At the Margin

Yesterday, Avedon said:

Just for the record, I have never in my life met anyone who quit working because their taxes were too high, nor have I ever even heard of someone who suddenly wanted to work harder because of a tax break.

But this argument seems to prove too much. I never heard of someone who didn't buy a product because it cost $0.01 more than he wanted to pay. Still, I take it that we all agree that price is a factor in purchasing decisions. Nobody would be shocked if I told you about a scenario where I was offered $150 to write a column and turned it down because I was too busy at the moment but then reconsidered when the offer was upped to $200. But of course the value in terms of take home pay of a $200 freelance gig is very different (especially in places with a progressive state income tax) depending on what tax bracket you're in. So I don't think it's crazy to think that marginal income tax rates could have an impact on people's willingness to take these kinds of assignments.

When you extrapolate out to something like Ezra Klein's example of a high-paid CEO it really is hard to imagine the tax incentives driving effort. The difference is that the quanta of effort available to a CEO a very large. Basically, you can either do the job and work the long hours it entails or else you can quit and do something entirely different. There's very little ability to respond to small changes in marginal tax rates with small changes in behavior.

And that's how it goes with all of this -- a change needs to be big enough for there to be some alternative course of action that it makes sense to take. In practice, we debate the top income tax rate within a very narrow band so we don't see much impact. But that's not because the theory that tax changes have consequences is crazy -- it's just because the contemplated changes are small.

Comments

In practice, we debate the top income tax rate within a very narrow band so we don't see much impact. But that's not because the theory that tax changes have consequences is crazy -- it's just because the contemplated changes are small.

But that's the whole point. Because in real life no one contemplates making large changes to the marginal rates, it's grossly misleading to talk as if these theoretical consequences should control the the policy choices.

Posted by gordonminor | February 9, 2008 12:34 PM

Back thirty or forty years ago, high-end marginal tax rates ran around 80% or 90% I think. Is there really any evidence that CEOs or other high earners of the 1950s or 1960s were significantly less productive than their current counter-parts? Their companies certainly were doing much better in many cases.

One benefit of such very high marginal tax rates is that there's less of an incentive for a CEO to spend all of his time trying to maximize how much personal wealth he can extract (e.g. steal) from his company, often at the expense of shareholders, employees, the public, or even the company's long-term survival.

For example, how much of the current collapse of so many big financial firms has been caused by the "double or nothing" mentality of their CEOs, brought about by low marginal tax rates?

Posted by RKU | February 9, 2008 12:34 PM

it's not that there isn't a tax level (let's call it near-confiscatory) that could cause someone not to take on extra work; it's that american tax rates aren't at those levels.

whether the highest marginal rate is 32% or 34% or 36% or whatever it turns out to be is completely irrelevant to decision-making about additional earnings: it is, in essence, your $.01 differential.

Posted by howard | February 9, 2008 12:35 PM

What you leave out of the question is moral hazard. I think the point Avedon is trying to make is not that the upper tax level is irrelevant, but that there is little moral hazard in moving the marginal tax rate from 35 to 39 because people will not stop working based on the change. The question becomes when will people stop working, and there are many more factors than simply the amount of money they earn.

You use your own example of being offered the $150 to write a column. You turned it down, but not because it wasn't enough money, but because the money combined with other variables did not match your need. You obviously did not need to $150, it would have been a luxury in your life. But also, you didn't think the other benefits you would get out of writing this column would offset the income related to effort. Most short story writers sell their work for minimal if any pay and consider it a great victory - which is often is. I publish in academia where it is actually considered low worth if you are paid. Academics desperately try to get their work published for free, because it adds to their reputation.

The upshot is, an individual is not going to stop working because of a marginal tax rate if he needs the money, only if the money is a luxury (and then he could stop working at any time anyway so the raise in the marginal tax rate has little meaning anyway). As a society what we need to do is figure out the moral hazard of raising marginal tax rates,which is dynamic, not make the argument that raising marginal tax rates will necessarily make people not want to work.

Posted by Wilbur | February 9, 2008 12:37 PM

What gordon said.

Neither Avedon nor Ezra is unaware of the theory of marginal value. They just argue that it's unimportant to the actual discussion of actual people and actual money that is happening in America.

Turning away from real people and real issues to issue a lecture on an effectively irrelevant theoretical discussion is the classic trope of the intellectually bankrupt economist. Don't do that.

Posted by DivGuy | February 9, 2008 12:46 PM

Depends. I have saved up enough money that I could plausibly retire early. But I figure I'll keep working. But what if the FICA tax went up significantly? I'd probably just retire.

Posted by Adam Herman | February 9, 2008 12:48 PM

I don't know but I wonder if Avedon wrote that in response to the insanity going over at McMegan's place since Thursday. That is one stupid group of regular commenters over there. While McMegan might have an economics degree, she doesn't seem to get the human condition part that throws a real monkey wrench into things(which I think either Barry Ritholtz or Krugman has talked about). The idiots over at McMegan's seem to think that guy that heads Blackstone, or others like him, will pack it in if we raise his taxes. That is so laughable. Warren Buffett has said many times that the rich aren't gonna pack it in just because they make 400 million a year instead of 500 million.

Posted by Joe Klein's conscience | February 9, 2008 12:48 PM

The question is far more relevant at the lower end of the tax tables than at the upper end.

Will my take home pay -- after payroll and income taxes -- for the part time job at the library be enough to pay for the daycare that I would need for my kids?

That decision is made all the time by women with kids. And tax rates are a part of the decision.

Posted by flory | February 9, 2008 12:52 PM

Another point: Usually I work, but my wife doesn't. Whenever we contemplate her getting a job, we then come to the conclusion that it might not be worth it. Every single dollar she earns would be taxed at a high rate. Add that to the expenses that go along with having a job, and it's hardly worth bothering. So taht's how our tax system affects my family's productivity today. it wouldn't take much more for me to also get out.

Posted by Adam Herman | February 9, 2008 12:52 PM

Matt Yglesias: Proudly standing for all those right wing ideas he learned in the Ivy League.

Posted by soullite | February 9, 2008 1:04 PM

RKU: I expect that a very high income tax rate would actually increase the likelihood that CEOs would steal: not "steal" in your sense of "recieve compensation approved by a board or shareholders based on the value they produce" but steal in the real-world sense of taking kickbacks and bribes under the table, evading taxes, or committing fraud. The greater the difference between compensation for illegal (tax-free) activity and legitimate activity, the higher the chance that those in a position to choose will opt for the illegal activity. This is why people in poor neighborhoods deal drugs instead of taking minimum wage jobs - money motivates people. And yes, I've read Freakonomics - but the broader point still holds.

This broader point negates any all-or-nothing argument about taxes. If you think changing the tax rates doesn't affect the economy, then you basically don't believe in capitalism, Adam Smith, or the science of economics. Of course people make decisions based on goods other than money, but money motivates people because it can be exchanged for things they desire. A small amount of money might only motivate someone a little bit, and it might be outweighed by things like time or happiness, but any change that increases or decreases people's take-home pay, even by 1%, will have economic effects. Maybe not changes like increasing productivity by 50% or causing people to quit working, but little changes - working a little longer, feeling a little more satisfaction with the job and producing more efficiently, seeking out new opportunities more eagerly, vs. taking more vacation or cutting out of work early, feeling frustration with inadequate compensation and not putting in quite as much effort, or not pursuing an opportunity. Of course people aren't robots - my girlfriend regularly turns down overtime even though she could make time-and-a-half because she wants to spend time at home - but on a macro level they respond to costs and incentives. Its foolish to pretend they don't for political reasons like wanting to pretend that soaking the rich is consequence-free or that we can have any social programs we want without compromising on anything else. There are trade-offs to any policy change.

Posted by cgaros | February 9, 2008 1:24 PM

What soulite said. If you imagine tax rates starting at 10% and going to 90%, you would keep less of each bracket but you would always have more money than if you didn't work at all. You never go backwards. And in fact in this goofy country, once you clear the FICA threshold, your take home goes up. In addition, its only the marginal rates that matter and the top rate now is too low to begin to influence behavior. And I have never met anyone who stopped working for tax reasons, unless they had enough money. Lower tax rates just would have had them stop working sooner.

Posted by Peter Driscoll | February 9, 2008 1:26 PM

Most labor economists think the "Frisch elasticity of labor supply" (i.e., how much you stop working when taxes go up) is about 0.1 - that is, raise taxes from 20% to 30%, and people reduce their hours by about 1%. Not a big effect.

But macroeconomists usually think it's a lot bigger, maybe about 1. So if that's the case, raising taxes from 20% to 30% would make people reduce their hours by 10%. That seems unrealistic to me, but it tends to roughly fit with the difference between European and American taxes and work hours.

One thing people often forget is that it's the poor, not the rich, who are the most likely to stop working (or go into the underground economy) when you raise taxes on them.

Posted by Mr. Noah | February 9, 2008 1:29 PM

That tax incentives can determine whether someone will work harder - or longer - is certainly true, at least in the lower to middle class income tax brackets. I worked at a meat-packing plant last summer and many of the workers would not do overtime work as they feared it would drive up their tax rates and they would lose a lot of the extra money they made to taxes. Now whether they actually would lose a lot of money, I don't know. I don't profess to be a tax expert. But this belief was certainly prevalent among the workers.

Posted by Josh | February 9, 2008 1:37 PM

JKC: McArdle doesn't have an economics degree, IIRC. She has an English degree and an MBA.

Peter Driscoll: What you're pointing out is what economists call the "income effect", as opposed to the "subsitution effect". When your wages rise, you can afford to purchase more of something, in this case leisure time. But when your wages go up, leisure time is more expensive, so you would naturally purchase less of it. Whether the income effect or the substitution effect wins out is an empirical question. If the income effect overrides the substitution effect, this means that leisure time is what economists call a "Giffen good".

Anyway, Avedon's claim is just bad armchair economics. There are plenty of people who quit working, or work less when their tax rates rise, e.g. low-earning spouses of high earners, who pay their high-earning spouses marginal rates, and for whom FICA taxes don't increase their potential SS earnings. (Though in this case, you have the income effect kicking in as well.)

Amongst the highest earners, I'd imagine that rock stars are the only group that can easily cut back their labor in response to higher marginal tax rates, by scheduling shorter tours. CEOs, hedge fund managers, and sports stars can only work less by retiring early. I'm not sure about Hollywood stars.

Posted by Brock | February 9, 2008 1:48 PM

Actually, the effect of taxes on labor supply is extremely well-established in the macro literature that looks across different countries and time periods. One can quibble that cultural differences between countries like the U.S. and Japan, on the one hand, and European nations, on the other, confound such comparisons, but objections like that just go to the magnitude of the effect, not whether it exists in the first place.

And the point of marginal analysis in economics is not that any *single* person will change their behavior in response to a small change in the tax rate, but that *someone somewhere* in the economy will feel the small change as the proverbial straw that broke the camel's back.

When marginal income tax rates were very high in the 1960s and 1970s (up to around 90% I believe), there was a well-documented labor supply effect not among CEOs (whose jobs, as someone pointed out, tend to be of the "do it or don't do it" variety) but among freelancers--writers, artists, actors, and the like--who worked on discrete projects. Hollywood actors would make at most one picture per year because they would earn only a dime-on-the-dollar for the second one. (In reality, there was some flexibility on this as tax lawyers came up with elaborate compensation contracts to spread earnings over a period of years.)

Posted by Ryan | February 9, 2008 1:49 PM

Brock:
Then how did she get to write at the Economist? Because of her MBA? Did she ever take any economics courses as part of that?

Posted by Joe Klein's conscience | February 9, 2008 1:56 PM

cgaros:

Remember, I was focusing on marginal tax rates at the *high end* not necessary on the tax rates that apply to the vast numerical majority of the population, for which extra dollars have large marginal utility, since they spend nearly everything they earn.

And there's also one crucial difference between a CEO "stealing" and stealing---the latter can put you in prison, while the former will just bankrupt your shareholders. The cost-benefit ratio is totally different.

Remember also there's quite a lot of empirical data in America, namely several decades during which the high-end rates were around 80-90%. Is there any particular evidence that crime, kickbacks, or other illegal (hence untaxed) business activity was dramatically greater during e.g. the 1950s than in more recent decades. That's certainly not my impression of that era.

I think one major problem is that many people seem to assume that "human nature" is almost wholly governed by "greed", namely profit-maximizing activity. Anyone who's familiar with the pattern of recorded human history OR with the modern framework of evolutionary human psycho-biology would never accept this.

Posted by RKU | February 9, 2008 2:06 PM

If this was true there would be bunching around tax rate levels. Basically people not wanting to earn money because their marginal rate goes up from 28 to 33. That isn't shown by the data.

Posted by Rob | February 9, 2008 2:06 PM

It's odd that many of the left (quite rightfully, I might add) get upset when the right discount the work of academics in, say, biology, but then go on to say equally inane things like the quote Matt came across in the original post.

As mentioned by other commenters, there is a long and detailed analysis in the economics literature of the marginal effect of tax rates on hours worked. Econ *theory* doesn't make a prediction one way or the other: perhaps a high tax rate will cause people to shift from work to leisure or perhaps it will cause people to work more hours so they can pay for necessities. That's why empirical work has been done to estimate which way the effect goes.

It turns out that there is a real, and substantial, effect of tax rates on hours worked. Guys like, for instance, Edward Prescott think that differences in marginal tax rates explain a large part of the difference between European and American hours worked. You can believe this or not, but simply ignoring the evidence and relying on ridiculous anecdotes is just as anti-intellectual as ignoring the work of biologists when discussing science.

Posted by cure | February 9, 2008 2:18 PM

I have in fact met someone who refused to work because of his perceived extra taxes.

50 years ago I was tutoring a union welder in algebra for a high school requirement. Sunday wage for union welders there and then was triple scale. This dude (We didn't say dude back then, I'm sure.) would somehow extrapolate the figures in his head to where he thought he would be paying more taxes for working than was his regular hourly wage and couldn't understand how it might average out to a better tax rate next April and refused to work Sundays.

I somehow got him through algebra with a passing grade anyway.

Posted by janes'_kid | February 9, 2008 2:56 PM

Then how did she get to write at the Economist? Because of her MBA? Did she ever take any economics courses as part of that?

SERIOUSLY!! I HAVE BEEN ASKING THIS QUESTION FOR YEARS!!!

Posted by Mr. Noah | February 9, 2008 3:34 PM

avedon's comment is simply inane. the non-working mother example already noted is the most obvious refutation, there's all sorts of empirical work on this, all sorts of folk decide (e.g.) whether to switch jobs or add another one based on the effect of taxes. working folk of all incomes aren't as dumb as the left seems to think, unable to do the basic math to figure out that moving up brackets can be counterproductive, etc.

Posted by dj superflat | February 9, 2008 3:47 PM

For the great majority of people, a reduction in income means that they have to work more, not less. If a taxi driver's hourly earnings go down so he can't make what he needs to live in ten hours on the road, he works eleven, not nine.

And the job of a CEO is not like the job of a factory worker. A CEO has power. Wielding power is fun. CEO's are not about to retire to the golf course because their marginal tax rates go up.

Posted by Bloix | February 9, 2008 4:20 PM

The best empirical analysis of the impact of tax rates on economic activity is Joel Slemrod's Does Atlas Shrug.

http://www.hup.harvard.edu/catalog/SLEDOE.html

He concludes that changes in tax rates can cause substantial tax shifting -- the nature of compensation packages change to take advantage of lower tax rates, on say capital gains or in-kind benefits (company cars etc.), but that there is little impact on actual labor supply or business formation.

It's a good book. Worth owning.

Posted by Dave Meyer | February 9, 2008 5:14 PM

I had an experience this past year similar to Matt's freelance example. I had about $10,000 in 1099 income for a second consulting job. I hated it, but the $10,000 for spending evenings and weekends for two months was worth it.

Well, I just did my income taxes, and the $10k is being taxed at about 45%. I thought the taxes would be pretty bad, but when you add up your top marginal rate, a 15.3% "employment tax," and your state and city taxes, it just might make me think twice about taking on the extra work in the future.

If my taxes were raised higher, then the extra work would be even less attractive, no question about it. This seems to refute the point that high tax rates don't affect one's incentive to work.

I would also point out how much more hateful it is to write the IRS a check for several thousand dollars than it is to have it deducted from your paycheck every two weeks.

Posted by chris | February 9, 2008 6:48 PM

It's easy enough to calculate how much more income a person would have to earn, after a tax cut, to pay the same tax as before. It's just the higher rate divided by the lower rate.

Thus, if the rate's cut from 40% to 36%, a person would have to earn 40/36 = 1.111 times what they earned before. That's 11.1% more, or 4.4 extra hours per week if wages are constant.

You can do it in reverse, too! So if your rate increases from 36% to 40%, you would have to make 36/40 or 90% as much, the equivalent of blowing off 4 hours a week, to avoid paying extra tax.

With that in mind, it's a little easier to ask how plausible a scenario is. How many people are really going to work that much extra? How many people really have a choice to blow off that many hours, even if they wanted to?

What if you're a salaried employee who doesn't see an extra dime, no matter how hard you work? What if you're a CEO, and essentially write your own paycheck?

Posted by Aatos | February 9, 2008 6:58 PM

Nice to meet you then, Matt.

I work 70 hours / 2 weeks instead of 80 because the marginal cost of those last ten hours is so much higher than the marginal cost of the first seventy. Basically, it boils down to the change in tax rate, cost & time of commute, and opportunity costs of what I can do with the extra hours.

Posted by independent | February 9, 2008 7:17 PM

Matt,

You need to keep in mind the marginal case, and who those people are.

For example: you have the guy who owns a successful Dunkin Donuts franchise. Does he invest in another? It all depends on the after tax return. The guy who runs a small CPA firm - does he hire that 10th employee - if the IRS is going to take 50% of that added revenue - then maybe it's not worth the effort.

I think that perhapse your Harvard education leads you to belive that the affluent are all made up of CEO's, Investment Bankers, Big-Law Lawyers, Management Consultants etc. While in reality the vast majority of the affluent are made up of the guy who own the chain of dry cleaners, or the McDonalds franchise - the chain of Jiffy-Lubes the freelance computer savant etc. And these are people who are in a position to work less if the rewards of working more fall too far.

Posted by Jmo | February 9, 2008 7:38 PM

Matt,

You need to keep in mind the marginal case, and who those people are.

For example: you have the guy who owns a successful Dunkin Donuts franchise. Does he invest in another? It all depends on the after tax return. The guy who runs a small CPA firm - does he hire that 10th employee - if the IRS is going to take 50% of that added revenue - then maybe it's not worth the effort.

I think that perhapse your Harvard education leads you to belive that the affluent are all made up of CEO's, Investment Bankers, Big-Law Lawyers, Management Consultants etc. While in reality the vast majority of the affluent are made up of the guy who own the chain of dry cleaners, or the McDonalds franchise - the chain of Jiffy-Lubes the freelance computer savant etc. And these are people who are in a position to work less if the rewards of working more fall too far.

Posted by Jmo | February 9, 2008 7:39 PM

Matt,

You need to keep in mind the marginal case, and who those people are.

For example: you have the guy who owns a successful Dunkin Donuts franchise. Does he invest in another? It all depends on the after tax return. The guy who runs a small CPA firm - does he hire that 10th employee - if the IRS is going to take 50% of that added revenue - then maybe it's not worth the effort.

I think that perhapse your Harvard education leads you to belive that the affluent are all made up of CEO's, Investment Bankers, Big-Law Lawyers, Management Consultants etc. While in reality the vast majority of the affluent are made up of the guy who own the chain of dry cleaners, or the McDonalds franchise - the chain of Jiffy-Lubes the freelance computer savant etc. And these are people who are in a position to work less if the rewards of working more fall too far.

Posted by Jmo | February 9, 2008 7:39 PM

Higher tax rates can make you work harder too -- if you want the same money as before, you have to work more with a higher tax rate. Income effect.

In any case, taxes probably matter more for investment in business expansion than for work effort. But only a small minority of capital gains tax falls on genuine new business expansion as opposed to asset speculation.

I found this post a touch annoying -- it's like the sole purpose was to establish Matt's bona fides as Reasonable Liberal who understands Taxes Actually Can Matter. Well, duh. The question is whether they actually do in the America we live in, at the levels we're talking about. How about doing some actual homework on that?

Posted by mq | February 9, 2008 10:34 PM

In the post above, I meant "taxes probably matter more for investment in business expansion than for work effort" to apply to the U.S. only. I don't see the evidence that U.S. income tax rates, the lowest among the major industrialized countries, have much effect on labor supply (although they affect the timing for capital gains realizations, option exercises, and stuff like that).

And yes, McArdle is an utter fraud as an economist. She doesn't even bother to learn the economics she could easily get by googling whatever topic she's posting on.

Posted by mq | February 9, 2008 10:37 PM

There's no serious doubt that high marginal tax rates discourage economically productive activity. That's why there's been a worldwide trend of reducing high marginal tax rates. It's been especially strong in Europe.

One holdout has been Denmark, which still imposes a top marginal income tax rate of 63%. As a result, many of its best and brightest young workers have been fleeing for Britain and Germany and other EU countries with lower taxes. Sooner or later, Denmark will almost certainly have to lower its rates too.

Posted by Mixner | February 9, 2008 11:14 PM

Isn't Ireland's economic miracle over the last 10 years pretty convincing evidence that lowering marginal tax rates (including corporate taxes) increases economic growth, GDP, etc. etc. etc.?

Posted by BladeDoc | February 11, 2008 12:37 AM