David R. Henderson  

Where Are We on the Laffer Curve?

PSST: Work-Sharing May Not Wo... Did the IGM Reject Laffer Opti...
A cut in federal income tax rates in the US right now would raise taxable income enough so that the annual total tax revenue would be higher within five years than without the tax cut.
This is one of the questions on the IGM Forum, a forum in which many of the country's leading economists give their responses and, if they so desire, their reasons.

Not one economist surveyed agrees with this claim. I don't either. Cutting marginal tax rates will somewhat increase taxable income. But the odds are very high that it wouldn't increase nearly enough to increase tax revenue. Whereas one can make a case that cutting some marginal tax rates would increase revenue--I have in mind the corporate income tax, for example--it is hard to make a case that cutting marginal tax rates on individuals would increase revenue.

To see why, consider someone paying a marginal tax rate of 25%. If the feds cut the tax rate by 10%--that is, 2.5 percentage points--then his incentive to earn an additional dollar rises from 75 cents on the dollar to 77.5 cents on the dollar. (I'm assuming he's past the Social Security threshold so that his marginal FICA tax is 0. I'm also assuming he's in a 0 state income tax state such as Texas. Finally, I'm just ignoring the 1.45% Medicare tax.) For the feds to make the same revenue on him, his income must rise by one ninth, or 11.1%. That seems highly unlikely.

Mark Thoma gives a particularly misleading title to his post highlighting these findings. He calls it "Laughing at the Laffer Curve." But nothing in the above says that the Laffer Curve is wrong. In fact, the Laffer Curve must be right. At a zero tax rate, the government will collect zero revenue. At a 100% tax rate perfectly enforced, there will be literally zero incentive to make income and so the government will collect zero revenue. At in-between tax rates, the government collects some revenue. QED.

So the question is not whether the Laffer Curve is right. The question is where we are on the Laffer Curve.

Comments and Sharing


COMMENTS (23 to date)
OneEyedMan writes:

The Laffer Curve is plausible even likely, but societies without taxes still pay for public projects through non-tax means and there are communitarian and religious orders that manage 100% tax rates (albeit with free exit).

Sonic Charmer writes:

Yes it's always weird when economists of a certain stripe 'laugh at the Laffwr curve' in such a way as to reveal that they don't know what it is (or pretend not to).

To me the importance of the Laffer curve observation is twofold: 1) somewhere, tax revenues max out vs tax rates, but also (relatedly), 2) tax revenues are a negatively-convex function of tax rates.

Everyone knows about and focuses on #1 (even folks like Thoma who interpret it wrongly) but few seem to pay attention to #2, which to me is actually the most important point. After all, while the existence of a max is theoretically interesting, it's not as if the government should be trying to maximize tax revenues to the exclusion of all other considerations. But the idea that (even if you ARE to the left of the Laffer curve) that second incremental x% increase in rates won't raise as much money as the first, ie that there are diminishing returns to how much tax revenue you can raise by increasing rates, no matter WHERE we are on the curve at any given time, seems very relevant to discussions of optimal tax rates yet rarely discussed or even grasped, seems to me.

Or is that just me?

liberty writes:

"At a zero tax rate, the government will collect zero revenue. At a 100% tax rate perfectly enforced, there will be literally zero incentive to make income and so the government will collect zero revenue. At in-between tax rates, the government collects some revenue. QED."

It is perhaps a trivial objection to point out that there are non-economic incentives to working such that, even at 100% tax rate, someone might still choose to work - especially if s/he has another income source, such as inheritance or capital gains. However, this objection is less trivial if it is applied at other places on the curve. In other words, we must keep in mind that a change in the marginal rate from 25% to 30% will discourage some people at the margin; but many others will not consider it noteworthy, as only such changes as prevent them from being able to earn enough to live comfortably will matter to them - otherwise it is the non-economic considerations which matter to them, such as enjoyment of the job, fulfillment, ability to help others, etc.

Mark Thoma writes:

1. I do laugh at people who say that if we cut taxes from current levels, revenues will go up. What is misleading about saying so? Nothing.

2. The assertion that GDP would be zero with 100% tax rates is not well thought out (and hence particularly misleading). It depends upon the underlying distribution mechanism. Suppose we tax at 100%, then redistribute to give everyone an equal share, an idealized form of the old soviet economy (think of it in goods/services if you like, and you can make the goods and services different from what the person would have chosen in a freer market, or give them a value exactly equal to their marginal product, but not the exact goods they would have purchased). Would GDP be zero? Nope. The connection between production and distribution must be considered.

3. Where did I ever say there was no max revenue tax rate? Nowhere. Saying I have is just plain wrong. The question is how to efficiently and equitably pay for the the things we decide, collectively, government should provide.

4. Just to emphasize the first point, I *am* laughing at the people who are trying to say that if we cut taxes from where they are now, revenues will go up, and there's no shortage of politicians saying this. I won't apologize for trying to rebut such nonsense.

So I don't think I'm the one making misleading claims here...

Grant Gould writes:

The "where are we on the Laffer curve" debates have definitely crossed the line into silliness many years ago. But the explicit 5 year horizon of the original question raises a more interesting line of approach.

The Laffer curve in effect measures elasticities, and as I was taught back in the dark ages elasticities are never constant but are (generally increasing) functions of time -- nothing is particularly elastic one minute out, whereas nearly everything is quite elastic after a hundred years. Correspondingly we should expect the peak of the one-year Laffer curve to lie well to the left of that of the ten-year Laffer curve (after ten years everyone has had a chance to adjust their hours worked, level of tax evasion, investment portfolio, and so forth, even leaving aside economic growth). On this view nearly any given tax rate is revenue-maximizing on some time scale, and the question we should be asking of a tax rate is on what scale it is revenue maximizing.

To put it another way, imagine the same-year-revenue-maximizing tax rate had been 80% in 1912. Hardly anyone would argue that that rate would have maximized tax revenues today -- the economy would be smaller and the magic of compounding the rate of growth suggests it would be much smaller; the hundred-year-revenue-maximizing tax rate would be somewhere nearer, say, 20%.

Revenue-maximizing tax rate versus time would be a much more interesting curve than the rather overplayed two-zeroes-and-the-mean-value-theorem Laffer curve. It would also put a much more informative number on the stance of tax policy at any given point ("candidate A prefers a 25-year-maximizing tax rate, while candidate B prefers a 50-year-maximizing rate...").

Mark Thoma writes:

One other note:

Making a big deal out of a title that is intended to be a play on words without noting that I included statements from Bob Hall about precisely where we are on the Laffer curve (even though I disagree we are close to the peak given -- recent research says otherwise) is what is "particularly misleading." I mean really, focusing on the title and ignoring the content that makes your points?

Also, you end with "So the question is not whether the Laffer Curve is right. The question is where we are on the Laffer Curve."

Uhm, that's the point everyone in the forum is making -- we're to the left of it, and there is little doubt about that. We can argue how close we ae to the peak, but that has nothing to do with the fact that cutting taxes will cut revenues.

So your question has already been answered -- that was the point of the post. Like it or not, all indications are that we are to the left of the peak.

Pat writes:

Mark Thoma,

If a country has a 100% tax rate but then writes everyone a check back for 20%, the net effect is that it's not a 100% tax rate, it's 80% (ignoring the costs of collecting taxes and writing checks).

The laffer curve still holds as long as you adjust it for the net redistribution.

If you tax someone 100% while getting nothing back in redistribution, they're not going to work.

Since the whole point of redistribution is to take from some people to give to others, there are going to be people who are net tax losers. We can't all win from redistribution.

MG writes:

We may be at a point in the Laffer curve where a shift to lower taxes reduces, in the short-term, taxes (government's share of GDP). But we may also, simulataneously, be at a point in the "Rahm curve" (as the Cato institute calls it!) were a shift to lower tax rates (even simple, broad based taxes) would increase GDP growth. In this case, there would be more wealth to share, even if the government's initial claim on incremental private sector wealth does not recoup immediately foregone taxes. I would certainly not laugh at two propositions: That redirection of investment away from public hands into private hands could produce enough additional growth in the medium term to make up for the lost revenues in the short term. And that distribution of wealth back to its creators (who can then get to decide how to recycle it) is fair.

CKE writes:

Certainly there is some truth in the Laffer Curve, in that taxes do motivate people to work or invest (or not), and this will affect the total future tax take.

But there are other things that get wrapped up in the idea. A 10% tax cut coupled with a 10% spending cut might very well produce more tax revenue 5 or 10 ten years down the road. But, in this scenario, two factors are involved: motivation and resources. People might be motivated to work and invest more. Plus, resources freed from government spending can be invested to grow productive capacity, and hence output.

Put another way, if I simply loan my tax savings back to the government so it can keep spending the same, I must work more and likely consume less in order to grow future income and tax recepts. This is a tall order for a tax cut. Investment that increases productivity is the real driver of growth, and tax cuts may (or may not) expand investment resources.

Ken B writes:
I mean really, focusing on the title and ignoring the content that makes your points?
Indeed. Why the two needn't be connected at all, titles are just personal expression. If you post a rebuttal titled "Sleazy David Henderson has a Hissy Fit" and the content of the post is just a recipe for carrot cake who can complain?
Thomas Boyle writes:

I find these hypotheticals that ignore realities like state taxes and AMT rather silly.

In order to do what they do, many people have to live in one of a handful of places, that (not coincidentally) have high state taxes (the taxes are high because they can be).

Someone paying AMT and living in California or the Northeast pays 35% federal (the "28%" rate on AMT is an effective 35% when allowance phaseouts are considered), plus 10%-ish state tax. That's 45% or more.

Even in Europe, 50% marginal rate is considered something of a threshold beyond which taxes are simply immoral (and eventually become socially acceptable to evade - which is another problem). Even with the Occupy Wall Street/the City movement, the UK recently backed away from 50% rates; only Sweden and Denmark have marginal rates above 50%.

At 50%, many people are right up there on their personal Laffer Curves. In California and the northeast, at a 50%-rate income, they're living in rented apartments, driving modest (if any) cars, and wondering how come they're being blamed for the country's ills. If the rates go above 50%, I don't know how many of them really will spend the remaining hours of their lives working primarily for the glorification of Congresscritters, and how many will dial it back and start to prioritize quality time with their families. I don't know how many already have.

At first, of course, people keep doing what they're doing; there's a fair bit of inertia. That's why liberals think tax rates (and regulations) don't matter. It's 10, 20, 30 years down the road that society pays the price.

There's a lot of ruin in a country. But, break the social compact, set punitive tax rates, and vilify the taxpayers, and it can indeed be ruined.

Tom West writes:

I think the main problem is that the Laffer curve is usually on brought up in the context of an excuse for cutting taxes while not having to cut any programs.

In other words, while the Laffer curve is not duplicitous concept in and of itself, in a political context, it's pretty much exclusively used in a duplicitous fashion. (We can cut taxes *and* raise revenue!) This has resulted in the concept being tarred by the context in which it is usually used.

So, no, the Laffer curve doesn't deserve the disdain it gets, but let's face it, its those who've tried to sell tax cuts as 'painless' who are responsible for besmirching its name.

(And just to be clear - this has *nothing* to do with those who sell tax cuts as a necessary pain in order to get out financial house in order, improve productivity, etc., etc.)

Andrew writes:
The assertion that GDP would be zero with 100% tax rates is not well thought out (and hence particularly misleading).

Mr Thoma:

The Laffer Curve states that taxable GDP will be zero not that GDP will be zero as you approach 100% tax rates. People will still engage in economic activity, there just won't be any remitted to government. See the wars on Drugs and Prostitution for real world examples.

Secondly, even if you redistribute all taxes equally, over time the revenue from 100% taxation will approach zero. The only people who will labor will be the fools who believe their labor will be subsidized by others. Eventually even the stupid realize they are getting less than they put in.

Radford Neal writes:

liberty writes: It is perhaps a trivial objection to point out that there are non-economic incentives to working such that, even at 100% tax rate, someone might still choose to work...

Yes, they might choose to work, but why would anyone choose to pay them, considering that they would work the same amount (on the same things) without being paid?

So tax revenue would indeed go to zero.

Lars P writes:

I want to point out that the Laffer Curve is three dimensional.

That is, effects of changes in incentives are not immediate, they play out over decades.

That is, when my tax rate is changed, it might not change most people's day to day behavior much in the short term, but in the long term it affects people's career and life choices, which may take a generation to fully propagate.

This of course makes the curve even more impossible to measure than if you only think of the immediate effects...

happyjuggler0 writes:

I realize there was only a semi-serious attempt to look at a 100% tax rate, but I will take it seriously. Here goes:

It is hard to wrap my mind around the idea of a country that has a 100% personal income tax rate. I also have a hard time imagining that any government that would impose such a rate would also not impose a 100% rate on corporate income to "make them pay their fair share", along with a 100% tax on capital gains and dividends, again for the same reason of making them "pay their fair share".

Therefore, I think it is reasonable to posit that there would have been an unprecedented amount of capital flight before that point; anything not nailed down would have left the country.

There also would have been a huge flight of high skilled human capital too. Why does the US currently have a large number of doctors who speak English with a heavy accent? Is it because of their humanitarian impulse to forgo serving in poorer countries in order to serve the health needs of the wealthiest country in the world? Or is it because we "showed them the money" while their country of origin was unwilling or unable to do so?

The city of Detroit long ago lost the vast bulk of its existing high quality human capital, while the vast bulk of its emerging high quality human capital (i.e. recent high school graduates) also leaves each year.

The city of Detroit still functions (sort of), but that is only because of a transfer of wealth from places that have chosen not to be on the wrong side of the "tax, labor regulation and crime" Laffer Curve.

If the US were to impose such 100% taxation, there would be no such wealth transfer, and as a result the economic implosion resulting from capital flight (including human capital flight) would be near instantaneous and devastating.

Therefore I am inclined to think that either US voters or elected US officials would stop their idiotic high tax experiment long before it got to that point.

Indeed I would argue that that has already happened, not just in the US but around the developed world. Starting in the late 70's, and carrying on into the 80's and 90's (depending on country) the experiment of running headlong into socialism or extreme social democracy (depending on country) came to a screeching halt precisely because it was either self evident amongst a majority that it was indeed failing, or because of capital flight (including human capital flight) that forced the hand of reluctant governments to dramatically reduce marginal tax rates.

To take the example of the US, we've been bouncing around between 28% and 40% for the top marginal tax rate (not counting entitlement taxes or state tax, or the invisible employer side of such taxes) for two and a half decades now. I would argue that that is our steady state, and that anyone who tries to take us out of either side of that 28-40% band will unceremoniously get booted out of office, and we will then moderate towards the other side of that tax band.

Tax rates upon whom? I suppose there must be research upon this question, but I did not notice it in the immediate posting and comments.

A desperately poor person might work with even 90% tax if the remaining 10% gives him his only way to get a meal.

But a person with some reserves works not for today's food but for longer-term and more abstract satisfactions. Such a person might be dissuaded from going to work if the government is going to take half, or some other percentage.

pyroseed13 writes:

Mark Thoma seems to think that the economists are disputing where we are on the Laffer Curve, but I'm not so sure they are. I'm seeing denial that such a curve even exists:

David Autor: "Not aware of any evidence in recent history where tax cuts actually raise revenue."

Austin Goolsbee: "Tax cuts lose revenue."

Kenneth Judd: "That did not happen in the past."

Anil Kashap: "Not true empirically in the U.S."

It seems that Darrel Duffe gave what I think is the only reasonable answer: "A lower tax rate applied to higher earnings could raise or lower tax revenue, depending on the extent of growth."

The one thing we know from observing tax revenues is that they rise during booms and fall during recessions. Since economists don't seem to think that tax cuts are contractionary, as evidenced by their response to the first question, why do so many find it implausible that a tax cut in some instances could increase revenues by stimulating economic growth?

Brandon Berg writes:

It's worth pointing out that "Where are we on the Laffer Curve?" is not a question with a binary answer.

Even when you're on the left side of the curve, the specific position---or more to the point, the slope---still matters. When you're far from the peak and the slope is steep, taxation is relatively efficient, with a low ratio of deadweight loss to revenues. But as you approach the peak and the slope begins to level out, you incur more and more deadweight loss for each additional unit of revenue. Raising taxes will still result in extra revenue, but it may still be contraindicated due to high cost in terms of deadweight loss.

Joe Cushing writes:

I've come to the conclusion that Taxes don't matter as much as spending does. You can cut the crap out of taxes but if the government is spending like make, it is still taking money out of the productive sector. Sure taking rates from 100% to 35% will make a big difference but taking rates from 35% to 20% and then borrowing trillions of dollars probably won't make a bit of difference in the growth rate.

Jim Glass writes:

Mark Thoma writes:

The assertion that GDP would be zero with 100% tax rates is not well thought out (and hence particularly misleading). It depends upon the underlying distribution mechanism. Suppose we tax at 100%, then redistribute to give everyone an equal share, an idealized form of the old soviet economy...
When I am offered a job that pays me zero -- due to a 100% tax on my earnings from it -- I have zero economic incentive to accept it ... especially if I have been promised that I will be able to live off of transfers from others.

Does Mark Thoma truly believe otherwise?

And a 21st century economist citing the Soviet model as a justification of anything ... really??

I do laugh at people who say that if we cut taxes from current levels, revenues will go up. What is misleading about saying so? Nothing.
What is misleading is he didn't say he was laughing "at people" who naively mistake the implications of the Laffer Curve, he said he was laughing at the Curve itself.

I laugh at the political partisans who naively mistake the implications of Mark Thoma's posts on his blog. So what is wrong with me writing "I Laugh at Mark Thoma"? Apparently nothing.

Steve Roth writes:

There is empirical research addressing this question.


“The estimated impact of a rise in the after-tax share is consistently positive, small, and precisely estimated” pp. 15–16). They find an elasticity of taxable income with respect to changes in the after-tax income share of 0.19.

"an elasticity of 0.19 implies that tax revenues would be maximized with a tax rate of 84 percent; that is, you could raise taxes up to 84 percent before people’s reduced incentives to make money would compensate for the higher tax rates."

See also here on the optimal level for maximizing per capita GDP growth:



Ken B writes:

Dan Mitchell has a good short introduction in this video.

Comments for this entry have been closed
Return to top