By Josh Barro, The New York Times, September 14, 2015
* * *
Jeb Bush says he wants to make the American economy grow at 4 percent a year, or nearly 2 percentage points faster than the Congressional Budget Office expects in the long run. The goal has been greeted with skepticism, but last week, four prominent conservative economists predicted his tax-cut plan would get him a good portion of the way there.
They say it would add half a percentage point to G.D.P. growth every year for a decade, eventually adding 5 percent to the size of the economy.
For decades, faster economic growth has been a key justification offered by Republican candidates promoting tax plans that include large tax cuts for the highest earners and owners of capital. These sorts of tax cuts, their economic advisers say, will produce more jobs, higher wages and better standards of living.
The problem is, nobody really knows how tax cuts actually affect jobs and growth — not even economists.
The economists making the prediction — John Cogan, Martin Feldstein, Glenn Hubbard and Kevin Warsh — did not run Mr. Bush’s tax plan through an economic model before arriving at their 5 percent estimate. Rather, Mr. Hubbard, the dean of Columbia’s business school and a former top economic adviser to George W. Bush and Mitt Romney, said they looked at previous studies in which economists evaluated other tax proposals, and placed Jeb Bush’s tax plan among them in what they thought was approximately the right place.
A model built by Alan Auerbach and Laurence Kotlikoff found that abolishing income taxes in favor of a consumption tax would add 9 percent to G.D.P. in the long run. Another model by John Diamond and George Zodrow found that a more modest tax overhaul considered by House Republicans last year, focused on cutting rates and eliminating deductions, would lift G.D.P. by 3 percent. Other proposals fell somewhere in the middle, and so, the Hubbard group thought, would Mr. Bush’s.
“This is four smart people who made an educated guess based on looking at other models,” said Joel Slemrod, a public finance economist at the University of Michigan. “If I had to pick a number, I would have picked a lower number.”
And that’s the thing: While most economists agree that tax policy can have significant effects on economic growth, they have a remarkable ability to look at the same body of evidence and disagree wildly on the size of the effects. Asking a handful of economists about how your tax plan will affect the economy is better than throwing darts at a dartboard, but it’s not necessarily that much better.
Mr. Auerbach and Mr. Kotlikoff have collaborated since the 1980s on influential models of how taxes affect the economy. Yet they disagree on how to evaluate Mr. Bush’s tax plan. Mr. Auerbach said he read the Hubbard group letter and immediately thought the growth estimates were too high, while Mr. Kotlikoff thought they were too low.
“0.5 percent a year definitely stuck out when I read it,” said Mr. Auerbach, a professor at the University of California at Berkeley. “I don’t really think it’s reasonable.” He notes the Bush plan retains significant taxes on investment income, while his model found the largest growth effects from abolishing such taxes. He also thought it would take much longer than 10 years for the positive economic effects of a major tax overhaul to be fully felt.
Mr. Kotlikoff, a professor at Boston University, said nearly the opposite: He thought the Bush plan, especially its provision allowing companies to fully and immediately deduct capital expenses from their taxes, would have large and swift economic effects. He said he thought the plan would grow the economy by more than 5 percent, and he thought most of those effects would be felt within the first decade, pointing to Ireland as an example of a country that experienced rapid economic growth after cutting corporate income taxes.
The diversity of views in the economics profession makes it easy to pick and justify the forecasts you like. With last year’s Republican House tax plan, there is that outside analysis Mr. Hubbard noted, finding it would expand the economy by 3.1 percent in the long run (though only by 2.2 percent after a decade). But the congressional Joint Committee on Taxation ran not one but eight models on that plan, finding it would increase the economy by 0.1 percent to 1.6 percent over a decade.
This is how the Joint Committee on Taxation and the Congressional Budget Office deal with the economics profession’s uncertainty about how taxes affect the economy: They provide a range of estimates, often such a broad range that it feels as if the report isn’t telling you anything at all.
Douglas Elmendorf, who until recently served as the director of the Congressional Budget Office, told me the main usefulness of these reports is the estimates they rule out: By saying a tax plan might cause a boost to G.D.P. from 0.1 to 1.6 percent, they are saying to ignore an outside report promising it will cause a rise of 4 percent. (Mr. Elmendorf joined several other economists in saying he thought the prediction of 5 percent in the Hubbard group letter was overly rosy.)
Mr. Slemrod also urged economists to talk in terms of ranges rather than point estimates when discussing how taxes affect the economy, to reflect the fact that these figures are simply educated guesses. But he understands why they don’t. “Washington wants a number,” he said. “Washington doesn’t like confidence intervals.”
In the case of Mr. Bush’s tax plan, using a point estimate is politically desirable because it builds in the economic feedback effects as a key justification for the policy. If the tax cuts really did cause the economy to grow by 5 percent, they would generate a lot of revenue to offset the cost of cutting taxes, and would do a lot to improve living standards even for people who do not benefit much directly from the tax cut.
If you instead consider a range of possible economic effects that includes very small ones, you have to be prepared for the possibility that the tax cuts won’t provide those benefits. That means the government will have to borrow much more money and cut more spending than anticipated. Or raise other taxes.
Voters should remember that we have been burned before by economists’ guesses on taxes, including guesses by the very economists who wrote the letter backing Mr. Bush’s plan. For example, in August 2001, Mr. Hubbard, who was then serving as the chairman of the Council of Economic Advisers, wrote for The Wall Street Journal about the first round of George W. Bush tax cuts under the headline “Tax Cuts Won’t Hurt the Surplus.” He said:
The recently enacted tax cut fits naturally into a growth-oriented agenda. Lower taxes reduce fiscal drag on the economy in the short term. Lower marginal tax rates provide longer-term incentives for work, savings, risk-taking and innovation. In this way, the current fiscal policy helps to provide for the future of Social Security.
That guess turned out to be both wrong and expensive.