Treasury Secretary Steven Mnuchin, left, and National Economic Council Director Gary Cohn met with members of the Senate Finance Committee earlier this month. (Alex Wong/Getty Images)

Dear Colleagues:

You recently wrote an open letter to Treasury Secretary Steven  Mnuchin quantifying the economic impact of tax reform. We are interested in and surprised by your analysis.  We share your commitment to the idea that well-designed tax reform can make the economy stronger and that careful economic analysis is essential. And we know that you all share our belief that such careful analysis is well served by discussion and debate of these issues that is at least as frank and vigorous as what we are all accustomed to in the average economics seminar. To that end, we think it would be useful to lay out some of the questions we have about your analysis:

First, many members of Congress are citing growth estimates consistent with your letter to claim that the tax cuts would pay for themselves and that the legislation being considered by Congress would not add to the deficit or debt over the next decade. Your letter, however, does not say that tax cuts would pay for themselves. Would it be fair to say that you agree with Martin Feldstein (who did not sign the letter) that these tax cuts will not pay for themselves and, in fact, would add about over $1 trillion to the debt over the next decade?

Second, can you explain how the studies you cite justify the conclusions you reach? You cite three studies to justify your conclusion that the annual growth rate would rise by 0.3 to 0.4 percentage points over the next decade. But two of those studies actually appear to have estimated substantially lower growth rates — potentially as low as a 0.01 percentage-point increase in the annual growth rate.

You cite a Treasury study of the George W. Bush tax reform commission’s Growth and Investment Tax Plan that you asserted found a 4.8 percent increase in long-run GDP. But the study you refer to provides estimates from three different models ranging from 1.4 percent to 4.8 percent increases in national income and does not express any views on which model is preferred. What was your reason for citing only the upper end of the range of Treasury’s estimates, from one model and ignoring Treasury’s other estimates? Also, why did you not mention that the middle of the range of Treasury’s 10-year estimates was 1 percent, a figure that would corroborate the views of critics of the tax bills since there would be only a 0.1 percentage-point increase in the growth rate? (Moreover, the Bush Commission’s tax plan differed in important ways from the plan now before Congress: It included permanent expensing, applied expensing to structures, raised taxes on investments that are already in place and was fully paid for. Do these differences affect the validity of your use of this plan as a model for the current Congressional bills?)

You also cite an OECD study that you say justifies the conclusion that long-run GDP would go up by 2 percent. But since you are explicitly talking about 10-year growth rates in your letter, would it not be better to use estimates from this same study that show that the effect in the 10th year is less than one-third of the long-run effect, translating into an annual growth rate of less than 0.1 percentage point? Moreover, how did you come up with the 2 percent long-run GDP number since the OECD study says that a 1 percent of GDP reduction in corporate taxes adds 1.25 percent to long-run GDP? Applying that estimate to the corporate rate reductions in the Congressional bills would yield only a 0.8 percent increase in long-run GDP, translating into a growth rate increase of 0.02 percentage point per year (and if you factored in the base broadeners the magnitudes would be half as large). Either way, the OECD study you cite also corroborates the critics of the tax bill.

Third, did you give thought to the impact of the corporate rate cut assuming that expensing as proposed in the bill was enacted? We suspect that much of the projected growth benefit from corporate tax reform comes from enacting expensing of equipment, which reduces the entity-level effective tax rate to zero on equity-financed investment and makes it negative if financed in part with debt.  In the presence of debt finance, textbook analysis would suggest that a cut in the corporate tax rate would raise the cost of capital because interest deductions would no longer be as valuable and thus discourage investment. Have you considered this important possibility, since most of the budget cost of the reform comes from the corporate rate reduction?  Moreover, even with the lower statutory tax rates in the bills if expensing ended after five years, as it does in the bill, effective marginal tax rates on equipment investment would actually be higher than they are today with bonus depreciation in effect. It would be ironic if lower corporate rates and an expiration of bonus depreciation actually discouraged investment at the margin relative to continuing current policy, but we believe this is likely.

Fourth, the pass-through provisions in the House and Senate bills would appear to create new sources of complexity in the tax code and violate the basic principle of tax policy that similar sources of income should be taxed at similar rates. A number of you have expressed concerns about the pass-through provisions in the past. Did you model the impact that these provisions would have on macroeconomic impact of the tax cuts?

Fifth, President Trump has expressed concern about the magnitude of the trade deficit and his chairman of the Council of Economic Advisers said that “a corporate tax cut to 20 percent would dramatically reduce the trade deficit.” In your analysis you reject concerns about the macroeconomic impact of budget deficits because, you argue, the United States will be able to attract capital from the global capital market. As a matter of logic, won’t increased capital inflows require an increase in the trade deficit, totaling hundreds of billions of dollars annually if they are to finance your projected 15 percent increase in the capital stock?

Sixth, apart from the question of global financing of the budget deficit, do you worry about the impact of enlarged deficits given projections of rising spending on entitlements and national security?  Do you think that it is realistic that Congress will actually sunset provisions like the expanded child credit or corporate expensing, or do you think that the true cost of this bill is likely to significantly exceed $1.5 trillion? One of the authors of your letter wrote that the Bowles-Simpson fiscal commission “was very good,” and several other signatories wrote that it should be the “starting point” for fiscal negotiations. Do you think that cutting revenue to less than 18 percent of GDP, as would happen under the congressional legislation, is consistent with the 21 percent of GDP in revenue that the Bowles-Simpson commission recommended for long-run fiscal sustainability? Moreover, we understand that you support entitlement reform, but do you believe that it is politically realistic for the Republicans to actually achieve your goal when it would entail asking seniors to sacrifice by cutting Social Security and Medicare shortly after Republicans argued that we could afford to add over a $1 trillion to the deficit for tax cuts that largely benefit corporations and high-income households?

Finally, one of you signed a previous letter stating that because outside estimates are often “not objective and not as well-informed as CBO’s analysts” that “relying on CBO’s estimates in the legislative process has served the Congress —  and the American people  — very well during the past four decades. As the House and Senate consider potential policy changes this year and in the years ahead, we urge you to maintain and respect the Congress’s decades-long reliance on CBO’s estimates in developing and scoring bills.” Your recent letter ignores many specific features of the legislation. To give just three of many examples that you did not incorporate into your analysis: The legislation would increase the after-tax cost of R&D, would increase asymmetric penalties on corporate risk-taking and would raise effective marginal tax rates for many individuals through the repeal of the state and local tax deduction. Do you think your analysis of a highly simplified hypothetical plan that is different from the actual legislation before Congress should serve as a substitute for the Joint Committee on Taxation (JCT) and the CBO producing analysis? Do you think it is responsible for Congress to vote on technically complex legislation in the absence of hearings or complete analysis?

Thank you for your attention to these questions.  We may disagree on the merits of particular proposals, but as professional economists we can all agree on the importance of critical discussion and debate to advance the improved understanding that is the basis of better public policies.


Jason Furman
Professor of Practice, Harvard Kennedy School

Lawrence Summers
Charles W. Eliot University Professor and President Emeritus at Harvard University