$44 TRILLION?….Yesterday I wrote a post about a Financial Times article claiming that a Treasury estimate of long-term deficits had been suppressed by the Bush administration. Today, via Instapundit, Lynxx Pherrett looks at the transcripts of the interviews that FT did and says their story was wrong.

Lynxx’s post combined with the FT interviews turns out to be pretty long and complex, but here’s my best attempt at a summary:

  • Social Security and Medicare deficits are normally calculated over a 75-year period.

  • The Treasury study in question was related to private accounts for Social Security and Medicare and the authors were concerned that the traditional 75-year horizon made proposals for private accounts look worse than they are. Basically, they say that the costs of private accounts are all within the 75-year period but some of the benefits fall beyond this period. So traditional accounting makes private accounts look more expensive than they are.

  • To fix this, they recalculated the deficits over a very long time horizon. When they did this, the total deficits naturally went up a lot. That’s where the $44 trillion number comes from.

  • This new analysis was indeed left out of the 2004 budget. The authors are circumspect about why, suggesting that the new Treasury team confirmed earlier this year (after Paul O’Neill was fired last November) wasn’t comfortable with the new approach and needed more time to digest it. On the other hand, it’s also likely that the new team’s decision was partly political since Bush was in the middle of trying to get his tax cut passed.

So what to think? The FT piece pretty clearly implied that the report was suppressed for political reasons, and while this is a reasonable guess it’s not really backed up by the interviews. They seem to have overreached on this one.

But what about that longer time horizon? Is it legit? I doubt it. The longer horizon seems to have been set up for an overtly political purpose (justifying private accounts), which makes me pretty cautious about accepting it. What’s more, I’m generally skeptical about even the 75-year horizon that the Social Security trustees use. In fact, when I blogged about this yesterday I restricted myself to a 30-year horizon.

Why? First, because too many things can happen in 75 years. Policies change, tax law changes, medicine advances, etc. There’s just not much point in projecting current policies over such long periods.

Second, because the longer the horizon the more sensitive you become to very small changes in assumptions. If you assume economic growth of, say, 2.75% instead of 3%, it makes a moderate difference over ten years, a bigger difference over 30 years, and an enormous difference over 75 years. Compound interest, you know. (The trustees are well aware of this, by the way, and provide multiple estimates based on different assumptions. The “middle” estimate is the one that gets all the attention.)

So there you have it. Fascinating, eh?

UPDATE: Max says the infinite horizons (“generational accounting”) are hooey. His reasons are roughly the same as mine except far more learned.

He also makes the point that Medicare is not a generational problem, it’s a problem with the healthcare system in general. I quite agree, and treating Medicare like some kind of future annuity is silly. Healthcare is an ongoing expense, and all that really matters is (a) how much will it cost each year and (b) how will we pay for it each year?

Our ideas can save democracy... But we need your help! Donate Now!