The Case for Low Capital-Gains Taxes Grows Weaker
(Bloomberg Opinion) -- By now, most informed observers know that economics research has gotten much more empirical since the early 1980s.
What might be less obvious is why this matters so much. To understand, it helps to look at the history of the debate over capital taxation.
In economics, capital taxation means the levy on income that people get from owning assets, like stock in a company. There are several types of capital taxes, but the main ones are corporate-income taxes, capital-gains taxes and taxes on dividends. These are different from labor taxes, like payroll taxes or ordinary income taxes, because they affect the economy in different ways. Theoretically, labor taxes discourage people from working, while capital taxes discourage them from investing in companies.
Standard theory also says that when people are discouraged from investing in companies, those companies will invest less in real, productive assets like offices, machine tools, delivery trucks and computers. Ultimately, lower business investment should mean a less productive economy, lower wages and a reduced standard of living in the long term. (Of course, that has to be balanced against the value of whatever the tax gets spent on; if buying more food for the poor means rich people will have smaller yachts, so be it.) The real question is how much this effect matters — if it’s minor, then taxing capital might be a good way to pay for government spending.
In the mid-1980s, when theory was king, two economists — Kenneth Judd in 1985 and Christophe Chamley in 1986 — independently came up with theoretical arguments for why capital shouldn’t be taxed. They were not the only economists to reach a result like this, but their papers were especially influential — “Chamley-Judd” has become a shorthand that economists use in discussions of capital taxation.
The Chamley-Judd result has colored the worldviews of thousands upon thousands of academically trained economists, including myself. Think tanks have used it to push for corporate-tax reductions. Economic policy advisers, at least in the U.S., have generally tended to think that individual income taxes, or consumption taxes, are more efficient than capital taxes, and Chamley-Judd is part of the reason.
But this classic result is almost certainly more influential than it deserves to be. One reason is that Chamley and Judd’s models involve lots of unrealistic assumptions. The models are of a type called general equilibrium, which attempts to model the entire economy at once. It was especially popular in the 1980s:
Modeling the economy is very, very hard unless you make a lot of simplifying, and often-unrealistic, suppositions. For example, Chamley and Judd assume that everyone is immortal — or at least, plans for the future as if they’ll live forever. This assumption helps the case against capital taxes, because the impact of these taxes is felt especially keenly in the distant future. It also makes the model easy to solve mathematically. But as a number of economists have shown, when you drop this assumption, the result that capital taxes are always bad falls apart. For example, in a 2012 paper, economists Thomas Piketty and Emmanuel Saez replace the assumption of immortality with the more realistic idea that people differ in how much they care about leaving money to their kids, and find that the Chamley-Judd result no longer holds.
A second, even more insidious problem of relying on theory is that it’s often difficult to connect to the real world. For example, in 2014, macroeconomists Ludwig Straub and Ivan Werning showed that even if you believe Chamley and Judd’s models are good ones, they don’t necessarily show what people think they do. If a certain parameter in the model, representing how much business owners are willing to endure fluctuations in their consumption from year to year, has a value less than one, then Chamley and Judd’s models actually show a positive optimal rate of capital taxation, instead of zero. Interestingly, empirical estimates generally tend to put this parameter at about 0.5 or lower, at least for the economy as a whole.
In other words, even the theory most commonly cited for no taxes on capital actually says that taxes probably shouldn’t be zero.
Of course, that doesn’t mean that capital taxes should be high, or that the basic intuition for why they’re harmful is wrong. It’s important to encourage businesses to invest for the future — not only does investment give people jobs and raise wages in the short term, but it raises living standards for the future. What papers like Piketty and Saez’s, or Straub and Werning’s, really tell us is that economic theory needs to be disciplined by hard data. That’s why the empirical revolution in economics is so important.
And what does the data say? In 2003, under President George W. Bush, the U.S. lowered the top rate for taxes on dividends — a type of capital tax — from 38.3 percent to 15 percent. Eight years later, economist Danny Yagan measured how much this tax cut had boosted business investment, by comparing C-corporations (which are subject to dividend taxes) with S-corporations (which are not). He found that the tax cut resulted in no change between the capital spending of the two types of corporations — in other words, companies ignored tax rates when making their investment plans. That result undercuts the case for low dividend taxes — and also for low capital-gains taxes, which are similar in nature.
The history of the debate about capital taxes shows that even the smartest economists struggle to make definite recommendations when theory is the only guide. When making high-stakes decisions on things like federal tax rates, it’s best to stick as close as possible to the data.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Noah Smith is a Bloomberg Opinion columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at Noahpinion.
©2018 Bloomberg L.P.