The bean counters who could kill health care reform.

Everywhere you look, health care reform seems to be chugging along. Insurers and drug companies are visiting the White House to show solidarity with President Obama. House Democrats are promising to pass a bill by July 31. But, if you talk to senior staff in the administration or on Capitol Hill, you'll detect anxiety over one tiny agency--an agency that helped kill health care reform in 1994 and has the power to do so again.

That agency is the Congressional Budget Office (CBO), which is just now weighing in on the debate. When Congress writes a bill, the CBO is the agency that determines how much implementing it will likely cost. And that's no small matter. For every extra dollar in new expenditures that the CBO projects, Congress must find a new dollar in revenue--or learn to live with a new dollar in deficits. Back in 1994, the CBO decided that paying for universal health insurance under the Clinton plan would cost more than the administration thought it would. That forced Bill Clinton and his allies to propose additional regulations on insurance prices, incurring a political liability that helped seal the plan's fate.

The CBO hasn't rendered an official verdict on health care reform this time around, because, officially, there's still no plan on which to render a final verdict. But the CBO began producing preliminary projections about three weeks ago, based on rough proposals that congressional staff have submitted. And, according to several sources familiar with the estimates, the news isn't quite what Obama and his allies were hoping to hear.

The good news for reformers is the CBO's determination that expanding health-insurance coverage would cost a lot less than many outside experts had predicted. Instead of a politically daunting $1.5 trillion, the CBO figures the price tag will be closer to $1 trillion, at least under certain parameters. But the reason for the lower estimate is a bit unsettling. Even with a requirement that everybody obtain insurance--a so-called individual mandate--the CBO assumes a that between a quarter and a third of the uninsured still woulnd't have coverage. That would leave the country short of universal coverage, the goal Obama and his allies have repeatedly cited.

Other preliminary judgments from the CBO are causing more consternation. To help defray the cost of expanding coverage--and to help make medical care more affordable generally--reformers have proposed creating electronic medical records, studying which treatments work best, and taking other steps that would make the business of health care more efficient while cutting down on unnecessary medical treatments. The CBO is less optimistic these moves will save money. What's more, the CBO may determine that, for accounting purposes, the money individuals spend on health insurance--or, at least, some portion of it--should count as part of the public revenue stream. That would give critics more grounds for labeling the Democrats' plan a costly expansion of government.

If that judgment sounds familiar, that's because it is. The CBO took nearly the same positions back in 1994--a fact not lost on either the White House or congressional leaders, who have communicated their concerns publicly and privately. One apparent purpose of bringing industry leaders to meet Obama this week was to showcase the potential for cutting costs; see, the administration seemed to be signaling, even the health care industry thinks it can save money by becoming more efficient. But it will take more than such appearances to move the CBO.

As it happens, one reason so many observers (this writer included) have been optimistic about reform's prospects this year has been a belief that the CBO wouldn't play the same curmudgeonly role it did during the Clinton-care fight. In 2007 and 2008, the agency came under the direction of Peter Orszag, who famously made health care reform a personal and institutional obsession. Worried that the rising cost of medicine would create a serious fiscal crisis for the country, he used his perch to sound the alarm--and used his agency to grease the wheels of change. He nearly doubled the number of experts studying health care and, convinced that their reading of the academic literature left them with an unnecessarily narrow and dated perspective, pushed them to explore what was happening out in the real world.

Like all bureaucracies, though, the CBO has its own inertia--and its own set of institutional concerns. Asked to do the impossible, thankless task of predicting the future based on ambiguous, incomplete evidence, the CBO finds itself under constant attack from critics--left and right, up and down in the Washington power hierarchy. Until recently, the CBO's big worry was conservative supply siders, who kept insisting the agency use something called "dynamic scoring" to prove that tax cuts could help pay for themselves. Going out on a limb is hazardous in this environment. The best posture is a claim of scientific integrity, backed by a strict reading of peer-reviewed academic literature. So, when Orszag nudged them to change, sometimes the careerists nudged back.

Under Orszag's successor, Doug Elmendorf, the CBO has continued and even intensified some of the study habits Orszag instilled, while working furiously to keep pace with the packed congressional agenda. But the agency still seems inclined to produce estimates that strike many outsiders as overly cautious. To take one example, electronic medical records make it a lot easier to compile data on which treatments work, which--in turn--will let you cut down on the use of ineffective ones. In other words, there is reason to believe the two innovations would save more money in combination than they might separately. But, according to people familiar with the CBO's thinking, the agency seems skittish about giving extra credit for the way the two reforms would interact. Harvard economist David Cutler, a key architect of Obama's campaign plan, just published a paper outlining the potential to save the federal government $600 billion over ten years--about half of what it would cost to pay for expanding insurance coverage. Again, the CBO seems skeptical.

Nobody disputes that there is room for honest intellectual agreement; plenty of smart people dispute claims about savings that reform will generate. But it's an open question whether the CBO takes skepticism too far--and whether such a super-strict reading of the evidence really serves the public interest. As Obama himself frequently points out, exploding health care expenses are going to cripple the government and, eventually, the country. If the CBO's estimates or the decision to classify premium payments as government revenue end up paralyzing lawmakers, effectively snuffing out reform again, it's quite likely that the country's fiscal situation will be worse--not better--ten or 20 years hence.

Of course, lawmakers wouldn't need such favorable scoring if their proposals were a little bolder. If Obama and Congress could get more aggressive about steering reimbursements away from doctors and hospitals that provide poor care, the CBO would undoubtedly credit those moves as generating additional savings. And, if reformers were willing to find more tax revenue--by, say, declaring that premiums for the most generous health plans were no longer fully exempt from personal income taxes--they could raise the money they need. But those are tough political tasks. Providers of medical care don't like their incomes going down; taxpayers don't like their liabilities going up. In the end, Obama and Congress will have to live within certain constraints that the CBO sets--adjusting their proposals to coincide with the agency's predisposition or scaling back what they hope to do. The result might be an imperfect law, but, as the Clinton-care veterans will tell you, sometimes that's better than no law at all.

Editor's Note: This article has been updated with new reporting about the CBO's projections.

Jonathan Cohn is a senior editor of The New Republic.