Will a wealth tax change the behavior of the very rich?

Do rich people respond to incentives?
Do rich people respond to incentives?
Image: Reuters/Eric Gaillard
We may earn a commission from links on this page.

When I studied economics as a graduate student, we took it for granted you could change behavior through tax policy. If you increased taxes, for example, people might work less, take fewer risks, or find clever ways to avoid taxes. But some economists are questioning this assumption and it could have a big impact on economic policy and the taxes you pay.

In the last few decades economists have relaxed their assumptions that people behave rationally, or in a consistent, logical manner when faced with various financial incentives. But now they are questioning if we respond to incentives much at all. The newest economics Nobel Prize winners Esther Duflo and Abhijit Banerjee argue, based on their experiments and various studies, that we don’t respond to financial incentives as much as economists think—sometimes not at all. The cite evidence that enrolling the poor on welfare doesn’t induce them to work less and survey evidence suggests that when a local economy collapses, its residents aren’t willing to move somewhere that has jobs.

Duflo and Banerjee concede that higher taxes results in more tax avoidance, and maybe even relocation to lower tax jurisdictions, but they are confident that taxes have no impact on how much people work. They conclude: “We should not be unduly scared of raising taxes to pay for[unemployment benefits and re-education programs]. There is no evidence that it would disrupt the economy. This is, of course, a touchy subject politically: The idea of raising taxes on anyone but the very rich is not popular. So we should start with raising the rates on top income and adding a wealth tax, as many have proposed.”

They theorize this is because we don’t just care about money but about dignity, as well. They are not alone. Gabriel Zucman and Emmanuel Saez, Berkeley economists advising presidential candidate Elizabeth Warren, are the ones proposing the wealth tax referred to by Duflo and Banerjee. This is a major departure from traditional economics which assumed wealth taxes are the least desirable of all taxes because they can have a big impact on behavior. It was believed they reduce the incentives to save, take investment risk, or start a businesses by reducing the returns from these activities. Income or capital gains taxes, or, even better, consumption taxes, are harder to evade and presumed to have less impact on behavior. But if you aren’t convinced tax payers respond to incentives, then a wealth tax can be an effective way to increase revenue and reduce inequality with no harm to the economy. Zucman and Saez argue the behavioral impacts are not large on the very wealthy and evasion would be limited. Warren proposes a 6% tax on wealth for billionaires and Zucman and Saez support very high income taxes, 75% for high earners, under the presumption it will not distort incentives and harm the economy.

It is hard to infer what drives behavior, and in the same way economists once overestimated how rational people are, they may now be underestimating the effect of incentives. For example, small incentives may have more impact than large ones. There is evidence workers save more when automatically enrolled into retirement accounts with savings rates around 4% to 8%, but once over 10%, employees start to opt out. A new paper by Stanford economist, and soon to be member of Trump’s Council of Economic Advisors, Josh Rauh and graduate student Ryan Shyu, estimate how many Californians moved following a 1% tax increase. The greater their  earnings (and the bigger their tax increase), the more likely they were to move. Californians who earned more than $2 million had a big spike in leaving California, but as income fell, so did the odds of moving.

Or there could be other incentives-driven behavior we don’t see. Duflo and Banerjee argue a big annual payment from the state of Alaska to its residents does not reduce employment there compared to other states without a similar payout. But that payment from the state can vary up to 100% each year because it is based on oil revenue. Labor income, or working, could be how Alaskans hedge income risk since they can’t count on the same state income each year. In Denmark, where savers passively contribute to their retirement accounts, they also acquire abnormally high amounts of debt which undermines that saving. If only retirement accounts are considered, it would seem the Danish are not moved by financial incentives and sock money away at the default rate. But a more complete picture of their household finances suggests something different.

What can all this conflicting information tell us? There may be some room to increase taxes and experiment with different kinds of policy, but some caution is necessary. Perhaps a 0.05% wealth tax could be an effective way to raise with out harming the economy or radically changing behavior, but a 6% tax (which is more than the typical return on investing), may have much larger, unpredictable behavioral effects.