leave me unaffected in such a case?

The point is that changing expected future benefits alters the time paths for consumption and saving (a premise, or at least a hope, of those in favor of privatization of Social Security)  which, contrary to the claim above, does affect people "in the non-worst-case scenario."

Update: Angry Bear has more, including rebuttal comments by Arnold Kling.