Paul Krugman has some words for job creators and other high skilled people: we don’t need you.
…textbook economics says that in a competitive economy, the contribution any individual (or for that matter any factor of production) makes to the economy at the margin is what that individual earns — period. What a worker contributes to GDP with an additional hour of work is that worker’s hourly wage, whether that hourly wage is $6 or $60,000 an hour. This in turn means that the effect on everyone else’s income if a worker chooses to work one hour less is precisely zero. If a hedge fund manager gets $60,000 an hour, and he works one hour less, he reduces GDP by $60,000 — but he also reduces his pay by $60,000, so the net effect on other peoples’ incomes is zip.
First let me just say that the extent to which what people earn is equal to their marginal product is greatly unappreciated, so before I disagree with Krugman I just want to pause and point out that this is truer than most people think. But it’s not true everywhere and always. Note that Krugman does not go so far to say that marginal product *does* in fact equal income, but that “textbook economics says that in a competitive economy…”, and that job creator praise is not “something that comes out of the free-market economic principles these people claim to believe in”, and that “Even if you believe that the top 1% or better yet the top 0.1% are actually earning the money they make…”.
He doesn’t actually say “people earn what they make”, nor does he say how good of an approximation to reality marginal product theory is. But in his economics textbook with Robin Wells, they are a bit more explicit, and do call the marginal product theory a pretty good approximation:
The main conclusion you should raw from this discussion is that marginal productivity theory is not a perfect description of how factor incomes are determined, but that it works pretty well. The deviations are important. But, by and large, in a modern economy with well-functioning labor markets, factors of production are paid the equilibrium value of the marginal product -the value of the marginal product of the last unit employed in the market as a whole.
This is a really important point, and I don’t disagree. But I think that many high skilled, high paid workers, and job creators in the U.S. can be an important deviation from this general rule. Many of these people work at moving the productivity frontier forward, and thus increasing the marginal productivity of other workers. After all, one of the important things that entrepreneurs do is find ways to increase the productivity of other workers so they can underprice their competitors. The process of creative destruction is not manna from heaven. I won’t pretend to know have all the answers about what drives this process, but entrepreneurs, job creators, and high skilled people are an important part of it.
Consider, for instance, that if we suddenly kicked out the top 10% of high IQ people (or 10% most productive people, or 10% most creative people, or whatever) in the U.S.. It strikes me as fairly likely that the total output of the remaining 90% would go down. Krugman seems to argue that this would not be the case. But even if you disagree with me in the short run, in the long-run the productivity increasing innovations these people would have made won’t show up, and the rest of us would have lower productivity as a result.
Now, instead of kicking out the top 10% of workers, just make them work less as a result of high income taxes. See my concern?
Lowered incentives of job creators and other innovators should be considered as one of the likely downsides to higher taxes.
Note that if you don’t think this is true, then what business do we have subsidizing higher education? If workers capture the entirety of their higher productivity, then I don’t see who gains by giving young people money to go to college rather than just cash.

23 comments
Comments feed for this article
Wednesday ~ November 23rd, 2011 at 12:27 am
Andy Harless
The baseline is that everyone earns their marginal product. How does reality differ from that baseline? For one thing, a lot of high-income people earn their high incomes by redistributing rents. These people are earning much more than their social marginal product and ought to be taxed heavily. Some other high-income people earn their high incomes by making other people more productive (and without fully capturing the impact of those increases in productivity). These people are earning much less than their marginal product and ought to be taxed lightly if at all. But it’s hard for the tax system to distinguish between these two groups. It’s not necessarily easy for the education system to distinguish between them either, but it’s comparatively easier. Thus high marginal taxes and education subsidies are both defensible.
Wednesday ~ November 23rd, 2011 at 9:00 am
Adam Ozimek
Andy, I agree with your broader point that some people redistribute rents and thus earn less than their marginal product, but I disagree that the education system is better at distinguishing between the two than the tax system. I think both are pretty poor at it, and I’m not sure in what way the education system even *tries* to do this.
Wednesday ~ November 23rd, 2011 at 9:39 am
Andy Harless
It’s why we subsidize science education more than law and business.
Wednesday ~ November 23rd, 2011 at 9:57 am
Adam Ozimek
Do we? In raw dollar amounts I’d bet we spend more on subsidized loans and grants for business and law majors than science majors. I want a pell grant or student loan does it matter which field I choose?
Wednesday ~ November 23rd, 2011 at 12:33 am
lfvoss
On the other hand, marginal rates can be so low that it decreases the incentive to work and earn more. In other words, “I’ve already got enough.”
And on the third hand is whether the top 10% IQ/creative/driven/hard working/whatever are driven by the amount of cash they have, which seems completely absurd. Especially at the top end, except as a means of keeping score with their golf foursome.
Wednesday ~ November 23rd, 2011 at 12:53 am
happyjuggler0
I was about to congratulate Karl for possibly his most astute and pertinent post ever (possibly excepting his posts regarding risk aversion/ insurance deficits amongst the global poor). Then I realized it wasn’t him…oh well.
The notion of a 50% chance of $1000 win/loss being asymmetrical, especially for the dirt poor, is still radically under-appreciated by most PhD economists, especially those who seem to view the economics world as a math problem involving homogenous agents , not a word problem involving heterogeneous agents….
Wednesday ~ November 23rd, 2011 at 1:21 am
Ben
Some questions:
Does taxation actually persuade top income earners from working more?
Isnt the marginal benefit of a dollar earned much lower for the upper income brackets?
Additionally, American productivity isnt in line with income, so doesn’t that show that there’s a flaw in assuming wages are determined by productivity?
Wednesday ~ November 23rd, 2011 at 2:06 am
Eli
Adam, agreed, but remember also Landsburg’s story about the rich man who can’t be taxed. Even if we grant that the super-rich are earning their marginal product, they are probably not consuming their marginal product. If the hedge fund manager works an hour less, someone is consuming $60,000 less, but most of that loss is borne by others.
Wednesday ~ November 23rd, 2011 at 9:02 am
Adam Ozimek
When I was writing this post I did think “Eli would be better at writing this one”. Maybe you should take a crack at it too.
Wednesday ~ November 23rd, 2011 at 9:56 am
Eli
Nonsense, it’s a good post; I just had that one comment to add.
Wednesday ~ November 23rd, 2011 at 2:53 am
Ryan hayes
Your hypothesis is that productivity is a dependent variable and positively correlated so just test it against the data; I don’t see the big deal and you’re probably right since productivity per employee has been going up at a pretty fast rate lately…
Wednesday ~ November 23rd, 2011 at 6:23 am
foosion
>>just make them work less as a result of high income taxes>>
There does not appear to be any persuasive evidence on this point. People may work less as a result of lower net compensation. People may work more as a result of wanting more net compensation.
>>what business do we have subsidizing higher education>>
It’s a good thing to live in a country with a more productive population. It makes people happier apart from any economic benefit.
Also, we appear to prefer in-kind payments rather than just cash transfers to the less well off. Food stamps as a random example.
Wednesday ~ November 23rd, 2011 at 7:48 am
Curt Doolittle
Adam,
This is one of the best posts of the year.
Are people of different productive value? Just go look at the data around the world. If you kicked out the upper 10% it DOES GO DOWN for everyone else. We have a number of examples, from Ireland to Lebanon. In some cases, it takes only a generation or two to restore it. But the institutions must recover as well.
BULLET POINTS:
- Higher IQ’s tend to work on much longer production cycles. i.e. the division of knowledge and labor is inter-temporal. – Hayekian triangles.
– Objective truth-telling is the one of the greatest predictors of GDP. (Kant strikes again) – Huntington.
– Education in the aristotelian (western) tradition or confucian tradition ‘floats all boats’ equally by about 20pts. Jury is out elsewhere. Literacy is more important than any other form of education since general knowledge available from reading can compensate in practice somewhat for lack of “G” – natural IQ.
– Individual IQ is highly correlated with lifetime income.
– Being a deca-millionaire is less correlated with IQ (it’s luck and hard work)
– Very high IQ people are less likely to become deca-millionaires. (they don’t need to because income is more easily available.)
- IQ’s over 105 are necessary to repair machines.
- IQ’s over 122 are necessary to design machines.
- IQ’s over 130 are necessary to competitively coordinate capital
- IQ’s over 140 are necessary to invent abstract ideas, and assimilate accumulated knowledge.
- The upper 20% determine everything meaningful in an economy – and GDP depends upon where the average IQ sits on the curve.
- The higher the average of the upper 20% the richer the economy.
- Property rights make it possible for IQ to be employed to increase productivity.
- The upper 1% consist of multiple categories of people – from small business owners who sell the accumulated efforts of 20 years of work, to lottery winners, and entertainers, to fund managers that are statistically lucky (since the data says almost none of them can repeat their successes, it’s luck to get there, and access to capital that keeps you there).
- Redistributing rents is a privileged position in society that requires good upbringing, very hard work, a great education, intelligence, and luck.
- The question is whether we should employ these people in redistributing rents, or whether we should do something useful with them.
- Job creation is a function of mobilizing opportunities in every social class — meaning roughly at every standard deviation of IQ within an economy.
- Communication of ideas across more than one standard deviation is subject to extremely high friction and across three is almost impossible.
Anecdotally: I made over 1M last year according to my tax return. But I didn’t see a dime of that money that was above my salary. It’s paper wealth. My spending is determined by my salary, not my asset manipulations. My spending is consumed largely by signaling costs that assist in getting access to that income stream. And the frequency with which I make that kind of money varies considerably. This year I’m only in the upper 2%. And undoubtably next year I’ll be in the upper 3%. And ten years from now I’ll only be in the upper 10% even if I’m lucky.
BTW, it was Krugman’s worst post in years.
Wednesday ~ November 23rd, 2011 at 10:01 am
Twundit (@Twundit)
What was the marginal productivity of the work done by Tim Berners Lee, and how did that compare to his income? Or how much does the world owe to CERN, where WWW was first created? The creation of the world wide web has been about the most valuable in recent times, but no person or company became rich or powerful from it.
Aren’t high incomes all about both willingness and ability to negotiate a high salary, next to having some skill in an area where enough money is at hand to pay that high salary?
How is it that bankers get such high salaries, other than the fact that they’re near to the place where (other people’s) money is being stored, passed around, or created by loans, combined with selling themselves well? Financial innovation in recent times has been a most economical destructive force. But it made many people and companies became rich and powerful, even too big to fail.
Wednesday ~ November 23rd, 2011 at 10:09 am
Becky Hargrove
As Eli pointed out, Landsburg did a great job of illustrating the fallacy of the intermediary in which production gain could ultimately be lost. I only wanted to add that the process can work both ways wherever irrational complexity comes between the ultimate product and the ultimate producer within the service private sector.
Wednesday ~ November 23rd, 2011 at 10:09 am
Twundit (@Twundit)
I’m wondering what a ‘job creator’ would be. I think a more apt word is simply ‘leader’. A good leader is somebody who helps others to become more productive, providing a vision and a sense of purpose, by convincing people of good things to do, and how to do them. One important property of good leaders is that they are largely self-motivated by their sense of purpose, or hunger for power, not really receiving an exorbitant income.
It may be that hedge-fund managers are largely income-motivated. But one could argue if they actually provide leadership, a vision, and a sense of purpose, or methods to improve the world we live in. Are they leaders, or parasites?
Wednesday ~ November 23rd, 2011 at 11:00 am
Lord
There are both the income effect and substitution effect and what evidence there is, is that the income effect is greater at the tax rates we have. I would incline to the belief that the income effect predominates heavily among real job creators while the substitution effect is significant only among rent extractors so higher than current rates would increase job creation, at least that is what the paper he refers to is saying.
Wednesday ~ November 23rd, 2011 at 11:30 am
Cross Marginal Product of Labor, Part II « Modeled Behavior
[...] responds I think that many high skilled, high paid workers, and job creators in the U.S. can be an important [...]
Wednesday ~ November 23rd, 2011 at 1:22 pm
Dustin Miller (@_DustinMiller)
From an old Jared Bernstein post(august 29th):
there’s no reason to expect people to respond to higher tax rates by working less. That is, they could just as easily decide to work harder to make up the loss in their after-tax income.
Microeconomics predicts two responses to higher taxes on the work effort of people. Response A is that they work less, because the “price,” or opportunity cost, of non-work, just went down…you lose less if, once your after-tax wage has gone down, you work less.
But the other response (B) is that you work more to make up for the lost income.* And there’s no reason, a priori, to think response A dominates response B. If anything, the literature, which tends to show small responses to tax changes, suggests the two responses offset each other.
Wednesday ~ November 23rd, 2011 at 9:00 pm
rturpin
It seems to me there is a stronger argument for Krugman’s point. Most of the work done by the 1% is bad. The vast majority of new businesses fail. Most new inventions are worthless. A new business direction usually is a mistake. Fortunately, the forward direction of capitalism does not depend entirely on how well its leaders do, but from the fact that capitalism implements an evolutionary algorithm. The many mistakes are kicked aside, while the rare success is allowed to run and built upon. Which raises a serious question about the importance of the quality of new business ideas and ventures. Does it matter how good they are? Or just that they provide enough variation?
Lamarckism doesn’t have much role to play in biology. Maybe it has more role in economics.
Sunday ~ November 27th, 2011 at 9:10 pm
SC Mike
rturpin
“It seems to me there is a stronger argument for Krugman’s point. Most of the work done by the 1% is bad. The vast majority of new businesses fail.” Does not that keep them from being in the 1% in most cases? That is, they make it to the 1% only if they succeed, no?
And in retrospect we can measure and find that a hedge fund manager makes $60K per hour. However, an additional hour may or may not result in an additional $60K, rather it may cost him/her much more than that.
Thursday ~ November 24th, 2011 at 7:07 pm
Johnnie Linn
Where is average product? The income, marginal product, of labor is about 70% of income. So, averaged across industries and occupations, the marginal product of labor is about 70% of its average product. If a worker “chooses” to work one hour less, that choice is part of an amalgam of employers bringing marginal product per dollar to balance accross all factors, hence the effect on other factors is not “zip”. If a hedge fund manager is being “paid” to manage funds (of others), the effect on the others is not zip. If he is managing his own funds, others are paid commissions on his trading. Again, effect on others is not zip.
Ben, hast hit it in regard to wages and productivity. Wage is marginal product and productivity is average product so they don’t behave exactly the same.
Thursday ~ February 16th, 2012 at 6:53 pm
August
Read Is War Necessary for Economic Growth by Vernon Ruttan. Major general purpose innovations do not come from the top 1%.