July 08, 2002
Emmanuel Saez Has Evidence That Supply-Side Effects Are Not Large

One of the nicest things to happen to us here in Berkeley's Economics Department this year was our successfully persuading Emmanuel Saez that he wants to move from Harvard to Berkeley. Here Emmanuel looks at the number of people found at the "kinks" in the tax code. If supply-side effects are large, then lots of people who would otherwise work more (and be in the higher bracket) will decide to quit when their forecasted income reaches the point at which the marginal tax rate jumps up. So Emmanuel looks for kinks, and finds less evidence of them than I would have thought (save for those reporting lots of self-employment income, but that's another taxp policy story).

It's simple. It's ingenious. It's well-executed. It's convincing. All in all, very nice to see.

Emmanuel Saez (2002), " "Do Taxpayers Bunch at Kink Points?"" (Cambridge: Harvard University xerox).

Abstract: This paper uses individual tax returns micro data from 1960 to 1997 to analyze whether taxpayers bunch at the kink points of the U.S. income tax schedule generated by jumps in marginal tax rates. Clear evidence of bunching is found only at the threshold of the first tax bracket where tax liability starts. Evidence for other kink points is weak or null. Evidence of bunching is stronger for itemizers than for non-itemizers. The large jumps in marginal tax rates created by the Earned Income Tax Credit generate very little bunching except for recipients reporting substantial self-employment income.

In the standard micro-economic model, the amount of bunching should be proportional to the size of the compensated elasticity of earnings with respect to tax rates. We introduce uncertainty and rigidities in labor supply choices to account for the empirical results. Numerical simulations show that, even in those cases, behavioral elasticities consistent with the empirical results are small.

Keywords: Bunching, Budget set kink points, responses to taxation.

Posted by DeLong at July 08, 2002 09:18 AM | Trackback

Email this entry
Email a link to this entry to:

Your email address:

Message (optional):


It is cool to see an economist trying to scientifically test a theory, almost as if economics was a hard science. But unlike what would occur in a real science, few if any economists are going to change their views just because somebody proves them wrong.

Posted by: Isaiah on July 8, 2002 11:06 PM

Very interesting, thanks! This'll help get people to put away that darn napkin.


Posted by: Steve on July 9, 2002 06:32 AM

Probably I just don't know what "supply-side effects" are. Does this
have something to do with "supply-side theory?" Obviously I don't
know for sure what is being talked about here.

What I thought the supply-side business was about was first noting
that different investments have different economic impacts. Getting
really general, and focusing on productivity, if we speak of the government,
then a dollar spent is going to have some average impact on
productivity, either to increase or to decrease. That could be
compared to the impact of a dollar spent by the "non-government,"
which is going to have a different average impact on productivity.

Or again maybe we're not comparing productivity but gross economic
activity. Obviously the comparison could be made for many factors
of interest.

So probably greatly oversimplified I thought the key idea of
the "supply-side" business was that expenditures by government
are less productive than by "non-government" and thus basically
one would want to minimize government expenditures generally,
but more specifically, in some contexts, one might cut government
expenditures and taxation and greatly accelerate economic growth.

So if I'm on the right track how does Emmanuel Saez's paper asserting
that taxpayers don't bunch at taxation kink points refute this?

Posted by: Mark Amerman on July 9, 2002 11:48 AM

Although the economic literature ususally looks at the realtionship between tax rates and labor supply but I believe a better empirical study would look at the relationship between tax rates and compliance or avoidance.

Posted by: EcoDude on July 9, 2002 02:39 PM

Although the economic literature ususally looks at the realtionship between tax rates and labor supply but I believe a better empirical study would look at the relationship between tax rates and compliance or avoidance.

Posted by: on July 9, 2002 02:39 PM

To Mr. Amerman above: This study refers to one of the standard arguements within supply-side economic stimulus theory, which is that the higher the taxes than there is a inverse proportional amount people will be inclined to exert or earn. What Mr. Saez seems (I am far, far from an economic brain, by the way, so take this as my humble interpetation) to be saying is that under the supply-side theory there should be points along the tax bracket curve where a bunch of people max out their income efforts and, in a sense, stop earning. His study seems to show that this is not the case.

Posted by: A. Mendoza on July 9, 2002 06:43 PM

Some thoughts about this paper, in disjointed order of importance:

* It was my understanding that the consensus is supply response is small with the first income of the family but much larger with second incomes such as of the spouse and kids. The data presented doesn't seem inconsistent with that, since the strongest supply response is with lower incomes and self-employeds which would be the second incomes. But considering the importance of second incomes to household welfare (and the economy) these days, I wouldn't automatically conclude the data indicates their response is so unimportant.

* The paper notes but underestimates, I think, the issue of higher-income people having knowledge of their break points. Of course people can't group around a kink if they don't know it's there. And I can say professionally that few people have any idea when they are moving from one tax bracket to another.

More significant than that even may be that they have so many kink points that *aren't* mentioned in the paper.
E.g, Interactions between AGI deduction floors and limits can add 3 marginal points to the tax rate by phasing out itemized deductions, 2 points re exemptions, 10 points re casualty losses, 7.5 points re medical expenses, 2 points re miscellaneous expenses, etc., etc., there are a host of others -- all of which kick in at different unique levels and in different unique mixes per taxpayer. None of these are listed in the tax bracket tables, but they're all just as real -- and combined they pretty much swamp the cited gaps between, say, the 28% and 31% brackets.
Frankly, it would seem a statistical miracle to me if people were found grouped around those higher brackets. Heck, most people don't even know that they are getting hit by the dreaded AMT, which wipes out everything above, until after the year is done when Turbo Tax delivers the news.

Now, if Congress designs a tax system so opaque that it is impossible for people to group around kink points at higher income levels, does that mean that a lack of grouping around kink points indicates a low supply response to taxes?

* It's risky to dismiss the effect of nonfilers on the data. The IRS estimates there were 55 million of them -- 30% of everybody -- the last time it dared to look, in 1996. The information it has on their profile is very anemic because, well, they didn't file to say. (And of the 55 million the IRS pursued only about 2% to find out. It doesn't publicize these numbers much.) I guess it probably wouldn't affect the result here ... but when talking about tax return data generally, people should keep the size of this data gap in mind.

* Some of the other papers on the page look interesting too. The Zilak, et. al. paper looks at labor response from a different point of view and reaches a different opinion. I'm going to read the rest as soon as I get the chance. Thanks for the pointer! Taxes are fun!

BTW, I know I run on too long in my comments here sometimes, so if there are any "guidelines for comments" such as "keep it under 99 words" let me know and I will obey.

Posted by: Jim Glass on July 9, 2002 08:46 PM
Post a comment

Email Address:



Remember info?