Tax Cuts and Growth: The Need for Precision

by Kody Carmody

Cover Photo Credits

As the Trump administration pushes for new tax cuts, our nation re-enters the perennial debate over how to spur economic growth. As part of this debate, there seems to be a widely-held belief among Americans that economists doggedly pursue higher GDP growth without regard to actual human well-being and that they want to do so by blindly cutting taxes and regulations.

Leaving aside the first point, which deserves a blog post of its own, holders of the second belief might be surprised to see the below 2012 IGM Economic Experts Panel responses:

Screen Shot 2017-11-01 at 12.55.48 PM.png

The IGM panel, which regularly asks ~40 of the most distinguished economists around the U.S. for their opinions on certain economic issues, includes participants from a diversity of backgrounds, ideologies, partisan affiliations, and ages. Even so, it seems strange that there could be so much uncertainty around whether a cut in federal income tax rates would cause greater economic growth. Every Econ 101 student walks away knowing that, in most cases, higher taxes distort markets and ­decrease economic output. Where does the disconnect come from?

To understand the causes of economic growth more broadly, we can look at the production function used in a standard neoclassical growth model:

Screen Shot 2017-11-01 at 12.56.29 PM.png

Essentially, this says that an economy’s level of output (y) is a function of its capital stock (k), skilled labor (h), and technology (A) with; capital’s share of income (α) is treated as a constant. Note that these are all in per capita terms, since per capita growth is what we care about. If the economy grows by three percent but the population grows by five percent, there will be less stuff being produced per person and the average living standard would fall.

Without getting into the weeds of the model, we can see a few potential sources of long-run economic growth: increasing capital per worker, increasing the skill or size of the workforce, and creating technologies that make our workforce more productive. Proponents of tax cuts will point to some obvious theoretical links between these sources of growth and taxes. For example, cutting income tax rates raises the returns to working—encouraging more work increases the intensity of labor (h). Economists call this a substitution effect—as the reward for working increases, holding all else constant, people will substitute towards work.

But it turns out that the story, at least for income taxes and work effort, isn’t is not as clear as that. Imagine you suddenly received a large raise in your salary—no matter how much you work, you will receive an extra $10,000 per year—you might rationally choose to work less. Economists call this an income effect—an increase in income decreases an individual’s need to work as much. An income tax cut, then, will lead to both substitution and income effects, and pure theory can’t tell us the effects of the tax cut on work effort or, in turn, economic growth. The net effect should also depend on the actual tax rate—a cut from 90% to 80% will have a much larger substitution effect than a cut from 20% to 10%, even though both would have the same income effect.

Empirics confirm that the real behavioral responses to income tax cuts, if they exist, are small. A 2012 review in the Journal of Economic Literature looks at the behavioral responses of high-income individuals and finds “no compelling evidence to date of real economic responses to tax rates.” This is confirmed more generally by a 2011 paper in the American Economic Review, which estimates that the discouraging effect of taxes on work is very low.

There is an important exception, however. A 2015 NBER Working Paper finds that tax cuts for low- and moderate-income individuals do increase work effort: “in fact, the positive between tax cuts and employment growth is largely driven by tax cuts for lower-income groups.” This effect heterogeneity is probably behind some of the uncertainty in the IGM survey; several of the responding economists had comments like Bengt Holmstrom: “results [are] highly dependent on which rates are cut.”

Income taxes are only one possible avenue for tax cuts, and work effort is only one of the possible avenues by which income taxes could affect growth. The above is just to show that the causal narratives are often more complicated than “lower taxes lead to growth.” In general, empirical research on the relationship between taxes and growth has turned up mixed evidence. One difficulty is that changes in taxes are often linked to other developments in the economy, and it is hard to tease out the effects of the tax change itself. A 2010 American Economic Review paper is probably one of the best efforts to disentangle these, and finds that tax cuts have significant and large effects on growth. In response, an NBER Working Paper uses a slightly different estimation strategy for the same data and finds only small effects in general as well as no effect for tax changes after 1980.

 There is also some prima facie reason to believe that tax cuts will only have small effects on growth. As Dietrich Vollrath, who wrote the book on growth economics (cited above), points out: tax cuts act on potential, not actual, GDP, and it takes time for the economy to converge its potential GDP. The rate of convergence is commonly estimated as around 2%. Vollrath considers a case where the U.S. is missing out on a full $1 trillion in economic activity solely due to taxes; doing some quick math, he shows that completely removing this inefficiency, raising potential GDP from $17 trillion to $18 trillion, would only lead to 2.62% GDP growth. Raising potential GDP by $2 trillion, an extremely generous estimate, would only lead to 2.73% growth.

In sum, there are many compelling stories we can tell about why tax cuts might lead to growth, and specific tax cuts do seem to spur very specific economic activity, such as income taxes on the poor and work effort. But there is only mixed evidence for a relationship between taxes and growth more broadly, and even if we assume that tax cuts have extremely large effects on economic output, the convergence of actual to potential GDP means that this large effect would likely only translate to a small increase in growth.

All of this is to say that economics is hard; as a corollary, policy is hard, and we should avoid over broad claims like “tax cuts create growth.” That being said, there are some remarkably consistent results in the tax literature. One is that, given current tax levels in the U.S., cutting taxes would not spur economic growth enough to raise total revenue, contrary to the view posited by Arthur Laffer in 1974; see question B from the IGM survey:

Screen Shot 2017-11-01 at 12.56.00 PM.png

Another is that deficit-financed tax cuts decrease growth in the long run, as recently experienced by the state of Kansas. The takeaway is that when evaluating tax proposals, the specifics matter. This is best illustrated by the IGM survey responses to question A, whether a cut in rates would lead to higher GDP: Austan Goolsbee, former chair of the President’s Council of Economic Advisors, writes “probably, but the evidence is not as strong as you would think”; Kenneth Judd from Stanford points out that “everything depends on which taxes are cut”; but Nancy Stokey of Chicago puts it best with her comment, “Vague question.”