As regular readers of my posts know, I’m not someone who dismisses Paul Krugman or who thinks that he has nothing worthwhile to say. Indeed, I think he’s one of the smartest bloggers around. So that’s why his post yesterday on Social Security has me scratching my head. He discusses the differences and similarities, in thinking about the long run, between doing something now about Social Security spending in the future and doing something now about global warming in the future. Getting into global warming will take me too far afield. Instead, I want to look at his arguments about Social Security.

Krugman writes:

Instead, they’re [the people he’s criticizing] pushing for things like a gradual rise in the retirement age and a change in the formulas used to compute benefits – things that will cut future rather than present outlays. Or to put it differently, they aren’t really trying to cut debt; they’re simply trying to lock us in now to the spending cuts they think we’ll eventually have to make anyway. And they never, as far as I can tell, really ask why it’s important to do this now.

But think about it; use Social Security as the example, although much the same argument applies to other programs. It seems probable if not certain that we will eventually either have to cut SS benefits (relative to current law) or raise additional revenue. So the threat, if you like, is that future benefits will fall short of what people now expect. To avert this threat, the usual suspects insist that we must gradually reduce the program’s generosity. That is, in order to guard against cuts in future benefits we must … cut future benefits. Huh?

OK, there are some arguments you could make; maybe the adjustment would be smoother, with less of a “cliff” when the trust fund runs out, if we set benefits on a downward glide path. But that’s a second-order issue, literally: we aren’t talking about preserving the overall level of benefits, we’re just talking about reducing its variance around a smooth trend. And given how uncertain we are about what the world will look like in 25 years, preemptively cutting right now could mean a gratuitous sacrifice of future benefits that may eventually turn out to have been affordable after all.

First, most of the advocates I know of who are advocating “cuts” in Social Security are not really advocating cuts. In the Social Security benefit formula, as real wages rise, real Social Security benefits rise. So the program could possibly be made sustainable without real cuts and without further tax increases if the formula were changed so that real Social Security benefits remain constant rather than increase.

Second, people can adjust better when they have more time to adjust. If the Social Security formula is altered for the future, people can have longer to save to make up for the higher benefits they would have got but will not get. That’s the argument for doing something about it now rather than later. Remember what happened in 1981 when OMB Director David Stockman tried to cut the early retirement benefit by about one third for people retiring only a few years later. That got nowhere. People looking at a one-third reduction in their retirement benefits who are planning to retire in a few years will not look on that kindly. But what if some previous Administration had announced in 1962 a gradual reduction in the early retirement benefit for people retiring in the early 1980s. Those people would have had much more time to plan.

Third and finally, take Krugman’s last sentence: “And given how uncertain we are about what the world will look like in 25 years, preemptively cutting right now could mean a gratuitous sacrifice of future benefits that may eventually turn out to have been affordable after all.” Yes, we’re uncertain about how the world will look 25 years from now, but surely Krugman must have seen some of the CBO projections for budgets in the 2030s. Under one reasonable set of assumptions, government spending as a % of GDP in 2037 would be a whopping 37% of GDP. This is up from its average over the last 60 years of about 20 or 21% of GDP. A whole lot of things would have to have gone right for us not to have the federal government spending a much, much higher percent of GDP. Planning here is asymmetric. If things go so well that in, say, 2037, government spending is only, say, 18 percent of GDP rather than the more-likely 30 percent or more, it will be easy to raise Social Security benefits then. But if it goes to even 25 percent or more, it will be very hard to cut Social Security benefits.