I have heard a lot about “trickle-down economics” lately. The President has taken to using it in speeches. And pundits have increasingly invoked the idea. Back in February, I was asked about the term when I testified before a House committee and had to confess that I have never met an economist who has advocated anything close to “trickle down” economics.The words “trickle down” imply that if you redistribute money to the wealthy, they will spend it (say, by hiring workers or by buying products) and it will somehow find its way into the hands of the poor. To the extent that any economists endorse such a notion, they are emphatically not free market economists.
This is not to say that there is no case for low taxation. There is a strong theoretical case for low taxation (so long as it is accompanied by low spending!). And it is backed by good empirical evidence.
But the case for low taxation is not—as the phrase “trickle down” implies—based on the idea that we should give money to a wealthy person so she can spend it. Instead, it is based on the idea that if we take money away from either a rich or a poor person when they engage in some activity, they will tend to engage in less of that activity.
If we tax work, people will tend to work less. If we tax consumption, people will tend to consume less. If we tax saving, people will tend to save less. The idea is rooted in basic microeconomics. Taxing labor, for example, makes leisure less expensive. So people choose more leisure. This is called the substitution effect.*
All this theory is well and good, but is there any evidence to back it up? Yes. Michael Keane offers a nice survey of labor supply and taxation studies in the December issue of the Journal of Economic Literature. He identifies at least two major patterns in the evidence:
- Women are more responsive to taxes than men (most economists think men are relatively unresponsive to labor taxes, especially in the short run).
- People—particularly women—are more responsive to taxes when they consider whether to work than they are when they consider how much to work. In the average study, the long-run elasticity for female labor is 3.6. This means that if a tax hike reduces after tax wages by 10 percent, female labor force participation tends to fall by about 36 percent. As Keane puts it, this is a “very large” effect.
In my view, both of these patterns make sense. Historically, women have been more likely than men to work at home and so higher taxes seem more relevant for them than for men (as more women work outside the home and as more men stay home, I’d expect this gender difference to narrow). It also makes sense that taxes have a larger effect on the decision to work at all than on the decision to work a certain number of hours. Most of us can’t tell our employers that we want to work 30 hours a week rather than 40. But we can tell our employer that we don’t want to work at all. And evidently a lot of people—particularly women—do tell their employers this when taxes are high.
So far, I’ve only discussed how taxes affect labor supply. But they may also depress consumption and investment. What is the overall effect on the economy?
One of the best recent studies is that by President Obama’s former economic advisor, Christina Romer and her husband, macroeconomist David Romer. The Romers set out to understand the effect of taxation on an economy. But they knew that there was a major problem: taxes are not randomly increased or decreased. Instead, politicians tend to keep taxes low when the economy is in recession and raise them when the economy is booming. This makes it very difficult to disentangle cause and effect. So the Romers painstakingly analyzed decades of presidential speeches and government documents to identify exogenous tax changes (i.e., changes that were undertaken for reasons other than the condition of the economy). They then compared the performance of the economy following such exogenous changes. They concluded that exogenous tax increases are “highly contractionary.” As they put it in the conclusion:
Our results indicate that tax changes have very large effects on output. Our baseline specification implies that an exogenous tax increase of one percent of GDP lowers real GDP by almost three percent.
Now here is the irony: As I note above, few if any economists advocate redistributing resources to the wealthy in the hopes that they will trickle down to the rest of us. But over the objection of economists—particularly free market economists—policy makers do this all the time. Think of President Bush’s TARP. Or President Obama’s decision to extend TARP to the auto companies. Or his excursions into venture capital. In each case, money was actually transferred from taxpayers to the (mostly) wealthy managers and shareholders of private firms.
If words mean anything, each of these policies—and not, say, an across the board reduction in marginal income tax rates—should be labeled “trickle-down economics.” But in politics, words often mean nothing.
* You might be thinking that the income effect offsets this: By taxing income, you not only make leisure less expensive, you also make people feel poorer. In response to feeling poorer, they may feel that they need to work harder to make up for the loss income. This works for an individual, but as economists James Gwartney and Richard Stroup long ago explained in the American Economic Review, it does not work for society as a whole. This is because governments do something with the money they collect in taxes. And the income effect of spending government revenue makes people work less. So at the economy-wide level, the income effect from spending offsets the income effect from taxing. All you have left is the substitution effect and that unambiguously reduces labor supply.