How Does CBO Prepare Its Baseline Projections and Cost Estimates?

Posted by Mark Hadley on
February 7, 2018

Yesterday, the House Budget Committee held a hearing about CBO’s economic assumptions, baseline projections, cost estimates, and scorekeeping. I submitted written testimony and made some remarks; then I answered questions from Members, along with my colleagues Wendy Edelberg, an Associate Director for Economic Analysis at CBO, and Teri Gullo, the agency’s Assistant Director for Budget Analysis. This blog posting summarizes my remarks, which drew from the testimony and from three recently released documents:

I began by pointing out that the House and Senate Budget Committees are the scorekeepers for the Congress, crafting the budget resolution and enforcing the budget rules. CBO prepares its current-law projections and cost estimates of legislation to help the budget committees discharge those duties.

Some people think that CBO makes the rules we follow. But in fact, most of the rules that govern the formulation of those products are set in law. Others originated in budget resolutions, House or Senate rules, or conference reports that accompanied budget legislation. In addition, some have been developed with the House and Senate Budget Committees directly.

Some of the rules require CBO to develop estimates using specified assumptions. When that happens, CBO also provides information from alternative perspectives that the budget committees have found useful. And we would like to hear from the committees about information that we could add or present differently to be more helpful.

Baseline Budget Projections

CBO’s baseline budget projections include estimates of spending, revenues, the deficit, and the public debt. CBO’s economic forecast covers the major economic variables—gross domestic product, the unemployment rate, inflation, and interest rates—along with a broad array of other economic measures. The baseline and the forecast are CBO’s best estimates of how the budget and the economy would evolve if existing laws generally remained in place. They are developed together to be consistent with each other, and both cover the 10-year period used in the Congressional budget process.

The budget committees typically use CBO’s baseline as the starting point for budget resolutions and as a neutral benchmark to assess the budgetary effects of some legislation. Baselines provide consistency and add stability to the budget process. They provide consistency because each cost estimate is built on a common set of assumptions. That helps the Congress consider the relative costs of competing proposals on the same subject and of proposals on different subjects. Consistency also helps the Congress ensure that legislation complies with the targets set in the budget resolution. And baselines add stability because they do not change with each new piece of economic and budgetary information; instead, that information is included when we produce the next baseline.

Cost Estimates

Cost estimates describe the budgetary effects attributable to a single piece of legislation. Each cost estimate tells a concise story about a legislative proposal’s likely effects on federal outlays or revenues, compared with what would happen under current law (that is, with what would happen if the proposal was not enacted).

CBO’s cost estimates are advisory. The budget committees enforce the budget rules, and they are not required to use CBO’s estimates.

The Congressional Budget Act of 1974 requires CBO to prepare cost estimates for legislation that has been ordered reported by authorizing committees. We also prepare estimates, when requested, at other stages of the legislative process.

The Budget Act directs CBO to give priority to committees, particularly the House and Senate Budget Committees. Given CBO’s staffing, it generally is not possible to satisfy requests for cost estimates that do not come from the budget committees, the committees with jurisdiction over a proposed change in law, or the Congressional leadership.

Rules Governing the Baseline

Section 257 of the Balanced Budget and Emergency Deficit Control Act of 1985 defines CBO’s baseline projections and spells out the rules for making them. In general, for mandatory spending and revenues, we are required to assume that current laws will remain in place. For example, for individual income taxes, we assume that the changes scheduled under current law will occur. However, there are three important exceptions under section 257.

The first exception deals with expiring taxes and spending programs. Under section 257, CBO is required to assume that certain taxes and programs will continue beyond their statutory expiration. Those include excise taxes dedicated to a trust fund, such as the federal gasoline tax. They also include certain programs with current-year outlays greater than $50 million—for example, the Children’s Health Insurance Program. (That is sometimes called the $50 million rule.)

That first requirement has advantages and disadvantages. Consider the $50 million rule. On the one hand, because expiring programs are often extended, it improves the accuracy of baseline projections of the deficit. On the other hand, the rule makes certain extensions of programs appear costless because the cost of extension is already in the baseline. To provide more information, CBO and the staff of the Joint Committee on Taxation show how the baseline projections would change if the provisions that we must assume would continue were instead assumed to expire.

The second exception deals with entitlement authority. CBO is required to assume that entitlement programs, including Social Security and Medicare, will be fully funded and thus that all scheduled payments will be made. For example, CBO must assume that scheduled Social Security benefits would be paid even after the program’s trust funds were exhausted and annual payroll tax revenues were inadequate to fund those payments.

That requirement too presents trade-offs. On the one hand, a baseline that shows benefits being paid regardless of the resources in the trust funds may be more informative about the budgetary challenges facing the nation if policymakers want to maintain Social Security in its current form. On the other hand, it is less informative about the challenges facing the program itself. In our products, we present both the projected deficit if all benefit payments were made and the percentage by which benefits would be cut if no additional funding were provided.

The third exception deals with discretionary spending. Section 257 specifies how CBO must construct the baseline for discretionary programs. It requires CBO to use specified inflators to increase the most recent appropriation for each program over time. However, to project total defense and nondefense discretionary spending, CBO is required to apply the caps that limit funding for those categories through 2021. And again, CBO provides alternative paths in its reports—for example, a path without inflation for discretionary appropriations.

Rules Governing Cost Estimates

Four rules apply broadly to cost estimates. First, CBO must consider effects across the entire federal budget. Therefore, cost estimates sometimes include effects on programs other than those specifically addressed by the proposed legislation.

Second, CBO’s cost estimates generally do not include the interest costs that would result from increasing the deficit or the interest savings that would result from reducing it.

Third, CBO reports point estimates—that is, specific dollar amounts—of the budgetary effects of legislation because they are necessary for the Congress to enforce budget rules. (But CBO tries to indicate which factors underlying an estimate may be particularly uncertain.)

Fourth, for appropriations and authorizations of appropriations, the benchmark against which cost estimates are made is current law. But that benchmark is not the baseline, in which—for individual accounts—spending in future years grows with inflation. For many programs, the benchmark is zero because there is no appropriation or authorization for future years under current law.

Two sometimes controversial principles relate to a limited group of cost estimates. Those principles involve “dynamic scoring” and when an activity should be considered federal for budgetary purposes.

CBO always tries to estimate the effects that legislation would have on people’s behavior—for example, changes in how many people would enroll in a program. But the effects of legislation on the economy as a whole, and in turn on the federal budget, are taken into account only when CBO produces a dynamic score for major legislation. Nevertheless, when CBO prepares its baseline economic projections, the agency incorporates the macroeconomic effects of all enacted legislation.

And when the Congress considers legislation that would establish a new program or mandate a new activity, CBO must decide whether to treat the associated cash flows as federal transactions. For most legislation, that determination is straightforward because it would be carried out by federal agencies. But when the determination is less clear, we look to the guidance found in the 1967 Report of the President’s Commission on Budget Concepts.

I ended my remarks by noting that CBO tries very hard to provide information that is useful to the Congress. As we look for ways to serve its needs better, we welcome the budget committee’s continued guidance and input.

Mark Hadley is CBO’s Deputy Director.