Nouriel Roubini has jumped into the social security privatization discussion (hat tip, MaxSpeak), and he's pretty rough on its proponents.

... when you carefully look at the facts, it becomes clear that the proposed partial Social Security privatization is literally a Con Man Smoke-and-Mirrors Shell Game that - in the form it has been proposed - will not lead to any of the alleged benefits argued by its supporters. It is amazing the amount of misinformation that one reads about social security privatization; apologists argue that:

- The current pay-as-you-go (PAYGO) Social Security system is bankrupt.
- Privatization would increase national savings and the accumulation of capital and thus lead to higher long run income and growth.
- Privatization would lead workers to be able to invest into higher return equities rather than the low return public debt of the current system.
- Privatization would be self-financing and have no long run transition costs as you reduce the large future implicit liabilities of the current system, in spite of the fact that you are creating large transtion costs via privatization. So, it is a free lunch.

As I see it, the problem with Nouriel's claims is these positions do not reflect the opinions of serious advocates of privatization.  Let's take them one by one.

First of all, is the current pay-as-you-go (PAYGO) Social Security system insolvent or bankrupt? The answer to this question is more complex than ideological sound bites.

I don't know about the "ideological soundbites," but I agree that loaded terms like "bankrupt" are not useful.  The point Professor Roubini readily concedes:

It is correct that the current system faces serious long-term deficits.

And those in favor of privatization should concede this point:

Resolving this unfunded liability has a cost of only about 1.89% of taxable payroll if action is taken today to fix the hole in the current system. In other terms, if we start today and increase payroll contributions (and/or reduce future benefits for current young workers and/or change the retirement age) we need only a permanent adjustment equal increasing the payroll tax by 1.89% (i.e. an amount that is about 1% of GDP) to ensure that the current PAYGO system is solvent forever (more precisely for 75 years as this is the long run horizon considered by the Social Security trustees). Thus, $10 trillion problem ($3.7 in NPV terms) is not as large or scary if we start acting today to fix the current system. It is totally manageable.

I don't think anyone is disputing the fact that there are many ways to bring projected social security liabilities back in line with revenues.  The question is how we do it -- that's why we are having this discussion.

Second, the administration plan to start partly privatizing Social Security would create a large further hole in the budget balance and sharply increase the budget deficit. Specifically, privatizing social security has a massive transitional fiscal cost: if the young workers contribute less to social security when part of their payroll tax is cut and goes instead to private accounts, you still need to pay for the benefits of the current old, retired and soon to be retired, who have earned their benefits via contributons while working.

No fair.  As suggested here, here, and here (among other places), feasible privatization almost certainly requires a reduction in promised future benefits, and nothing is changed by making the remaining implicit public benefits explicit by issuing formal IOUs.

Third, a social security privatization financed by debt (increased deficits) is, as even the strongest supporters of privatization like Larry Kotlikoff admit, only a shell game that does not lead to the benefit of increased national savings and capital accumulation in the long run.

Right.  I think that's what I just said, but part of the reform is, of course, to reduce the total amount of social security liabilities.  No shell game there.  (Nouriel at this point in the essay does distinguish between "intellectually honest academic supporters" of privatization and "Con Man politicians," which almost  makes me feel better but for my natural inclination to believe that the designation "Con Man" is too often a synonym for "Straw Man.")

Fourth, what about the $10 trillion ($3.7 in NPV terms) of unfunded liabilities in the current PAYGO system? Don't they imply that the current system is bankrupt and that current workers will not get
their benefits when they retire?

Specifically, arguing that privatization reduces the liabilities coming from the promise of future benefits for the current young (and it is thus self-financing over the long run)is false as it confuses the two problems ("privatization" and "prefunding") to give you a Con Man Three Cards Monte game privatization. As in the Three Card Monte con game, it bamboozles and fools the hapless current worker in believing that the alleged higher returns on stocks will more than compensate the losses from having two losses: reduced benefits to fill the unfunded hole ("prefunding") and the extra current and future tax cost of paying for the transition to the privatized new system ( thisis solely the effect of "privatization"). It is a total Smoke and Mirrors game.

I agree that privatization -- the diversion of some payroll contributions to private accounts -- is logically distinct from the "prefunding" or unfunded liability issue.  And I certainly concede it is a disservice to claim that the former, in and of itself , resolves the latter.  But I repeat -- who is  making that argument?  Responsible advocates of privatization are simply not arguing that there is a free lunch to be had -- and until we see specific plans and how they are  sold to the public, we may as well assume that most advocates are responsible.

And here is, once more, a position I just can't agree with.

And the recent pathetic attempt by Republicans in Congress to hide the transition costs by putting such costs off-budget or by using other accounting scams turns the privatization con game into a total farce.

It's not crazy or dishonest to view this as a way to implement a sensible accounting system that recognizes the fundamental irrelevance of a deficit that does not change the true value of the government's long-term liabilities -- an irrelevance, by the way, that is fundamental to almost all of the arguments that Nouriel makes.  It would be better to rationalize the entire approach to fiscal accounting.  But barring that, it would really be a mistake to, for example, run afoul of budget rules designed  to constrain government behavior just because there is a paper transaction that has no economic meaning.

Furthermore, Nouriel's position on this would have been stronger had he not seemingly endorsed another bit of shady accounting.

But given the current large build-up of the Trust Fund, the current system will have the funds to pay all expected benefits until 2042. Thus, the problems with Social Security, from a cash flow point of view do not start until 2042.

The "Trust Fund," of course, is nothing but IOUs that the government writes to itself.  It, in fact, does not represent assets accumulated by the government, but rather the "promise" that we will spend the surplus payroll taxes today, and pay for future benefits out of general federal revenues (i.e. the income tax).  Could anything be more potentially misleading than to call this particular fiscal scheme  a "Trust Fund"?

Now, I know Nouriel is not trying to mislead anyone.  But I do think the discourse will be improved if opponents of privatization resist the temptation to suggest that the other side is.

UPDATE: I should have linked to this, from the president's chief economist.  It sure doesn't sound like an attempt to fool anyone.

"Let me state clearly that there are no free lunches here," said N. Gregory Mankiw, chairman of the Council of Economic Advisers, at a conference on tax policy here.   

"The benefits now scheduled for future generations under current law are not sustainable given the projected path of payroll tax revenue," he added. "They are empty promises."

Mr. Mankiw's remarks suggested that President Bush's plan to let people put some of their Social Security taxes into "personal savings accounts" would have to be accompanied by changes in the current system of benefits.