Behind Washington's Numbers Game

ASK most Americans to pinpoint the center of power in Washington, and few would lead you to the corner of 3rd and C streets. But here, in a frumpy building on the fringes of Capitol Hill, is one of Washington's most influential forces: the Congressional Budget Office (CBO).

For this 20-year-old nonpartisan agency, the low-profile address is not a problem. In fact, the more attention CBO gets, the harder its job becomes.

Each time the president or a congressional committee drafts a spending bill, CBO is responsible for "scoring" it, or projecting its cost. As lawmakers pursue historic budget reforms, these once-routine calculations have suddenly become Herculean struggles with staggering social and political consequences.

This month, as Congress submits a proposal to balance the federal budget in seven years, CBO faces one of the greatest tests of its objective mandate. If it bows to political pressure, or creates that impression, CBO could lose its credibility. Without it, there would be few dispassionate voices left in American politics.

"Without CBO, it would be hard for Congress to put together its own budget," says CBO director June O'Neill. "What you've seen this year would be impossible. Congressional committees would choose the numbers they like the best, and consequently, they wouldn't be believed."

According to Robert Reischauer, Dr. O'Neill's predecessor, "The institution's whole ethos and reason for existence would be threatened if it strayed from the mission of being objective and nonpartisan."

The first signs of CBO's stormy future emerged during last year's debate about President Clinton's universal health-insurance plan. For the first time, critics from both parties tried to sway CBO's decisions about specific accounting policies - each side charging that a judgment against them would prove that CBO was partisan.

In the end, the argument was diffused when CBO concluded that the Clinton plan's "regional insurance alliances" should be classified as government agents, not private entities.

This bookkeeping strategy added billions to the proposal's cost and helped to convince many members of Congress, especially on the Republican side, that CBO was nobody's dupe.

'Fictional' nonpartisans

But even before the health-care debate, criticism of CBO had begun to surface. Before they conquered both houses of Congress last November, Republicans like House majority leader Dick Armey of Texas often complained that CBO's accounting principles had a liberal bent. "One of the biggest fictions in this town is that the CBO is nonpartisan," he said.

The Republican complaint, which still simmers today, centers around CBO's methods for calculating the cost of tax cuts. Currently, CBO treats tax cuts as net losses in government revenue: that is, for every dollar given in tax breaks, another dollar must be saved elsewhere in the budget.

But supply-side Republicans like Representative Armey espouse a method of accounting known as "dynamic scoring."

They contend that some tax cuts, particularly cuts in capital-gains tax rates, spur economic growth and raise overall tax revenues. Thus, tax cuts should not require equivalent cuts in other spending programs.

Shortly after the election, Republicans set out to replace Mr. Reischauer, a Democratic appointee, with an economist who shared their supply-side views. Their eventual choice, economist June O'Neill, had a solid conservative track record.

But to the surprise and chagrin of many Republicans, O'Neill has refused to embrace dynamic scoring, thus complicating Republican attempts to balance the federal books while providing a $245 billion tax cut. In addition, both parties have taken issue this year with CBO's conservative projections of economic growth.

No purges, yet

Nevertheless, Republicans have not purged CBO's staff as they once threatened to do, and at times have even praised the institution.

In June, a CBO estimate concluded that Clinton's proposal to balance the budget in nine years would fall short by about $200 billion - a calculation that gave Republicans enough political cover to ignore the president's proposal entirely.

To CBO staff members, the fact that their estimates draw fire from both sides of the political spectrum is an indication that they are doing their job well.

"Usually when the dust settles, there's a general feeling that what we've done is trustworthy," O'Neill says. "CBO has established considerable credibility with Congress and the outside world as well. People believe that our estimates are our best shot."

Yet the job of analyzing budget proposals has become inescapably political. According to David King, a congressional scholar at Harvard University's Kennedy School of Government, many of the assumptions CBO is forced to make in determining the costs of programs "cut to the core of differences between liberals and conservatives."

For instance, in order to make cost projections, he says, CBO must predict how welfare recipients might react to work incentives, or what percentage of Medicare recipients might opt for managed care.

In this regard, CBO's challenges will only grow. Starting next year, as part of the unfunded mandates law signed by the president, CBO will have to determine how any new government regulation will impact state, local, and tribal governments, and the private sector.

Although the agency's budget has been increased from $20 million to $23.5 million, and staffing from 200 to 230, the new responsibility thrusts CBO into an uncertain role.

Tough number-crunching

Privately, CBO staffers say it will be next to impossible to complete cost-benefit analyses for regulations like the Clean Air Act, whose benefits are hard to measure in dollar terms. In addition, they say, there is no wealth of data on the cost many regulations would impose on private companies.

"CBO tries to be very prudent, to base its estimates on accepted practices and knowledge of procedures in political science and economics," Reischauer says. "The danger is that [these] responsibilities are beyond [CBO's] capacity to fulfill. Some estimates required under the unfunded mandates law are beyond the state of the art."