Dear Lawrence Summers and Jason Furman,
We read with interest your Dear Colleague letter in The Washington Post in response to ours in the Wall Street Journal.
You ask if we have considered several points you make. The answer is yes — we have considered each before coming to our conclusion. Responses to each of your points follow:
First point you raised: Our letter addresses the impact of corporate tax reform on GDP; we did not offer claims about the speed of adjustment to a long-run result (though official revenue estimators will obviously need to do so for short-run analysis). We did not approach Marty Feldstein as a signatory, as he generally is not a “signer” of letters, and he can certainly speak for himself. In fact, he did in Project Syndicate on Nov. 27: “I dislike budget deficits, and I have long warned about their dangerous effects. Nonetheless, I believe the economic benefits resulting from corporate tax changes will outweigh the adverse effects of the increased debt …”
Second point, Part (a) you raised: The studies we cite all find that reductions in corporate taxation have important positive effects on economic growth. Ultimately, we are confident that the estimates in our piece are closer to the mark and are at the same time broadly consistent with other estimates from empirical studies of effects of corporate tax changes on growth.
Second point, Part (b) you raised: We refer to the estimate we believed most accurately reflects likely saving responses and thus capital accumulation.
Second point, Part (c) you raised: We state explicitly in the letter that the figure calculated on the basis of the OECD study is a long-run estimate (the OECD study estimates effects on GDP per capita, not GDP per se). We wanted to be consistent in the choice of cost score. Choosing one, higher-cost estimate to show that the deficit effects are large and another, lower-cost estimate to argue that the growth effects would be small would be inappropriate. Second, because the underlying study is an OECD study, we used a second recent OECD study that estimated the effects of changes in effective marginal tax rates on corporate tax revenue as a share of GDP. One can then apply the estimated effect on GDP per capita of a revenue change of that size, having estimated the percentage-point change in the effective marginal tax rate that would result from the proposed changes. This approach seeks consistency in working with cross-country estimates, rather than picking and choosing from incomparable models and regressions. By this method, the proposed changes would raise long-run GDP per capita by approximately 1.8 percent.
Third point you raised: We incorporated the proposed expensing provisions in our user-cost modeling, and the range of estimates we present reflects these provisions. Moreover, as research on investment has demonstrated, expiring provisions can accelerate investment plans, bringing economic activity forward, even though the long-run, steady-state effect is lower than if those provisions were permanent. We incorporate this effect in the user-cost modeling and growth accounting exercise.
Fourth and fifth points you raised: The purpose of our letter was principally to address effects of corporate tax changes on economic activity, as, in our judgment, the bulk of growth effects come from contemplated corporate tax changes and many individual tax changes were (and are) in flux.
Sixth point you raised: The nation’s fiscal health is a very important topic for consideration by our leaders. Fiscal policy ultimately centers on two choices: (1) the size and scope of government and (2) the ways and means with which government spending is financed. Put that way, the question is less one of taking projections as given and more one of making long-run spending choices in a manner that supports sustainable economic growth and permits investment in key public goods.
Robert J. Barro, Michael J. Boskin, John Cogan, Douglas Holtz-Eakin, Glenn Hubbard, Lawrence B. Lindsey, Harvey S. Rosen, George P. Shultz, John B. Taylor