Press reports indicate that President Donald Trump’s budget, scheduled for release tomorrow, will assert that administration policies can deliver a balanced budget in 10 years by combining sharp cuts to anti-poverty and safety net programs with growth from unspecified or minimally detailed tax and regulatory reforms. According to these reports, the forthcoming budget assumes that the rate of economic growth will reach 3 percent by 2021. In contrast, the Congressional Budget Office projects a growth rate of 1.9 percent for the same year.

Assuming large growth effects from policies that have yet to be specified in detail certainly qualifies as fantasy budgeting or, in Washington terms, a magic asterisk. But what’s even more striking about the anticipated budget plan is the expected assertion that revenue-neutral tax reform will contribute to deficit reduction. Administration officials and congressional Republicans have been explicit that they plan to credit the revenue feedback from any growth delivered by tax reform against the cost of tax reform. Yet counting the revenue feedback from growth in assessing whether a tax reform proposal increases or decreases revenues means that there is no additional revenue to reduce the deficit below the level that would be realized under current law.

In short, the Trump administration seems prepared to double-count the gains from its magic asterisk.

A back-of-the-envelope calculation suggests that the assumed growth rates could add $2 trillion to revenues relative to the CBO’s current-law baseline. Thus, a budget that purports to achieve balance predicated on this growth could be short by well more than $1 trillion—even if the growth projections were realized—by failing to recognize that the revenues from tax-reform-induced growth are going to be used to offset the cost of that reform. (The exact overstatement of revenues would depend on how much growth is attributed to tax reform and how much growth is attributed to other policies in justifying the economic assumptions.)

This estimate of the overstatement of revenues would be conservative even if large growth effects were realized, as it ignores both the implausibility of the administration’s growth forecasts and the large revenue losses that would result from the tax reform plans that the Trump campaign and the administration have put forth previously. The Tax Policy Center, for example, estimated that then-candidate Trump’s plan would cost $6 trillion while delivering less than $200 billion in revenue feedback from growth in the first decade of its implementation—before ultimately harming growth in the long run by running up the debt and thus reducing investment. These estimates suggest that the administration’s budget documents could be understating deficits by well more than $6 trillion relative to the actual impact of its policies.

Even by the standards of the federal budget, which operates at a scale that is sometimes difficult to comprehend, these numbers are large. If only $1 trillion is attributable to double-counting the gains from tax reform, that’s still more than the savings that the American Heath Care Act realizes by taking health insurance away from 14 million Americans through Medicaid cuts. Moreover, the CBO’s most recent deficit projections under current law total $9 trillion for the next 10 years. Assuming the administration will achieve balance by the 10th year but not before, the Tax Policy Center’s estimates of the Trump tax plan suggest that the president and his economic policy team could be claiming to reach balance while actually making the 10-year deficit outlook worse—even with harsh cuts to anti-poverty programs and policies that sharply reduce the number of Americans with health insurance.

How will it be apparent that the administration is not merely assuming implausible growth from its policies but actually double-counting those same growth projections? Traditionally, the president’s budget is presented on a set of post-policy economic assumptions. That is, the economic assumptions underlying the budget assume the enactment of the president’s policies. Judged against this set of economic assumptions, a revenue-neutral tax plan with implausible growth effects, such as those previously promised by the administration, should appear as a large tax cut, measured in the low trillions of dollars. Since the administration is pointing to tax reform as a justification for the rosy economic assumptions, this is the most likely approach—and the most likely criteria by which the budget should be judged.

While the traditional approach to budgeting includes the impact of proposed policies in setting the economic assumptions, there is an alternative approach that would be more consistent with the administration’s rhetoric on tax reform. In fact, the Obama administration used this approach when it counted deficit reduction resulting from economic growth generated by certain elements of immigration reform as a policy impact in its budgets. Under this alternative approach, the economic effects of a policy change are ignored when setting the economic assumptions. But by ignoring the growth impacts when setting the economic assumptions, those growth impacts can be included in the budget estimate for the proposal without double-counting the gains.

Thus, while the CBO concluded in 2013 that immigration reform would add 3.3 percent to gross domestic product 10 years after enactment, the economic assumptions underlying the last budget submitted by the Obama administration incorporated growth of only 0.7 percent attributable to immigration reform in its economic assumptions. The difference between 0.7 percent and 3.3 percent reflected the growth that had been included in estimating the budgetary impact of the proposed immigration reform. Critically, if the Trump administration takes this second course in estimating the impact of tax reform on the budget, then budget documents would appropriately show no impact of tax reform on revenues—but the administration would then need to justify its economic assumptions without reference to tax reform.

The analysis above takes administration officials’ previous public statements as informative about the direction of policy. Another option would be to recognize the budget as a statement of administration policy that supersedes those previous statements. Under this view, including the growth impacts of tax reform in setting the economic assumptions for the budget while showing no revenue impact of tax reform would amount to a statement that the administration now believes tax reform should be revenue-neutral based on conventional scoring—excluding impacts on growth—and any gains from growth should thus be used to reduce the deficit.

Recognizing the long-term fiscal challenges the country faces, this approach would be a wise policy choice on the part of the new administration. And in combination with Treasury Secretary Steven Mnuchin’s previous statements ruling out a tax cut for the upper class—the so-called Mnuchin rule—this addendum would provide a solid foundation for real tax reform. Yet it is unlikely that this is the course the administration intends to pursue.

So when the administration presents the budget tomorrow, watch not only for rosy economic assumptions to improve the deficit outlook but also for double-counting of the benefits of those economic assumptions. If the administration does in fact double-count the benefits, then recognize and understand the implied policy content of that choice. And if the administration tries to justify its harsh cuts to anti-poverty and safety net programs on the basis of clear-eyed fiscal accounting, then remember that these same officials are anything but clear-eyed when it comes to accounting for their plans for tax reform.