Tax reform is not a game. If major US tax reform is enacted anytime soon, it will affect all Americans.

Most will see some change in their personal income taxes, and everyone will be affected indirectly as businesses and the economy adjust to tax reform.

But, as with most games, the scoring of tax reform is critically important.

The score estimates how much money the federal government will receive in tax revenue in the future and who will pay those taxes, which in turn affects the economy.

The score for any tax proposal is the difference between how much tax revenue the federal government is estimated to collect (usually over 10 fiscal years) after enactment of new tax law compared to the revenue the government would have collected if the law had not been changed.

Scores can be positive (government revenues increase) or negative (government revenues decrease). The key word here is estimated.

Revenue scoring is a three-step process.

Step 1: Construct an economic baseline, which is a forecast of key economic aggregates such as Gross Domestic Product (GDP), investment, interest rates, inflation, and population under current law.

Step 2: Estimate baseline revenues based on the economic baseline.

Step 3: Estimate government revenues under proposed law and the economic baseline. The revenue score equals estimated proposed-law revenues minus estimated current-law revenues. The Congressional Budget Office (CBO) undertakes Steps 1 and 2, and the staff of the Joint Committee on Taxation (JCT) generates the official score in Step 3.

Economists generally agree that large-scale tax changes, such as those proposed by the Trump administration and Republican congressional leadership in the Unified Framework for Fixing Our Broken Tax Code, can have measurable effects on economic growth.

Forecasts that include the feedback effects of economic growth on tax revenue are called “dynamic scores.” Congress has directed that dynamic scores be used for “major legislation” – proposals that would have gross budgetary impact, before taking into account macroeconomic effects, of 0.25 percent of GDP in any year over the next ten.

Smaller proposals are estimated under the constraint that GDP does not change between current law and the proposal. These estimates are typically called “static” estimates, although it would be more accurate to call them “macro-static,” as “static” scores take into account dynamic shifts in economic activity across sectors or markets and/or changes in the timing of economic activity.

The controversy is not about whether a revenue score should be dynamic or static; the official scores of tax reform proposals use dynamic scoring. The controversy arises over how “dynamic.” In other words the size of the macroeconomic impact.

The non-partisan staffs of the JCT and CBO use extensive econometric analysis to determine the feedback effects of economic growth on tax revenues. That consistent, careful and unbiased modeling of dynamic feedback demonstrates that the feedback is very real, but generally it is not as big as some would like.

A brief article cannot prove that dynamic feedback is small (or large). But the following does show how much feedback is required for a major tax cut to pay for itself, which is what the President and congressional leaders claim the Framework would do. To hear Treasury Secretary Mnuchin describe it, the plan “will not only pay for itself, it will pay down the debt.” The math is simple, but the actual numbers are hard.

Reducing the corporate tax rate from 35 to 20 percent (as proposed in the Framework) means that government revenue from existing taxable profits would decline by roughly 43 percent (1 – 20/35), not including any macroeconomic feedback. (Eliminating some of the existing loopholes and preferences in the corporate code would change that percentage.) That tax cut would prompt growth from three direct sources.

Corporations currently subject to US tax could invest more and generate more profits. New businesses could start up, and pay new taxes. And corporations could shift investment and production from outside the US tax system to inside. There could also be indirect increases in tax revenue, as new income to workers and owners also generates some new tax revenue. But, for the reduction in tax rate to pay for itself, taxable profits (and the related personal taxable income) would have to grow through economic feedback by 75 percent (35/20 – 1), which is a very high hurdle. And that is just to pay for itself – not to reduce the deficit or debt over time.

Virtually everyone agrees that tax cuts don’t pay for themselves.

That means to avoid increased deficits the offsetting revenue or spending cuts have to come from somewhere. We cannot afford to simply trust what snake oil revenue projections politicians are promising.

That’s why lawmakers have to rely on scores from the non-partisan green eyeshade folks at CBO and JCT.

The total national debt exceeds 100 percent of GDP. Interest to service the public debt is estimated to be more than $800 billion in ten years’ time. That’s more than the current defense budget and more than triple the $240 billion interest last year.

Tax reform is important, but only if done right.