The House has instructed the Joint Committee on Taxation and the Congressional Budget Office to factor in the macroeconomic effects of tax law changes when calculating the official budget score of revenue bills. But are existing models up to the task of what’s commonly called dynamic scoring? A group of experts assembled today by the Tax Policy Center and the Hutchins Center on Fiscal and Monetary Policy at Brookings generally agreed that current models have serious limitations, but disagreed on whether they are good enough to produce acceptable revenue estimates for major tax bills. This arcane debate over how congressional experts score tax bills is enormously important. Until now, the scorekeepers included behavioral responses to new tax law but did not try to figure how tax-driven changes to the overall economy would impact revenues. GOP backers of dynamic scoring are confident that adding macro effects would show that tax rate cuts would increase growth and lose less revenue—making the politics of rate-cutting tax reform vastly easier. Many critics say they'd love to know how tax changes affect the economy but argue existing models simply are not up to the task. And they fear that requiring dynamic scoring would force JCT and CBO analysts to project the unknowable . At the Brookings program, former CBO director Doug Holtz-Eakin argued that “it doesn’t make any sense to exclude growth effects,” since major tax law changes certainly have some. When pressed on the limitations of the current state of the art, Doug said, “We know just as little about this as anything else.” But most panelists were far more skeptical about dynamic scoring. Major macroeconomic models do a poor job estimating the effects of tax law, and any exercise must confront what Chye-Ching Huang  of the Center on Budget and Policy Priorities called the “uncertainty, flaws, and heroic assumptions” that will inevitably accompany any attempt to estimate revenue effects with those tools. My TPC colleague Len Burman argued that these models often can’t even project whether a specific tax change would gain or lose revenue. Senior JCT economists Pam Moomau and Nick Bull pulled back the curtain, at least a bit, on how their shop uses dynamic analysis and how it’s beginning to deploy these models for revenue estimating. But outside analysts say they need to know more. “I still don’t understand the changes,” said Jane Gravelle, a tax expert at the Congressional Research Service, “I have to speculate.” Last year, JCT used two different models to estimate the revenue impact of Dave Camp’s tax reform—and came up with very different results. The plan would either generate $50 billion in new revenue over 10 years (basically the same as current law) or a hefty $700 billion, depending on which model JCT used and what assumptions it plugged in. Outside analysts were especially skeptical of the larger estimate, particularly because of its assumptions of the effects of tax law changes on labor markets and intellectual property. Panelists on both sides of this debate agreed on four ideas: If it is used at all, dynamic scoring ought to be limited to only a handful of big bills; JCT needs more resources to do this work well; It needs to be more transparent about the assumptions it uses; and it ought to report both dynamic and traditional revenue estimates so lawmakers and outside analysts better understand the differences. Experimenting with dynamic scoring will make it possible for those outside experts to critique the work—which will inevitably improve the quality of the analysis. But JCT may have to Beta test its modeling on big, complex, and highly controversial tax bills in the white heat of congressional debate. And the real world consequence is that Congress may make important decisions based on weak or inaccurate estimates.