How the CBO Scores a Bill: A Step-by-Step Explanation

The Congressional Budget Office (CBO) is Congress’s official scorekeeper for the fiscal impact of proposed legislation. When a bill is referred to the CBO, analysts follow a systematic process to estimate its 10-year cost or savings: establishing a baseline, projecting the real-world behavioral and economic effects, and weighing competing methodological choices. The same bill can receive different scores depending on when it is scored, what baseline is used, and whether “dynamic scoring” is applied.

The Baseline: Your Starting Point

Before the CBO can measure the impact of a bill, it must establish a baseline—a projection of what revenues and spending will be over the next 10 years if no new law passes. This baseline incorporates current law, including expiring provisions, scheduled tax increases or decreases, and inflation adjustments built into existing policy.

For example, if a tax credit is set to expire in two years under current law, the baseline assumes it expires. A bill that extends that credit is then scored as costing less than it would appear to, because the baseline already accounts for the revenue loss from expiration. If a bill makes the credit permanent, the CBO calculates the incremental cost relative to the baseline trajectory.

The baseline is set at a specific point in time—usually when the bill is referred to the CBO—and reflects the Budget Office’s own economic forecasts for inflation, GDP growth, unemployment, and interest rates over the 10-year window. If the economy shifts sharply or Congress passes other legislation, the baseline may be recalculated, potentially changing the score of the same bill.

Step One: Reading the Bill and Defining the Scope

The CBO begins by analyzing the legislative text in detail. The analysts identify every provision that affects federal revenues or spending, determine which agencies or trust funds are affected, and establish the bill’s effective date and duration. A bill might affect multiple programs—Medicare, Medicaid, defense contracts, the Internal Revenue Service—each requiring separate projection work.

The CBO also identifies potential loopholes, phase-ins, sunset dates, and interactions with existing law. If a bill repeals one program while creating a similar one, the scorers determine whether beneficiaries will switch smoothly or leave the system entirely, changing the net fiscal impact.

Step Two: Quantifying Direct Costs and Revenues

Once the scope is clear, CBO analysts project the direct fiscal impact. For a spending bill, they estimate how many beneficiaries will use the new program, what the average benefit payment will be (often based on similar existing programs or pilot data), and how these figures will change over time. For a tax provision, they estimate how many taxpayers will be affected and the average revenue change per taxpayer.

These projections rely on historical data, program records, and economic relationships the CBO has documented over decades. For example, if a bill increases the earned income tax credit, the CBO draws on past experience with similar credits to estimate take-up rates, average benefit sizes, and the distribution across income groups.

Step Three: Static vs. Dynamic Scoring—The Methodological Fork

This is where the same bill can receive different scores.

Static scoring assumes the bill’s direct mechanical effects occur in isolation. If a tax cut reduces revenues by $100 billion, the static score is a $100 billion loss to the budget deficit. If a spending bill distributes $50 billion to beneficiaries, the static score assumes exactly that outlay.

Dynamic scoring goes further. It models how the policy change affects behavior and the broader economy. A large tax cut, for example, might stimulate additional investment, work effort, and entrepreneurship, raising economic growth and generating higher wages, which in turn raises tax revenues relative to the static baseline. Alternatively, a policy might reduce incentives to work or invest, lowering growth and revenues.

The CBO can produce both static and dynamic estimates for the same bill. Dynamic scoring is more ambitious and contestable—it requires assumptions about macroeconomic multipliers, labor supply elasticity, and other behavioral parameters that economists dispute. Static scoring is transparent but potentially misleading if a policy materially changes incentives.

Most CBO scores are static. Dynamic scoring is optional and typically used for major tax or spending bills with broad economic effects. When it is applied, the CBO labels the assumption clearly and explains its magnitude.

Step Four: Accounting for Behavioral Responses Within the Dynamic Framework

Even within a primarily static score, the CBO accounts for some behavioral changes—those with strong empirical support. For example:

• If a tax deduction is introduced, some taxpayers will increase their deductible spending to claim it; the CBO models that shift using historical elasticities.
• If a subsidy for a good is introduced, consumption will rise; the CBO estimates the demand elasticity from past data.
• If a program’s eligibility threshold is raised, some beneficiaries will be disqualified; the CBO estimates how many will exit.

These are microeconomic responses baked into the direct cost estimates. They are distinct from macroeconomic dynamic effects—effects on overall GDP growth, interest rates, and aggregate tax revenues—which are modeled separately and presented as an adjustment if dynamic scoring is applied.

Step Five: Scoring Across the 10-Year Window

Most CBO scores cover a 10-year period, as mandated by the Congressional Budget and Impoundment Control Act. This window was set decades ago and is now somewhat arbitrary—it captures most bill effects but misses long-term phase-ins and permanent entitlements fully mature.

The CBO projects year-by-year costs or savings, accounting for phase-ins, temporary provisions that expire, and changes to eligibility or benefit levels over time. A bill might cost nothing in year one (effective date is late in the fiscal year), rise to $10 billion annually by year three, stay flat, then drop if a sunset clause is triggered in year nine.

The total 10-year score is the sum of annual impacts. It is presented both as a total figure and as a breakdown by year and by affected program, allowing Congress to see both the cumulative impact and the trajectory.

Step Six: Sensitivity Analysis and Uncertainty Ranges

The CBO rarely presents a single score as definitive. Instead, it publishes a range or notes key uncertainties. For example, a bill creating a new health insurance program might be scored at $50 billion over 10 years, but with a note that the estimate could range from $35 billion to $70 billion depending on take-up rates and cost trends.

The CBO also performs sensitivity analysis: “If the labor supply response is twice as large as we assumed, the score would change by $X.” This transparency helps Congress understand where the uncertainty lies and whether the estimate is robust to different assumptions.

Step Seven: The Published Score and Limitations

The final CBO score is published as a memo, typically 10–30 pages, with a headline figure (e.g., “$25 billion deficit increase over 10 years”) followed by detailed tables and explanatory text. The memo documents every assumption, data source, and methodological choice.

However, CBO scores have real limitations:

• They cannot predict Congress’s future behavior. If a bill’s funding depends on Congress passing a follow-up appropriation, the CBO assumes that will happen—but Congress might not.
• They rely on historical data, which may not predict future outcomes if circumstances change.
• Long-term projections are inherently uncertain. A 10-year estimate is more reliable than a 30-year projection, but both involve substantial guesswork.
• They do not measure non-fiscal effects: environmental impact, health outcomes, national security, or distributional effects (who bears the cost and who receives the benefit), though the CBO often publishes supplementary analysis on these.

When a Bill Is Scored Multiple Times

As a bill moves through Congress—from committee markup to floor amendment to reconciliation with the Senate version—it may be scored multiple times. Each version receives a new CBO estimate, and lawmakers often use differences in scoring as evidence that a provision does or does not “pay for itself.”

A common tactic: A legislator proposes an amendment, requests a score, and announces that the amendment “costs less” than the original bill. This is true in a narrow sense—the incremental amendment has a smaller score—but it may still increase the budget deficit, because both the original bill and the amendment increase deficits relative to current law.

The sequence of scoring can also matter. If Congress passes one bill that changes the baseline, and then refers a second bill to the CBO, the second bill is scored against the new baseline. The same legislative idea might cost more or less depending on what came before—not because the idea changed, but because the baseline shifted.

See also