Asymmetrical Information - Megan McArdle

10.15.12

Do Higher Taxes Make Us Work More, or Less?

No matter what you may think, there's no obvious answer.

One of my quixotic blog battles has been against the nonsense of revenue-raising supply-side tax cuts. It's not that it's impossible for tax cuts to raise revenue--at some level of taxation, higher taxes really are counterproductive, because they depress the work incentive too much.  You will not raise much money with a 100% effective tax rate.  But there's no evidence that US tax rates are anywhere near that level, and considerable evidence that cutting marginal tax rates in America has, in fact, cost the government money.  

This is in part because of the naive mental model that supply-siders tend to have about how taxation affects work effort.  It's true that when your tax rate goes up, you now get less for every hour of labor.  All else equal, that will make you less willing to trade away an hour of labor.  This is what economists call the subsitution effect: when the relative prices of things change, you reallocate your consumption of various goods (in this case, leisure), in order to maximize your utility at the new prices.

But all else is never equal.  If we tax away some of your hourly wage, that means that you earn less for working an additional hour.  But it also means that you have less money from all the hours you're already working, which is to say, that you're suddenly poorer.  And what do people do when they suddenly get poorer?  

Often, they work more to make up the gap.  

This is what economists call the income effect: your desire for various goods changes as your income changes.  Few billionaires invest as much in Kraft Mac and Cheese as the average lower-middle-class household.  And it applies to trading leisure for work.  Public policy can influence this--people on disability rarely work at all, because doing so threatens their benefits and Medicaid.  But generally, the fewer material wants that you can fill with your current work output, the more incentive you have to work extra hours.  

These two countervailing effects make it difficult to tell ex ante whether raising taxes will increase work output, or decrease it.  Empirically, we think it decreases it, but not by enough to raise revenue unless tax rates are much higher than they are now in the United States.  This is the standard liberal rebuttal to the supply-siders in the GOP, and it's pretty strong.

Now Garrett Jones has an excellent post that complicates the story still further:  

Here's the Wisdom of Prescott: If the government raises the tax rate on you and all of your friends, and then divides up all the tax revenue and dumps it from a helicopter, what do you get?  Well, none of the money gets wasted, so the tax hike doesn't have a direct income effect (there's a small indirect one I'll ignore here). 

If the tax hike is used for pure redistribution from the "average person" back to the "average person," then the tax hike doesn't make the "average person" poorer: The government is taking money out of everyone's right pocket and slipping it into their left.  

But if the income effect is gone, what's left?  The disincentive to work: The pure substitution effect. 

So here's Prescott's Big Idea: 

If higher taxes are wasted, then a tax hike has a small, ambiguous effect on employment.

If higher taxes are spent wisely, then a tax hike causes a big fall in employment. 

Not quite what you expected, was it?  

Which is why we should all be careful of recourse to Econ 101, or even Econ 301, arguments.  There may be an even more counterintuitive intuition waiting around the corner.