CBO's Blurry Medicare Estimates

Look closely at the Congressional Budget Office (CBO) estimates of the Senate's Medicare bill (S. 1). The first estimate, released June 17, was based on early specifications by the Finance Committee. It had a $41 monthly premium (in 2006), $151 billion in total premium collections over 10 years, and $455 billion in total benefit costs.

The second estimate, released a day later, incorporated several changes made to the legislative proposal by the committee. After those changes, the monthly premium in 2006 was reduced to $35 and it remained lower throughout the rest of the 10-year budget period. However, the total amount of premiums collected was unchanged at $151, identical in every year.

How could that be? Well, maybe CBO estimated more beneficiaries would enroll if the premium was lowered, a reasonable assumption. Price times quantity equals cost (or in this case Premium Rate times Number of Enrollees equals Premium Collections). So although it's sort of eerie that the total premium collections were identical in every year, it's mathematically possible that the total could still be $151 billion.

With an increased number of enrollees, surely the total benefit costs would go up, right? Actually, no. CBO estimated that the 10-year benefit costs fell slightly from $455 billion to $452 billion between the June 17 and June 18 estimates.

Hmmm. Maybe the benefit package was made less generous. Actually no, it wasn't, at least not in any easy-to-see way. You could look up the legislative language (I did). Didn't see any benefit reductions.

Could there be some other technical change to the legislative language that could simultaneously explain lower monthly premiums, identical annual premium collections, and slightly lower benefit costs? Some benefit change that is not obvious or apparent? Actually, yes. Chances are, some combination of policies changed enough to cause the strange set of corrections to the CBO estimate.

Has CBO explained any of this? No. Except for the changes to the numbers, the write up of the two CBO estimates is identical.

Should CBO explain such a change? Yes.

In the old days (the early 1990s), CBO produced bound reports on major health care estimates, with detailed tables and chapters explaining the figures and other analysis lawmakers might consider before enacting a bill. More recently, they produced written cost estimates, which explained the estimates thoroughly (although they did not usually contain additional analysis or perspective).

Last year, CBO's estimate of the House Medicare bill dispensed with the "Basis Of The Estimate" section, the part where they (would have) explained how the estimates were done.

This year, we don't even have the crudest explanation of why a major Medicare estimate changed, fairly significantly, from one day to the next.

I know these sorts of estimates are hard, and the pressures are great. I used to help do them.

But the current CBO process smells fishy. Mysterious estimates changing without explanation. No published explanations.

These sorts of problems are just the tip of the iceberg for CBO. When curious minds look into why this year's Medicare bills are structured as premium-based, stand-alone drug plans, administered by risk-bearing private health plans with an unorthodox but tightly prescribed benefit package, they should look to CBO. That's the only sort of market-based Medicare proposal CBO has been willing to even try to estimate over the last several years.

It doesn't matter that many outside analysts believe such a benefit structure is impractical and unworkable. Not to mention costly. Solving the workability problems inherent in this sort of benefit could easily double the stated $400 billion federal cost.

But when Congress decided to make a deal, that's the type of benefit they had to use. No other Medicare reform has been analyzed. Not the Medicare Commission's premium support model. Not the FEHB-style system with a defined contribution/high option drug benefit outlined in the "Breaux-Frist I" bill. Not the Medigap drug option originally favored when the "Breaux-Frist II" bill was being drafted. Each time, Medicare reformers got the same reasons CBO couldn't help: It was too uncertain, too difficult to estimate, there wasn't sufficient time. But without a CBO estimate, those proposals were dead in their tracks.

So we're stuck with a stand-alone, premium-based, unorthodox but highly specified drug benefit delivered (hopefully) by private risk-bearing plans.

Oh well, it's not as if the nation's fiscal future is at stake. Hmmm.

Links:
CBO Estimate of S.1 as modified (June 18, 2003)
CBO Estimate of S.1 (June 17, 2003)
Centrist Policy Network Legislative Resources Page
Centrists.Org Will the Senate's Medicare PPO Program Work? (6/11/2003)

Posted by Jeff Lemieux at June 20, 2003 12:25 AM