Skip to main content
On the Record

Comments on IMF’s Presentation on Fiscal Policy and Growth

Douglas W. Elmendorf

Thank you, Adam, I’m delighted to be here, and I’m pleased to have a chance to comment on this very interesting paper by a terrific team from the Fiscal Affairs Department at the IMF.

The paper presents an important catalog of fiscal policies that could strengthen economic growth, and I commend it to policymakers in this country and around the world. The paper covers many specific policies to boost the quantity and quality of labor, the quantity of physical capital, and multifactor productivity—which is essentially the efficiency with which labor and capital are combined. The paper describes the policies with great care, it offers many examples of the successful implementation of those policies in different countries, it provides an empirical analysis showing the pick-ups in economic growth that have followed the implementation of those polices, and it gives the results of simulations of policies from a so-called endogenous growth model.

I agree entirely with the main message of the paper—that changes in fiscal policies could significantly boost output and incomes in many countries. I hope that policymakers and their staffs in every country will read the paper and think hard about ways in which their countries’ fiscal policies could be improved.

However, I would like to use my time to offer two cautions about interpreting the findings of the paper: First, I am skeptical of the paper’s assertion that better fiscal policies could raise the annual growth rate in advanced economies by three-quarters of a percentage point in the medium- to long-term. Second, I think the tradeoff between the goal of raising overall economic growth and the goal of raising living standards for lower- and middle-income people is starker than the paper admits. Let me address those points in turn.

The first issue is how much promise fiscal policy holds for strengthening overall economic growth. High aspirations have their advantages, but I worry that the numbers in the paper are promising too much. The paper suggests that, in advanced economies, budget-neutral tax reform could add about a half-point to economic growth for many years, and a shift of government spending toward infrastructure investment could add another quarter-point for many years. And the paper adds that even bigger gains are achievable in emerging market economies and low-income countries.

Three-quarters of a percentage point is a very big number in the business of medium- and long-term economic growth. For example, the Congressional Budget Office projects that the potential labor force in the United States will increase by half a percent per year over the coming decade. If that rate were boosted by a half-point, it would return the rate halfway to what it was in the 1980s when women’s labor force participation was rising rapidly and baby boomers were not yet retiring. That’s a lot to ask from tax reform.

I believe that tax policies are bad enough now in some countries that reform could increase sustainable growth rates by a half-point or more, but I would be surprised if that were true in most countries. Again using the United States as an example, suppose that the effective marginal tax rate on labor earnings was cut by 5 percentage points, which would be a pretty big change from revenue-neutral tax reform, partly because achieving that drop in the effective marginal rate would require a larger drop in statutory rates for various reasons. Based on an extensive literature on the response of labor supply to changes in tax rates, CBO has said that such a reduction would raise labor supply by 2 percent or less. If the capital stock increased by the same amount, then GDP would be 2 percent higher. That corresponds to less than a quarter-point per year on the growth rate, not a half-point, and for 10 years, not indefinitely.

Similarly, based on the analysis of infrastructure investment that CBO has done for the United States, raising GDP growth by a quarter-point per year for 10 years would require literally trillions of dollars of additional infrastructure investment in that period.

Now, CBO may be wrong. I think the endogenous growth model used in this paper would predict a larger and more sustained increase in the growth of GDP from both tax reform and infrastructure investment than the more traditional model used by CBO, and the synthetic control method used in the paper implies larger effects. But I don’t think economists know enough about the process of economic growth to have great confidence in specific sorts of models or empirical techniques. Without that confidence, I think economists should talk about policy changes raising the level of output or the growth rate of output for a limited period rather than about raising the growth rate for an indefinite period. And I think economists should be careful not to overstate the potency of positive policy changes, because subsequent disappointments can undermine people’s confidence in economic analysis.

My second caution about the paper is that I think there is a more difficult tradeoff than the paper states between the goal of boosting overall economic growth and the goal of boosting the well-being of lower- and middle-income people.

To be sure, not every policy involves such a tradeoff. For example, there is growing evidence from the United States that certain policy interventions for low-income children generate high returns in their income as adults. Such interventions include preschool education, regular access to health care, and some types of housing assistance. In addition, support for lower-income students to attend college leads to higher income for them later in life. Increasing spending for those purposes—leaving aside the question of how to pay for it—would probably improve both growth and equity.

However, paying for that rise in spending would require a combination of greater federal borrowing, higher taxes, or lower government spending for other purposes. Those changes might well hamper overall growth or equity. That problem is especially acute in the many countries that have experienced sharp run-ups in public debt during the past several years and are experiencing significant population aging.

Moreover, many other important tax and spending policies present more direct tradeoffs between overall income and the incomes of people of modest means. One current example from the United States is the Affordable Care Act. According to CBO, repeal of that act would increase the level of GDP by three-quarters of a percent, primarily by taking away the implicit tax that comes from the phase-out of health insurance benefits as one’s income rises. However, repealing the act would also mean that 24 million more people would not have health insurance. That is a significant tradeoff.

Another example is tax reform again. The staff of the U.S. Joint Committee on Taxation estimated last year that Congressman Camp’s comprehensive plan would probably raise GDP by something less than 1 percent. The effect was not larger in part because Camp set objectives of revenue and distributional neutrality relative to the current tax code, and those objectives and others meant that the proposal would actually increase the cost of capital for domestic firms and thereby decrease the stock of business capital. If, instead, Camp had abandoned that distributional goal, he could have designed a tax system with a larger positive effect on economic growth. That is also a significant tradeoff, and it is not unique to his plan or to the U.S. tax code.

It is certainly true that economists should work hard to develop policies to “mitigate equity-efficiency tradeoffs,” as the report says. The report discusses a number of important areas in which policies can be improved in this way, including corporate and individual income tax policy, benefits for unemployed and disabled people, and more. Further work on those issues is a critical mission for policy analysts in the years ahead.


Still, I think economists should say directly to policymakers that some, but not all, policies that would increase the size of the overall economic pie would diminish the shares of the pie going to people who are already receiving smaller-than-average shares—and similarly, some, but not all, policies that would increase the shares of the economic pie going to lower-income people would diminish the size of the overall pie.

Those points are not new, of course. We are celebrating this year the 40th anniversary of Arthur Okun’s famous book about the tradeoff between equality and efficiency. But being upfront about that tradeoff is especially important today because we have seen in the decades since Okun wrote that increasing total income in a country may not increase the incomes of most people in that country. Therefore, if policymakers want to help lower- and middle-income people, they need to consider explicitly the effects of policies not only on total income but also on the incomes of lower- and middle-income people.

I will conclude here simply by thanking again Vitor Gaspar and the team of authors of this report for a very important contribution on a critical set of policy issues. Thank you.


Get daily updates from Brookings