TOPICS > Health

Cost Estimates Prove Key to Health Reform Debate

BY Lea Winerman  November 12, 2009 at 2:48 PM EST

U.S. Capitol; file photo

Whenever Congress wants to know how much new legislation will cost — or save — the government, it asks the CBO. Elmendorf will put a dollar figure on the health care bill Reid hopes to bring to the Senate floor as early as next week.

It’s been an uncommonly high profile year for the CBO, founded in 1974 to provide objective budget analyses for Congress. The approximately 235 analysts and economists who work for the agency generally crunch numbers in relative obscurity. But this year, as President Obama has vowed to keep the price tag of a massive health care reform overhaul under $900 billion, CBO has become the arbiter of that goal, with politicians and even the public eagerly awaiting its estimates, or “scores,” of each new proposed bill.

The Speaker of the House and the President Pro Tempor of the Senate appoint the CBO director, but the agency is nonpartisan, lending credence to its estimates.

“It’s specified that [CBO] should be independent of Congress and objective, and that’s very important in terms of the credibility that CBO cost estimates have,” says Allen Schick, an expert on the federal budget at the Brookings Institution.

But the agency’s high profile this year has also come with criticism, as some analysts and policymakers have questioned its power and its track record in providing accurate cost estimates for health care legislation.

What goes into a cost estimate?

CBO analysts are required by law to provide a cost estimate for every bill reported by a congressional committee, but the agency’s work — and influence — sometimes extends beyond that, Schick says.

“In many pieces of legislation, particularly when they’re important, there’s a more interactive process between the CBO and members of Congress,” he explains. “Members of Congress will bounce off CBO staff what will happen if the language is this or that, if we expand or contract the numbers who are eligible, [...] if we tweak the legislation one way or the other.”

So the agency’s cost estimates can influence the shape of legislation, as well as its eventual chance for passage.

Health reform legislation is particularly complex and difficult to analyze because it relies on predicting how people and companies will behave in the future, in response to changing laws and new public programs.

CBO analysts begin by gathering the relevant data, explains former CBO director Rudolph Penner, now at the Urban Institute in Washington, D.C. For example, because the legislation now under consideration in Congress would provide subsidies based on income levels, analysts must have accurate information about the income distribution in the United States, as well as insurance coverage rates at those different incomes.

Then, the analysts develop a model to determine how the changes in the health care system will affect consumers’ behavior, as well as the behavior of doctors, health insurance companies, and other players in the system.

“The root of the model is that [...] we think [people] are making purposeful choices about how to get health-care insurance that are sensitive to the cost and quality of health insurance,” Elmendorf said in an interview with the Washington Post in September. “There are estimates that economists and health experts have developed over years of the way that individuals and companies respond to different options. We build those estimates [...] into the model.”

Then analysts use that model to estimate the budget impact of providing different levels of subsidies, offering a public option, requiring individuals to buy insurance or employers to provide it, and all of the other aspects of the bill.

Accuracy and controversy

It’s an extraordinarily difficult task. Which is why, Schick says, the CBO may be credible but it is not always accurate.

Health care economist Jon Gabel, of the National Opinion Research Center at the University of Chicago, puts it more strongly: “The CBO has ‘mission impossible,’” he says.

Ideally, Schick explains, analysts estimating costs in an uncertain future would come up with a range of possibilities rather than a point estimate — determing, for example, that a bill’s cost might fall somewhere between $600 billion and $1 trillion.

“But Washington has no stomach for ranges, it wants to know exactly what it will cost,” Schick says. “And the exact number is always the wrong number.”

In fact, Elmendorf took pains to point out in a Washington Post interview that the CBO is well aware of and tries to convey the uncertainty in its estimates.

“I have rarely given an interview or testimony or written a letter on health care where the word ‘uncertainty’ didn’t appear many times,” he said, adding “I think the caveats we apply to our analyses are often left behind.”

This summer, Gabel and other analysts and reform advocates charged in the op-ed pages of several U.S. newspapers that the CBO was particularly likely to overestimate the costs and underestimate the savings of health care bills. Gabel cited research that the CBO had, for example, overestimated the cost of the 2003 Medicare Part D drug benefit program by nearly 40 percent, and had underestimated the initial savings from the 1997 Balanced Budget act, which reduced Medicare spending, by more than 100 percent in the program’s first two years.

The problem, Gabel contends, is that CBO comes up with its savings estimates partly by looking at how much money similar programs have saved in the past. But when there aren’t enough historical examples to go by, those savings are unknown.

That’s true for many aspects of health care reform now being considered, such as efficiency savings from converting from paper to electronic health records, or from relying more on comparative effectiveness research that would determine the most effective and cost-effective medical treatments. “Too often,” Gabel wrote in an August New York Times op-ed, “‘unknown’ becomes zero.”

“I think they use cautious methods, and they’re underestimating the amount of savings that can take place,” he says.

But Penner argues that there’s good reason to be cautious. Overestimating costs, he points out, doesn’t jeopardize the federal budget. “That’s a happy outcome, and Congress can always use the unexpected savings to reduce the deficit or cut taxes or enhance benefits.”

And Schick points out that there have been times that the CBO cost estimates have been too low rather than too high, such as for the 1996 Welfare Reform Act.

Such questions of accuracy are more than academic, he says, because the CBO’s work can decide the fate of legislation: “Even though they’re technocrats rather than politicians, it’s evident from this year’s health care debate [...] that they have a very large impact on the legislation that emerges.”