Stuart Altman

Content Section

From Newsweek

Why the CBO's Estimates Shouldn't Count for Much

By Jeremy Herb

Democrats breathed a sigh of relief Wednesday after the Congressional Budget Office declared the Baucus health bill would reduce the deficit by $81 billion. But the $829 billion CBO estimate is unlikely to be accurate if the legislation is enacted. While the CBO puts together its most comprehensive prediction possible, it often gets it wrong with big health legislation. It's not a lack of expertise or bias that causes the predictions to miss the mark, says Stuart Altman, a Brandeis University economist: "The problem is what we're asking them to do is impossible." Health-care legislation is the toughest to score accurately, says Robert Reischauer, former CBO director, because unlike laws that change the tax code or budget new building projects, there are often no data to examine.

When President Obama declared that health reform should have a $900 price tag, the CBO's scoring became a decisive, make-or-break factor. Their first $1 trillion estimate of the House bill stymied plans to move forward with a public option, while the less ambitious Baucus bill will likely sail through committee with the CBO's blessing. "[Obama] has basically said that they're going to be the scorekeepers," says Dean Baker, of the Center for Economic and Policy Research. "He's endowed them with a lot of power." Three reasons why we shouldn't put too much stock in the CBO's predictions:

Built-in skepticism. One reason the CBO was created in the 1970s was that Congress was excellent at predicting huge savings and low costs from new ideas, but too often the savings never arrived and the costs skyrocketed. The CBO helped curb that practice, but critics say the organization's built-in skepticism has moved it too far in the other direction. Their model is "Show me the beef," says Reischauer, currently president of the Urban Institute. But that skepticism also makes it more difficult to pass large-scale legislation when the CBO doesn't give credit for generating savings. In 1994, the Clinton administration threw fits when the CBO scored its employer mandate as a tax. If today's system were in place back in the 1960s, says Harvard medical professor Robert Blendon, "nobody believes Medicare would have been passed."

The 10-year estimate. The CBO's estimates force legislators to find funding for their bills 10 years out. The problem here is twofold. First, the CBO is trying to predict economic realities in 2019 (did anyone have a clue of the 2008 economic crisis back in 1998?). "The ability [for economists] to project out two years is hard," says Altman. "Once you're out past two or three years, you're in never-never land." The CBO must also assume current laws will still exist, giving Congress a chance to game the system. In the Baucus bill, for example, the Medicare sustainable growth rate, which sets the rates doctors are paid by Medicare and is adjusted each year, will supposedly drop 25 percent in 2011. No one believes this will happen, but the CBO must score the bill as though it will.

No evidence. When legislation creates new kinds of health models, or even expands a local model to a national one, the CBO often has little or no evidence for its projections. With health legislation, as opposed to, say, building new bombers, the CBO must try to determine costs for both the system and how changes would affect human behavior. When Congress proposed a "doughnut hole" in Medicare benefits coverage, something no commercial policy had ever used, the CBO had to guess how the system would react. It overestimated the cost by about 35 percent, according to some estimates. In the current reform debate, there's disagreement between the CBO and the White House's Office of Management and Budget over how many billions can be saved by tweaking Medicare benefits, with the CBO coming in on the low end of the spectrum. Who's right? The safe bet is that no one knows.

With Dan Loeterman

View As Single Page