Greg Mankiw Criticizes Lieberman, Bush, and Krugman...

Greg Mankiw criticizes Joe Lieberman, George W. Bush, and Paul Krugman--but, I think, gets important elements of the situation wrong:

Greg Mankiw's Blog: Krugman blasts Lieberman (unfairly): Lieberman (and Bush) said: "Every year we wait to come up with a solution to the Social Security problem... costs our children and grandchildren and great-grandchildren $600 billion more."... [W]ere Lieberman and Bush telling the truth or not? It helps to understand what that $600 billion figure represents.

It all comes down to the economics of discounting. Let's say I have a $200 debt to a bank that is due in 10 years, and the interest rate is 7 percent. The present discounted value of that obligation is $100 today. That is, if I were to extinguish the debt right now, it would cost me $100. If I wait a year, the repayment obligation would then be only 9 years away. At that point, extinguishing the debt would cost me $107. One might say that waiting an extra year costs me an extra $7. Or one might not. After all, paying $107 next year is equivalent in present value to paying $100 today. One might say I haven't cost myself anything by waiting.

Now suppose my friend Joe offers me some financial advice: "Greg, you should really pay off that debt now, because waiting a year will cost you an extra $7." How should I respond?

(a) "Yes, Joe, you are right."
(b) "No, Joe, that's not the best way to think about it."
(c) "Joe, you are a liar."

I think (a) and (b) are defensible points of view. But Krugman chose (c).

Three comments:

First, in this case (a) is not a defensible point of view. Some $200 billion of the $600 billion number is a simple inflation adjustment--2006 dollars have less purchasing power than 2005 dollars. In any other situation, Greg would sharply downgrade someone who believed that changes in the value of the measuring rod that is money were changes in real income and wealth. The statement, "Yes, Joe, you are right," is simply not true.

Second, with respect to (b), what is the best way to think about it? It is not quite right to say, as Greg does, that "paying $107 next year is equivalent in present value to paying $100 today... I haven't cost myself anything by waiting." By waiting a year, you've given up your capability to use this year's resources to help settle the debt. The real cost of delaying Social Security reform is that we gave up the our ability to change this year's taxes and benefits to start dealing with the shortfall. (Of course, since the never-fully-spelled-out Bush plan never envisioned changing this year's taxes and benefits...) A good plan would have raised taxes a little bit this year to prefund a little bit more of the long-run shortfall.

Jason Furman--who is, all by himself, a large and well-trained professional analytical staff--has done the math:

The cost of delaying Social Security reform an additional year is about $160 billion.... [W] missed the opportunity to tax 2005 GDP or reduce Social Security benefits in 2005. According to the Social Security Trustees, taxes would need to be raised by 1.3 percent of GDP to achieve solvency over an infinite horizon. So we have to do about $160 billion more in adjustment because we did not address the problem last year.... Here's the CBPP note on the cost of delay. http://www.cbpp.org/4-28-06socsec2.htm... [T]he same methodology shows that the cost of the tax cuts goes up by $1.1 trillion annually.... I also wrote something making the obvious point that you could only talk about a cost of delaying if you delayed a plan that had prefunding http://www.cbpp.org/4-28-06socsec.htm...

Jason is, of course, right. The right number--the number you put down for full credit on any public finance exam--is $160 billion. Put down $600 billion, and you'll get little or no credit for the question.

Thus I'm not at all sure that (b) is defensible either. We economists believe that analyses that confuse nominal and real magnitudes and that do not focus on opportunity cost are not just "not the best way to think about it" but are wrong.

Third, what is the real point of Lieberman's saying the "$600 billion" number, the number that Greg Mankiw's coworker in the White House Chuck Blahous told him to say, rather than the "$160 billion" that Paul or Jason or I or Peter Orszag or any of a large number of other economists interested in raising the level of the debate would have told him to say? The point is that the $600 billion number was created because it was (a) big, and (b) misleading and wrong in a subtle way that could not be easily explained to a journalist who was either an English major or who has wiped all memory of Econ 101a from his or her brain.

The purpose of the $600 billion number is to mislead those who hear it into thinking that the urgency of Social Security reform is much greater than it in fact is.

Is it a lie? It's a lie in the same sense that Bill Clinton's claim that "there is no relationship" with Monica Lewinsky is a lie: it's a statement calculated to mislead and deceive those who hear it, made by people who have an interest in clouding men's minds, making the worse appear the better cause, and otherwise lowering the level of the debate.

The right way to put it is, I think, the way Stephen Colbert does. The statement "Every year we wait to come up with a solution to the Social Security problem costs our children and grandchildren and great-grandchildren $600 billion more" is not a true statement, it is a statement that has "truthiness." It's not a statement that anyone who wants to belong to the reality-based community would ever say.