The Congressional Budget Office provides nonpartisan and objective analysis to help Congress make informed budgetary and economic decisions. As part of that work, our analysts often develop models that are simplified representations of relationships between variables in the economy and the federal budget.
CBO's models differ in their complexity, from spreadsheet calculations to large-scale simulation models that account for a broad set of factors. The number of models produced each year varies. Many of the models are continually developed, refined, and maintained by the agency's analysts; new models are created each year in response to the interests of Congress.
As part of our ongoing commitment to bolstering transparency, we make our analyses—including the models and data underpinning them—accessible to policymakers and the public. That effort includes explaining the agency's analytical methods, releasing data, posting models and computer code, providing tools to estimate the effects of policy alternatives, and developing data visualizations. To improve our models, we subject them to reviews by experts outside of CBO, who provide feedback.
When practicable, we post data, computer code, and documentation associated with our models on the web-based platform known as GitHub. Users can examine those resources and explore outcomes under alternative scenarios.
We have also created several easy-to-use modeling tools for exploring alternative legislative and policy scenarios based on our models, examples of which are discussed below. Those accessible modeling tools, in addition to the more advanced modeling resources posted on GitHub, are listed on CBO's Models page.
Resources Available on GitHub
GitHub is an online platform that allows people and organizations around the world to share their computer code and other data. In May 2020, CBO began sharing data, code, and other documentation on that platform.
The agency currently has six repositories hosted on GitHub, each containing documentation that provides a detailed overview of a project's codebase. One repository contains the agency's capital tax model, called CapTax, which CBO uses to estimate how changes in tax law affect businesses' incentives to invest in capital, such as equipment, buildings, and other assets. The model can help users understand how federal tax laws may affect business investment by allowing them to test, compare, and visualize the impact of tax rules on different types of capital.
Another repository enables users to replicate the evaluations that CBO regularly conducts to assess the accuracy of its projections of key budget components, such as outlays, revenues, deficits, and debt.
CBO's GitHub page also features a model that helps the agency's analysts estimate how quickly private health insurance premiums will change over time. The premium growth model allows users to replicate CBO's projections of premiums and see how those projections are affected under alternative scenarios. The results derived from that model are used in CBO's health insurance simulation model, called HISIM2, which simulates how people respond to changes in health insurance policy.
Modeling Tools Available on CBO's Website
CBO's website features tools that allow users to adjust policy assumptions or economic conditions to see how outcomes of certain analyses may change. Reflecting CBO's modeling, the tools help policymakers and the public understand the bases of the agency's projections and the potential effects of policy choices.
For example, as part of the agency's distributional analysis of the 2025 reconciliation act (Public Law 119-21), CBO published an interactive tool that allows users to explore how the law will affect economic resources available to households grouped by income deciles (or tenths). Users of the tool can adjust the years and types of resources displayed—such as federal taxes, cash transfers, in-kind transfers, and states' fiscal responses—to see how the law's effects on household resources would change over time.
Additional interactive tools enable users to estimate the debt-service costs associated with specific legislative proposals in relation to CBO's baseline budget projections, to see how changes in economic conditions might affect the federal budget, or to examine a "waterfall" model that demonstrates how the agency projects federal discretionary spending over a 10-year horizon.
CBO's interactive force structure tool allows users to adjust the Department of Defense's budget to see how funding levels affect the size and mix of military forces—or, conversely, to modify force levels to see how those changes would affect the budget.
Going Forward
As part of our ongoing efforts to enhance the transparency of CBO's work, we continually assess opportunities to post additional modeling resources and develop new tools, while adhering to strict federal standards to ensure that sensitive information is not released.
We will continue to develop new modeling tools, update the computer code for our existing models, and publish information that explains the models and methods used to produce cost estimates and other analyses. The agency's staff routinely discusses those models and methods with Members of Congress, their staff, and external organizations to gain feedback that helps maintain and improve the quality of CBO's work.
Phillip L. Swagel is CBO's Director.