Tuesday, September 2, 2008

Bob Barr on Social Security reform

Don't say we don't give equal opportunity to all presidential campaigns. Libertarian Party nominee Bob Barr, a former Republican congressman from Georgia, writes in today's Washington Times on how he would address Social Security reform. To be honest, it was better than I expected. What I expected was a big pitch for personal accounts coupled with free-lunch language that would cover up the costs involved on transition to accounts. While we got the pitch for accounts, Barr didn't cover up the costs:

The principal fiscal challenge arising from private accounts would be the so-called transition cost, which is financing Social Security benefits for current beneficiaries while current workers are shifting tax payments into private accounts. There is no escaping this budget pain.

However, under my proposed reform - with a quick shift to private accounts and phasing new workers out of Social Security entirely - costs peak early and begin to diminish, in contrast to the present system, in which the red ink explodes and never ends.

Benefit cuts are undesirable, but trimming payments is more fair than hiking taxes, which would only exacerbate the system's existing infirmities. First, we should adjust wage indexation, which increases benefits faster than inflation. Second, we should raise the retirement age, since life expectancy has increased by 18 years since 1935.Third, we should consider other benefit adjustments, such as fixing the consumer price index, which overstates the impact of inflation.

Other spending cuts - in corporate welfare, nonessential domestic programs, and military outlays for defending everyone but the U.S. - also will be necessary. If these reductions are insufficient to fully fund the transition, it might be necessary to consider short-term borrowing. Though not desirable, these obligations would be more than counterbalanced by eliminating today's massive unfunded liabilities as a result of moving to a self-funded system.

Say what you will, Barr has proposed a more complete and specific Social Security plan than either of the main party candidates. (I guess you can afford to speak your mind when you're not expecting to win…) In any case, click here to read the whole article.


Bruce Webb said...

All I can say is that Barr, Lyle and Tanner could all use some fact checkers. Not every error was particularly important in itself but taken together are indicative of the sloppiness that pervades much editorializing on Social Security, most people just don't seem to think it important to check their numbers against the most recent data releases. For example I have no idea where Barr came up with $15.8 trillion. I would understand $15.2 trillion or $17.4 trillion or $13.6 trillion, all are in context legitimate numbers, but it is not like $600 billion is just a rounding error.

And don't tell me Michael Tanner of all people doesn't know better:
"Even if Congress can find a way to redeem the bonds, the trust fund surplus will be exhausted by 2040. Overall, the amount the system has promised beyond what it can actually pay now totals $15.3 trillion. Setting aside some technical changes in how future obligations are calculated, that's $550 billion worse than last year."
2040 is the last year in which the Trust Fund will run a surplus for the year as a whole. But it still will be running surpluses for a portion of 2041. Is it really important that Tanner says "by 2040"? Well maybe not, although it adds a curious blend of uncertainty and specificity to the equation. And once again the difference between $15.2 trillion and Tanner's $15.3 trillion is maybe not all that important. On the other hand for the way that Tanner actually formulated his sentence the right figure is $13.6 trillion, and once again $1.7 trillion is more than a rounding error. And then we have the "Setting aside some technical changes". Well come now, either you believe the OACT is doing the best job it can in which case those 'technical changes' should be assumed to be corrections to the model, or you can be a paranoid like me and doubt the models altogether. But Tanner seems to want to have it both ways. In point of fact the Trustees in 2007 put what I suspect to be the equivalent figure at $14.4 trillion and 3.7% of payroll while in 2008 put that number at $15.2 trillion and 3.5% of payroll. Which would give Tanner an even scarier $800 billion deterioration to play with. On the other hand an alternate measure of Unfunded Liability leaves the PV unchanged at $13.6 trillion, while the measure over the 75 year actuarial window shows an improvement from $4.7 trillion to $4.3 trillion. Which just goes to show you can play all kinds of games using even the official numbers. But for the life of me I don't see where Tanner came up with $550 billion.

Now I get the difference between an op-ed and a policy brief and in such matters you can over-qualify and footnote yourself to death but still it would be nice to keep the numbers in line with the official ones. For example this by Tanner is just flat out wrong:
"Of course we could always raise taxes or cut benefits enough to bring the system into balance. But we would have to raise payroll taxes by roughly 50 percent."
Well no, even if you abandon the revision in immigration assumptions the increase needed immediately would be 2 percentage points or 16%. If you assume that all real incidence falls on the employee you could double that to 32%. Or you could use the official figure of 1.7% and get that increase to 13%. But I see no way to manipulate those numbers to make it come out to "roughly 50%". On the other hand it sure is a lot more scary then telling people they might need to fork over .85% of their gross to make the system actuarially sound.

I could nit-pick Tanner's argument further but really there is no point, as long as people are simply allowed to place out of context numbers into their opinion pieces we never will get to a place where a fully informed populace can make the right policy decisions.

(And I find it quite disturbing that a person in Tanner's position would stoop to implying that the Special Treasuries are phony IOUs.)

Andrew G. Biggs said...

My guess (although I can't be certain) is that Tanner was referring to the growth of the unfunded obligation that occurs from year to year through the inclusion of interest). The passage of time would tend to increase the value by around $700 billion, according to the 2008 TR.

That said, I don't tend to use the $13 trillion number very much, and the growth of that number is easily misunderstood for two reasons. First, part of that growth is due to inflation, which should be corrected for (eg, my compounding by the real rather than nominal interest rate). Second, while the unfunded obligation grows so does the tax base, albeit at a slower rate. So what you want to know is how does the ratio of the unfunded obligation to the tax base (currently around 3.2%, I think) change? It does rise, but not nearly as quickly.

Bruce Webb said...

Shoot lost my comment. Lets try again. These figures from the Trustees represent Unfunded Obligations for all particpants in dollars and as a percentage of payroll and GDP as well as the offset from future participants:
2003: $10.5 trillion, unstated, unstated, $.0
2004: $10.4 trillion, 3.5%, 1.2%, $-.8 trillion
2005: $11.1 trillion, 3.5%, 1.2%, $-.9 trillion
2006: $13.4 trillion, 3.7%, 1.3%, $.1 trillion
2007: $13.6 trillion, 3.5%, 1.2%, $-.8 trillion
2008: $13.6 trillion, 3.2%, 1.1%, $-1.5 trillion

This is rather a curious series. What it seems to suggest is that the key current challenge to Social Security is not Boomers, they are funded through 2041 and take up only a small part of the $4.3 trillion unfunded liability through the 75 years window. And it is not Gen-X, they will mostly be shuffled off to Buffalo and points beyond by 2082. Nor it is futue kids , if we could isolate them as a group they are increasingly scheduled to cost less than the current tax level projects to deliver (which does not necessarily translate to paying more in taxes than they get in benefits, theoretically we could handle that by a multi-tier of taxation levels). Nope the real culprits seem to be the Echo Boomer/Millenials and their kids who would be the ones scheduled to collect benefits between about 2060 and 2100.

Which is to say our real task is not to digest the first pig through the python (Boomers) but instead the second one (Millenials and their kids). Which to say the least is kind of a reversal of certain conventional narratives. There would appear to be a small light at the end of the hundred year tunnel.

Andrew G. Biggs said...

I'm not sure what you can infer about the costs of supporting different generations from the infinite horizon figures alone, or from how they change over time. Combined with the trust fund exhaustion date you can come to some conclusions about who's NOT a problem (at least from the perspective of trust fund accounting). Generational stuff gets hard to figure out in present value numbers like the actuarial deficit, since even if far-off cohorts produce the largest dollar deficits the fact that they're far off means the present value of those dollars is reduced.