The Case for Dynamic Scoring

For decades, government agencies have used conventional scoring to measure the fiscal impact of policies. However, this approach often leaves out critical information—particularly when it comes to understanding the broader, long-term effects of policies on different economic groups. That is because conventional scoring assumes that policies cannot change macroeconomic variables. The Congressional Budget Office (CBO), Joint Committee on Taxation (JCT), and other organizations have expressed interest in using dynamic scoring more often to get a better picture of economic realities.

Dynamic scoring offers a more comprehensive picture than conventional scoring by considering how policy changes affect economic behavior and, in turn, how these behavioral shifts feed back into the economy. Many economists and prominent policy wonks are strong proponents of dynamic scoring.

A Brief History of Budget Scoring

Budget scoring has been a key part of policy evaluation since the Congressional Budget Office was created in 1974. Conventional scoring was the default, focusing on immediate changes to revenues and expenditures while assuming that the broader economy remains unchanged. In other words, a tax cut is assumed to have no effect on GDP. This method is useful for providing simple, first-pass estimates but is insufficient for rigorously evaluating policy changes.

Conventional scoring takes a static view of the economy, assuming that key variables like employment and investment remain constant. For example, a tax cut would be evaluated solely by how it changes government revenue under current economic conditions, ignoring any changes in behavior that the policy might trigger.

This narrow approach is particularly problematic when it comes to understanding distributional impacts. Without accounting for shifts in labor supply or investment, conventional scoring may overestimate the cost of a policy or misrepresent its benefits. The result is a picture of policy effects that is, at best, incomplete and, at worst, misleading.

Indeed, framing policy effects with a conventional score fails on a key dimension. The stated goal of many policies from tax cuts to innovation subsidies is to increase output, investment, and so on. And yet, CBO and JCT typically rule that out by assumption, which makes it very difficult to evaluate policies on their own terms.

By the 2000s, it became clear that conventional scoring wasn’t capturing the full picture, particularly in cases where policies had significant economic effects over time. In 2015, the House of Representatives introduced a formal requirement for dynamic scoring for major legislation—typically bills expected to impact revenues or spending by more than 0.25% of GDP. Unlike conventional scoring, dynamic scoring includes the ripple effects of policy changes on economic behavior, such as labor supply, investment, and productivity.

Dynamic scoring offers a much broader view than static scoring by factoring in how individuals and businesses will respond to policy changes. For example, a tax cut doesn’t just reduce government revenue—it might also encourage businesses to invest in new equipment or hire more workers, leading to higher productivity and wage growth. Over time, this increased economic activity can offset some of the initial revenue loss, providing a more accurate measure of the policy’s true cost. As a particular example of that, the 2017 Tax Cuts and Jobs Act was initially scored as costing around $1.5T, but dynamic scoring shaved $300B off that bill.

Dynamic scoring is particularly valuable when trying to assess how policies impact different income groups. For instance, conventional scoring might show that a corporate tax cut primarily benefits wealthy shareholders, but dynamic scoring reveals that long-term wage growth could benefit workers across all income levels. The extent to which this is true depends critically on the underlying model.

The Importance of Dynamic Scoring for Distributional Analysis

Understanding how policies impact different income groups is central to crafting effective economic legislation. Dynamic scoring makes this possible by capturing the longer-term ripple effects of policies. Conventional models, by contrast, often only look at immediate, direct impacts.

Take tax policy, for example. Conventional scoring would estimate the impact of a corporate tax cut solely by calculating the short-term loss in revenue. But dynamic scoring would also take into account how that tax cut might encourage investment, spur wage growth, and ultimately improve the financial position of lower- and middle-income workers over time. In other words, dynamic scoring allows us to see who truly benefits from the policy—and by how much-–rather than simply looking at who pays today.

Consider the importance of dynamic scoring for a topic like immigration reform. This is the primary example considered by economists Doug Elmendorf, Heidi Williams, and Glenn Hubbard in a recent paper. Under conventional scoring, a policy to increase green cards for high-skilled STEM workers would seem to increase the federal deficit, as it focuses only on the cost of providing public services to those workers. But when dynamic scoring is applied, the analysis shows a significant reduction in the deficit. Why? Because high-skilled immigrants contribute more in taxes than they receive in government benefits, making the policy a net gain for the federal budget.

This highlights the core strength of dynamic scoring: it provides a clearer, more comprehensive understanding of how policies play out over time. By capturing the full economic impact of immigration reform—both in terms of contributions to the workforce and long-term tax revenues—dynamic scoring paints a much more accurate picture than conventional methods.

More generally, there are other reasons to think that the current approach to scoring is lacking. For example, in a recent report for the Mercatus Center, Keith Hall highlights a deficiency in “integrity scoring,” which is how the government accounts for efforts to reduce waste, fraud, and abuse. Currently, scoring does not account for efforts to reduce waste, which means that the government implicitly favors maintenance of inefficiency when it scores federal programs. Although this is a separate issue from dynamic scoring, it is perhaps equally important.

Dynamic Scoring Should Be the Standard

Dynamic scoring isn’t just a technical upgrade to budget analysis—it’s a necessary tool for understanding the long-term and distributional effects of policies. As the CBO and JCT continue to refine their dynamic models, it’s time to make dynamic scoring the standard for evaluating major legislative proposals.

Without it, policymakers risk making decisions based on incomplete or misleading information, particularly when it comes to how policies impact different economic groups. Whether the focus is tax reform, immigration policy, or infrastructure investment, dynamic scoring provides the comprehensive analysis required to navigate the complexities of modern fiscal policy.

At the same time, there are serious challenges to integrating dynamic scoring. It requires writing down a model of the economy, which inherently makes policy analysis less transparent. Nevertheless, it is necessary for the government to invest in the infrastructure required to publish dynamic estimates as often as possible.