A confluence of personal and work-related responsibilities made for very spotty blogging last week.  I'm back, but there are few pieces of unfinished business to which to tend.  First up is my earlier post regarding Greg Mankiw and Phillip Swagel's March 3 pot-shot at critics of President Bush's desires on social security reform -- more specifically, Brad Delong's and pgl at Angry Bear's complaints about my comments.

At issue -- or at least what was at issue in my post -- was the following passage from Mankiw and Swagel:

The introduction of personal accounts will involve some transition financing, but this increase in the budget deficit won't place a new burden on future generations. These deficits are just an acknowledgment of promises that were made long ago.

In his original comments on the subject, DeLong followed this up with an earlier passage from Mankiw's textbook about deficits increasing the demand for loanable funds, raising interest rates, and so on.  I cried foul at this comparison, and Brad cried foul back in a comment to my post.

You chopped your quote from me off too early. The next paragraph is:

"We worry about the budget impact and effect on national saving of the Bush Social Security plan because there is a chance that it will significantly diminish national saving and lower economic growth. If holders of defined-contribution private accounts regard them as close substitutes for their other assets in a way that promised defined-benefit standard Social Security was not, then the Bush plan will be yet another budget-busting shift of the tax burden that will slow economic growth.

Well, creative editing is a hanging offense if it distorts another's position, and I had no intention of doing so.  But my criticism was much more narrowly drawn.  What I objected to was not the omitted paragraph above, but the implication that somehow Mankiw was not being true to his school by claiming that the specific form of debt that might be involved in the conversion of future benefits will be economically neutral.   Pointing out that most instances of deficit finance are non-neutral does not imply that it is so for every policy.  Mankiw (and I) can live quite comfortably with both statements, and to present them as if they are in conflict is pure misdirection.

DeLong is, of course, correct in noting that it all comes down to whether "holders of defined-contribution private accounts regard them as close substitutes for their other assets in a way that promised defined-benefit standard Social Security was not," a point he repeats in his march 12 post "Mankiw 0, Liberals 3."  (I can't seem to get the permalink thing working.)  But my second point remains.

That second point is this.  If debt issued to represent promised future benefits are not viewed as perfect substitutes by beneficiaries, then it is hard to see why the same sort of assets held in the social security trust fund should be. 

pgl notes DeLong's comment about substitutability, but adds this:

Mankiw claims some households are liquidity constrained. Now I’m not sure if the Bush plan would relax those liquidity constraints precisely because Bush refuses to give specifics. But if it does and if one believes liquidity constraints are important as Mankiw does, then how can one appeal to Ricardian logic.

I'm not sure if Mankiw is still appealing to this argument or not -- it's not in the Wall Street Journal article in question. Certainly if the debt were to take the form of individual-specific "recognition bonds" that are otherwise non-marketable, there would be no impact.  That, though, is an issue related to the substitutability of promised benefits, about which I agree reasonable people can disagree.

The main point of my post was that I find it difficult to reconcile the argument that converting promised benefits into explicit debt is non-neutral with the position that those promises are inviolable.  In the case of liquidity constraints, the existence of promised benefits have no economic impact, as the consumption and saving plans of constrained individuals are not contingent on the receipt of future benefits.  (Unless, of course, the future income represented by those payments is the reason for the constraint, which I find hard to believe is what anyone has in mind.)

In short: If I really believe that I am going to be paid the benefits promised at the time I pay my social security taxes, it is an act of bad faith on the part of the government to renege after I have done my life-cycle planning contingent on the expectation of the receipt of those payments.  But if that is so, issuing a paper promise won't affect my behavior.  If that piece of paper does affect my behavior, then I must never have believed the promise in the first place. I guess I'll leave to the philosophers to determine whether a promise not believed is really a promise.  But I think the economics are clear.

P.S. pgl also takes on Mankiw's recent (subscriber only) New Republic article.  For the record, Mankiw does not claim there is a free lunch on Social Security reform in his Wall Street Journal article.