Paul Krugman’s latest gets my vote for his most incoherent column ever. As I understand his argument, it goes like this:
- Computers are good at routine tasks.
- Therefore the rewards to performing routine tasks are falling. This is true at all skill levels.
- Therefore education does not always make people more productive. It makes people more productive only when it trains them to do tasks that are not better done by computers.
- Therefore we need stronger labor unions and universal health care.
Say what?. The basic thesis — that there’s no point in learning to do something difficult if a computer can do it better, and that this is significantly affecting the returns to certain kinds of education — is an interesting one. The moral, of course, is that you can’t imitate your way to prosperity. If we want to be rich, we have to innovate.
So to encourage innovation, you want to strengthen the unions? To encourage innovation, you want to reduce the relative reward to innovation, by insuring that everyone gets the same health care regardless of their social contributions?
Now, you might suppose that Krugman was thinking something along the following lines: Large swaths of American workers are being rendered unproductive by computers. Somehow or another, we have to support those people even though they’re not producing much. Unions and universal health care will keep them afloat.
But that can’t be what Krugman was thinking. I’m sure of this, because I happen to know that Krugman has a Ph.D. in economics. Therefore he must surely be aware that you can’t divorce incomes from productivity. Sure, you can redistribute, but you can’t redistribute more than what gets produced. If the problem is that our old skills are no longer productive, then our incomes must fall unless and until we acquire different — and less computer-replaceable — skills.
There are certainly good arguments for universal health care. (There are also good arguments against it.) But the “argument” that it’s somehow going to make people more innovative escapes me. And as for the idea that unions will somehow encourage workers to be more creative — I’d love to see either the theory or the evidence behind that one. Or at least a hint of what, if anything, was in Krugman’s head when he wrote this.
I saw his graphs (actually, Autor’s graphs : see http://krugman.blogs.nytimes.com/2011/03/06/autor-autor/), and I wondered – how much of the effect is really due to automation of highly-skilled jobs, and how much is due to increased international competition for highly-skilled jobs?
It’s important to know the answer. If college-degree-level jobs are disappearing because of automation, perhaps he’s right, and we should abandon traditional college education for some ‘soft-touch’ training.
However, if college-level jobs are suffering higher competition from overseas, surely the correct response is to strengthen and fix the US higher education sector.
I imagine it would disappoint some if the USA downgraded its higher education sector, only to find, in 30 years time, that all those technical skilled jobs are being done not by computers, but by humans overseas.
See also: http://blogs.the-american-interest.com/wrm/2011/03/07/paul-krugman-gets-it-half-right/
Amen
Krugman makes exactly that point in one of his excellent pop-econ books. (Recommended to all, by the way. They are Krugman at his best.)
It’s not a coherent argument – it’s Krugman the shill, not Krugman the economist.
Aren’t you setting up a straw man just so that you can have fun knocking it down?
It is pretty clear that Krugman is not going from #3 to #4. He clearly says “So if we want a society of broadly shared prosperity, education isn’t the answer — we’ll have to go about building that society directly.” The link between innovation and trade unions is artificially introduced only in your interpretation of the article.
Again, there is no proposal for divorcing incomes from productivity or “redistributing more than what is produced” in the original article. All I see there is a suggestion that, if education is not going to provide a path to basic health care and minimum prosperity levels even in the future where technology and computers are going to skew productivity and incomes even more than the present, then unions and universal health care are good policies to back. Whether the total productivity is expected to be higher or lower than the present is not even being gone into here, I think.
I can’t seem to summon up a defence for this one. He seems to answering the question of inequality due to superstar wages with an argument about education.
The problem he is pointing out – jobs carried out by well educated people being replaced by computers, is interesting. The answer has got to be to train those minds so freed up to do something with the output from computers to add even more value.
From the NYTimes article: “studios examined six million documents at a cost of more than $2.2 million, much of it to pay for a platoon of lawyers and paralegals who worked for months at high hourly rates.” Who, apart from lawyers, would consider this a good use of resources?
At the moment, computers are hopeless at some tasks easily done by humans. It will be interesting to see what happens when computers get better than humans at even these tasks.
You are an old-school economist. Maybe ypu are going to be replaced by a machine.
.
What if the machines allow us to live in a word where the individual produtcivity is not that important? maybe they can produce for us.
I think you’re right that his column is incoherent, but I don’t think he is arguing the points in your last paragraph. It’s a simple call for redistribution, nothing more.
Well what are people supposed to do when computers can do most of our work for us? Sounds like a great thing honestly for most the world as a whole, but it also seems like competition for the remaining jobs would skyrocket…prices would hopefully fall though to represent our gained productivity, so that would be good.
To be fair Krugman could be saying that when more productive computers displace less productive workers the total produced rises and we can alot the amount the workers *would have produced* to their benefit.
This will still disincent innovation but perhaps not entirely.
Not that I buy the argument, which amounts to ignoring the value of “creative destruction.”
I don’t really want to defend Paul Krugman, but I do believe there is a fallacy in Landsburg’s argument.
“Therefore he must surely be aware that you can’t divorce incomes from productivity. Sure, you can redistribute, but you can’t redistribute more than what gets produced. If the problem is that our old skills are no longer productive, then our incomes must fall unless and until we acquire different — and less computer-replaceable — skills.”
Umm… even if some individual’s relative productivity is going down (and it isn’t really), overall productivity is going up. The person who had a job is actually getting replaced by a computer that presumably does his job faster and cheaper. Consider the extreme case of a scenario where computers are basically able to do everything, making our own relative productivity zero. Does that mean we all starve to death, even if we can have computers capable of creating limitless supplies of food?
I’d say a better argument might be that people want some sort of obsolescence insurance, perhaps in the form of job retraining funds.
I have heard one argument for universal health care encouraging innovation, which is that currently people who might be off inventing things or founding start-ups are sticking with their day jobs in order to not lose their insurance.
Ack, I meant a better argument for Krugman might be that people want obsolescence insurance.
Landsburg’s argument that we need to encourage innovation as opposed to discourage it is not affected by whether total productivity decreases or increases.
What math_geek said … you can certainly redistribute from productive (semi-)robots to (relatively) unskilled workers. There are arguments for why you *shouldn’t* (robots r ppl too!), but it’s not some kind of economic impossibility, and certainly not one manifest simply by virtue of having an econ PhD, as Steve_Landsburg tried to imply.
What good is productivity if you can’t access the output? Think about it.
You ask for “a hint of what, if anything, was in Krugman’s head when he wrote this.” Here’s my guess: modern democratic societies are built upon the existence of a wide variety of occupations that pay well enough to permit a middle class existence. Computers are replacing the intellectual workers today just as machines replaced skilled craftsmen in the industrial revolution. The prospect is a new society with vast amounts of wealth concentrated in the hands of a tiny minority, and the mass of society reduced to unskilled manual labor and a subsistence existence. The alternative says Krugman, is more power and a larger share of the wealth to those who aren’t capitalists.
Prof. Landsburg,
This post is assuming you are not being sarcastic here (not a a high confidence assumption on my part)…
Krugman’s first point is not that people are becoming unproductive. Its that computers have reached a price/performance point to eliminate many currently high paying jobs. Krugman feels that this will continue to enrich the elite few only. Like all leftists this is his BIGGEST/ONLY fear. He doesn’t touch the subject of innovation. That was just you.
Continuing, Krugman is clearly trying to divorce incomes from productivity — Phd in Economics be damned! That is his mission since joining The Times. Yes, total national income must be tied to total national productivity. But to Krugman the relationship at the individual level is much less important. Krugman is developing very socialist views with regards to 1.) how much of that total national income should be distributed via gov’t fiat. and 2.) how to try to keep productivity/income high despite #1.
Mike H writes: if college-level jobs are suffering higher competition from overseas, surely the correct response is to strengthen and fix the US higher education sector.
If we invent a machine that can do legal work much cheaper than a lawyer, surely (ceteris paribus) it makes more sense to have less people going into law school. Some of those can produce electricity and the other inputs required by the machine, others are free to pursuit something else.
Likewise, if your trading partner can do legal work much cheaper than your lawyers, it makes more sense to have less people going into law school. And have some of them produce whatever is your trading partners like.
Thinking of trade as competition is not the right way to go about it: think of your trading partners as machines you can choose to use, and so collaborate to increase your combined output. Steven Landsburg explained this very well in Armchair Economist.
“It is pretty clear that Krugman is not going from #3 to #4.”-Krishna
I think it is clear that’s exactly what Krugman does. Here’s the full text from the two paragraphs in question.
“But there are things education can’t do. In particular, the notion that putting more kids through college can restore the middle-class society we used to have is wishful thinking. It’s no longer true that having a college degree guarantees that you’ll get a good job, and it’s becoming less true with each passing decade.
So if we want a society of broadly shared prosperity, education isn’t the answer — we’ll have to go about building that society directly. We need to restore the bargaining power that labor has lost over the last 30 years, so that ordinary workers as well as superstars have the power to bargain for good wages. We need to guarantee the essentials, above all health care, to every citizen.”
Mike H:
“However, if college-level jobs are suffering higher competition from overseas, surely the correct response is to strengthen and fix the US higher education sector.”
I noticed that you use the terms college-level and highly skilled interchangeably.
For sure there are some college degrees that also serve as vocational training. E.g. Computer Science, Engineering, etc.
There are also vocations which required highly skilled workers, but no college. E.g. Carpenters, Electricians, etc.
If by higher education sector, you mean providing education to trades (computer programmers, civil engineers, plumbers, carpenters, electricians, etc.) so that these trades can perform their crafts more efficiently, then I’m with you. However, these trades seem to be doing pretty well currently so my assumption is that this education is happening now.
If by higher education sector, you mean colleges and universities at large, I don’t see how this equates to skilled workers. Is someone with a History or Business degree working as an insurance claims adjuster, help desk supervisor, etc. a highly skilled individual because of their college education?
Paul Krugman even has this bias:
“Yes, we need to fix American education. In particular, the inequalities Americans face at the starting line — bright children from poor families are less likely to finish college than much less able children of the affluent — aren’t just an outrage; they represent a huge waste of the nation’s human potential.”
Perhaps instead of focusing on every child going to college and making sure we have plenty of college-level American jobs, we should focus on identifying a child’s talent and provide an education that better accentuates the child’s talent. Whether this is Autobody Painter, Electrician, Civil Engineer or Economics Professor.
Hey Steve, do you play requests?
Cuz I also would like to see Landsburg’s response to math_geek.
More generally, Krugman seems to be expressing a concern I share: that a libertarian world is a world with a small group of wealthy people and a vast sea of poor people — or, at least, “poor” by the standards of the US middle class. I suspect that many laws — progressive taxation, unionization, social safety nets, unemployment comp., civil rights, etc. — may reduce productivity. And I suspect that many of these laws also redistribute wealth in a manner that enables a middle class to exist (and to populate Landsburg’s university).
I’d be curious to know whether Landsburg believes that the US middle class would erode in the absence of laws designed to transfer wealth to them. I’d be curious to know if Landsburg can identify relatively efficient means for achieving the goal of maintaining a middle class. And I’d be curious to know whether Landsburg regards maintaining a US middle class as a worthy goal.
(While I’m at it, I’d like to hear Landsburg’s review of Anatham, but that’s a whole ‘nuther discussion….)
Rowan,
“I have heard one argument for universal health care encouraging innovation, which is that currently people who might be off inventing things or founding start-ups are sticking with their day jobs in order to not lose their insurance.”
I have a story to share with you. Back in the mid 90’s I was doing some work in China. At that time I noticed that working couples tended to have one spouse working for the state and the other spouse worked for a western firm. When I asked about this I learned that the reason was that the western firms paid very well, but you couldn’t get housing unless you worked for the state which tended to pay low wages. Hence bifurcated couples. It seemed really strange to me at the time.
A few years later, it began to dawn on me that this arrangement really wasn’t so different than what was happening in the states. The vast majority of new businesses fail because of under capitalization, but the ones that have the best chance of making it were the ones where one spouse started the business, and the other worked for a company with medical benefits.
Anyway, it’s just an anecdote, but I’m inclined to agree with your point.
On another note, back in 2005 Danial Pink wrote a book called “A Whole New Mind.” I’m not sure I recommend it per se because he makes a lot of assertions without data, but it will stimulate some thinking. Pink’s premise is basically that Abundance, Asia, and Automation are all changing the skills that will be required in the marketplace of the future. I found his diagnosis thought provoking, and his prescription wanting. He and I spoke about it at length. In reading the Krugman piece and supporting blog notes, I realize I was concerned about the fate of the middle class, much as Krugman is, at the time Pink and I spoke. Pink was not concerned, but let’s just say the example Pink used to make his case that new solid middle class jobs (his example was web designer) has not really panned out as a middle class wage earning position (much of it has been automated – ironically).
I think it’s incoherent but not in the sense you identify. He is saying “if you want a broadly shared prosperity”, stop kidding yourself that education is going to deliver that. You’re going to have to do it by fiat and with coercion. That’s actually relatively coherent and I agree with him that we should dispel the myth that changes in education result in changes in prosperity (I think the relationship runs in reverse just as much).
The incoherence is that, yes, fiat and coercion can cause something to be “broadly shared” but that something would not be “prosperity”.
Rowan:
I have heard one argument for universal health care encouraging innovation, which is that currently people who might be off inventing things or founding start-ups are sticking with their day jobs in order to not lose their insurance.
I think this is a huge problem, and a huge disincentive to innovate. But it’s not caused so much by the lack of universal health care as by the bizarre way we’ve tied health insurance to employment status.
Steve:
I don’t understand your comment about the bizarre way we’ve tied health insurance to employment status.
It seems that until 2008 American Citizens allowed market forces to determine the health care system for working age Americans. Is the “we” in “we’ve” market forces?
It is hard to say that it’s the government or citizens are the “we” since government and public were pretty much been on the sidelines.
Are you saying it’s bizarre that American citizens let market forces determine the health care system for working age adults?
@Ricardo Cruz
If we invent a machine that can do legal work much cheaper than a lawyer, surely (ceteris paribus) it makes more sense to have less people going into law school
Yes. But that’s not the point of my question. Autor shows a nice graph showing that skilled salaries are now going down. Krugman alleges that this is because skilled jobs are being automated. I merely ask “is that really the cause? Or is it more to do with overseas competition?”
The correct policy response depends greatly on the cause, no?
@Will A : guilty as charged. However, I stick by my main point, explained above.
@DividedLine : My goodness – that’s exactly what my wife and I are doing!!
It’s kind of scary when a well respected Economist wears his heart on his sleeve and writes this sort of populist nonsense.
How are strong trade unions and universal health care supposed to stop ‘hollowing out’?
More fundamentally, where in Economics is there a law saying there has to be a well paid middle class by virtue of a prior investment in college degrees?
This would be like saying ‘we Saudi Arabians are facing a crisis. Young people graduating with Sharia degrees no longer have well paid jobs as Imams. Clearly we need to intervene directly in the Economy by supporting the camel jockey’s union and introducing universal burqa cover for men as well as goats.’
The more I read Krugman, the more I am alarmed by his Marxist instincts. In my opinion the greatest failure of modern economists is allowing other, lesser professions to define Marxism as the moral economics and Capitalism as the immoral economics. Can we really be surprised when a new and ignorant generation takes up the call to “dream the impossible dream” once more? Isn’t it better to tilt at windmills and be badly beaten than to sell your soul to the Devil?
Will A: Our health insurance is tied to our employment status largely because the tax code rewards employer-paid health insurance. So “we” means our policy makers.
Mike H writes The correct policy response depends greatly on the cause, no?
I am pretty much an amateur, but I don’t see why it does. In one case, you have machines that, given some input, will produce what you want. In another case, you have people that, given some input, will produce what you want.
The cheaper computers are, the less sense it makes to deploy workers to competing labour. The same thing applies to trade.
@Ricardo I agree with the general principal, but not with the specifics. Humans in other countries are just humans. If they are taking our jobs now, we can compete with them. It may be best to try to do so.
Machines are machines, and are continually being re-engineered. If they are taking our jobs now, it’s likely we have no chance of competing with them. It would therefore be best to specialise on something else.
Therefore, the correct policy response depends on what is actually causing the drop in skilled salaries.
Mike H: if you want to “protect” those jobs, sure, in one case, you want to self-impose an embargo, while in the other case, you’ll want to shut down factories.
“A libertarian world is a world with a small group of wealthy people and a vast sea of poor people — or, at least, “poor” by the standards of the US middle class”. That sounds like a pretty good description of the world we live in now.
With regards to the difference between foreign competition and computer competition. My first thought was it makes no difference, the response should be the same. The answer is to educate to do what the competition cannot, whether it is foreign workers or computers. It is true that what you specifically teach may be different in each case.
People the world over are similar, so the capabilities are similar also. So you may choose to do what the competition chooses not to do. India can do lawyering, and the USA can do something else, engineering for example, or movie making. Ultimately, the Indians will be as wealthy as the USA, and everyone is richer than they are now.
On a wider view, the foreign competition adds to the total welfare partly because the foreign worker profits from his income. A loss for one is a gain for the other. What if we view comuters as the competition? The computer has no capacity to profit from its productivity, so where does the benefit go? It goes in providing human workers with better materials to work with. No-one talks of the power loom profiting from its output.
In the short term, the profit goes to the owner of the machine, be it power loom or computer. I think this will lead to a less equal society, with a very wealthy top layer and an expanded subsistence class – as happened in Britian during the industrial revolution. The mill owners were fabulously rich, and the workers were able to only scrape a living. For many of these workers life was probably better as a peasant, cropping strips of land for subsistence.
But things got better, how much through the invivible hand and how much through regulation is fuel for many an argument. So is this what will happen in the current computer revolution? I see no reason why not, except for one thing. It is possible that the computer will be better at everything humans can do, and it will not be possible to train to do what the competition cannot or will not. If this point is reached, then economics (for humans) is dead, and we will be dependent of the charity of the machines.
Computing consolidates wealth into the hands of those that own the machines. If your job has been replaced by a computer I suppose you are out of luck; but what if you are still part of the process?
If you are paid 100K by a firm to write stock trading algorithms that bring in millions per annum, shouldn’t you unionize to get your fair share – why should the profits of digitization go to the few? Perhaps this is what Krugman is driving at.
I’m confused as to how Krugman thinks education will be less valuable. If computers are replacing alot of jobs that once required a college degree, won’t people just study tech-related fields or things that computers/robots can’t do? Also, a college education helps you think critically and solve problems. I spent hours computing integrals in calculus, but I don’t think that’s because my math prof hadn’t heard of mathematica.
Another thing, the light bulb ruined the candle industry, and countless chandlers were laid off. I doubt anyone would say that it was bad for the economy overall. New jobs will open up for the skilled workers
I share this view. But I’d also note that insurers found (and find) it profitable to offer group plans to people who have characteristics that are unlikely to correlate with risk, or negatively correlate with risk. Employment status has proven to be one such characteristic. Basically, insurance companies bypass the process of evaluating each person’s risk and let the company’s HR department do it for them. And the insurer can be doubly confident that the level of risk in the pool is not disproportionately large assuming that the insurer can induce a large percentage of the employees to buy insurance (thereby avoiding adverse selection.)
Thus, even in the absence of goofy tax policies, I’d expect that employers, employees and insurers will find it in their mutual advantage to tie insurance to employment status. And I’d expect that insurers would charge higher rates to people who are not in such nice risk-neutral or risk-negative pools. A situation not so different than we see today.
Presumably there’s empirical evidence on this question somewhere? Surely not every nation makes employer-provided health insurance tax-free….
I was wondering where this ‘hollowing out’ business comes from. My memory is that it was linked to the notion of ‘elite globalisation’- i.e. big corporations selfishly outsourcing stuff rather than patriotically providing jobs for middle class people in their country of origin.
In other words, its a bit of demagoguery, like ‘sweated labor’ but aimed at middle class people.
Opposition to the Governor of Wisconsin appears to being orchestrated with ‘hollowing out’ being used as a catch phrase.
A google search brings up this comparison of Japan and Wisconsin from 2001 http://docs.google.com/viewer?a=v&q=cache:J3KMmwOkHzsJ:citeseerx.ist.psu.edu/viewdoc/download%3Fdoi%3D10.1.1.11.5062%26rep%3Drep1%26type%3Dpdf+hollowing+out&hl=en&gl=uk&pid=bl&srcid=ADGEESgqdLdntzIw5gU05fBTnTTXgmdvfrh9vrTMuApfQVL7BvcnXSBDBmlg5Ys7uCDas0pQ_hKVcktiXYujNjFm7aDRSUBwr1vg10DNcTPjBD_0TNfQxmn6E99SvIj84HKFAwHiRXYL&sig=AHIEtbQpo9Fv1oeG3JiQwoC-QxKPPKwxzg
Does anyone see any merit whatsoever to this sort of argument? I ask for information merely.
@ nobody.really:
“Thus, even in the absence of goofy tax policies, I’d expect that employers, employees and insurers will find it in their mutual advantage to tie insurance to employment status. And I’d expect that insurers would charge higher rates to people who are not in such nice risk-neutral or risk-negative pools. A situation not so different than we see today.”
What you say makes perfect sense to me. If you are saying that market forces would produce this system outside of our tax code, do you feel that the tax code is the reason for our system?
Or do you think that it is possible that our tax code was used to reflect the fact that market forces had already created the system? I.e. policy makers wanted to give tax breaks to the majority of their constituents who where already in a system that favored employment.
Health care benefits were originally introduced by employers as a way to weaken unions. After World War II, the UAW was planning to follow some other unions and offer union members health insurance as an incentive to join the union. Seeking to weaken the UAW’s drawing power, General Motors offered health insurance to employees in 1949. Thus, I would dispute that “market forces” created the tie between employment status and health insurance. It was originally created out of the conflict between employers and unions, and shortly thereafter codified in the tax code to support employers’ positions, as Steve wrote.
Al V: I do not believe your history is accurate. Employer-provided health care originated as a dodge around the WWII wage controls. Employers weren’t allowed to give raises, so they gave health care instead.
@Harold “The computer has no capacity to profit from its productivity, so where does the benefit go?”
I’m reminded of a novel I read a while back. In one scene, the protagonist is trying to use his phone to unlock a door. The phone informs him that he can’t crack the lock, but the lock might be happy to accept a bribe. So he transfers 200 bucks to the lock, and into the house he goes….
Later it’s discovered that the machines are pooling whatever money they get to fix up unwanted machines (feral robot coffeemakers etc), but that’s a very minor sub-plot.
Let’s put the argument in IQ terms. Computers are very good at those skills that constitute the measured IQ: memory, pattern matching, extrapolation.
I don’t know how much an increase of computing capacity is worth in terms of IQ points, but computing capacity is growing exponentially by Moore’s law, and soon or later there will be IQ 100 computers. That means than 50% of humans could be replaced by a robot.
Not much later, there would be computers with 120, 140 and higher IQ, meaning that most high-level jobs would also be replaced.
Education would not solve the problem, since education cannot increment IQ.
Some few people would inherit and own robots, and be fabulously rich, the others, not owning robots, and without the possibility to earn enough to buy one, would starve.
Al V. and Steve:
Thanks your explanations. Either of these or some combination of the 2 make more sense as a genesis of our current health care system.
Being that I’m an obvious economic noobie and being that I’m giving up blogs for lent (starting 2 days late), can either of you point me to decent book giving an overview of what market forces are. I’m kind of surprised that a conflict between employers and employees aren’t a market force.
I plan on reading the Armchair Economist, but it would be good to have a couple of other books to read during my blog fast.
The first 3 parts are true and good, the 4th is idiotic. The answer is training people to do things that are less routine! Train people to make new algorythms, not just use them.
According to the Economic Policy Institute, from 1989 to 2010 productivity increased 62.5% while real hourly compensation increased about 20%. This suggests to me that 1) there’s little evidence that workers are no longer productive, 2) incomes are already divorced from productivity, and 3) we’ve got a lot of room for redistribution before we bump our heads against the productivity ceiling.
To rephrase my earlier question, which would you prefer:
– A world without a US middle class?
– A world with a crazy-quilt of populist economic regulations designed to create but conceal subsidies to the US middle class?
– A world in which the bulk of the US citizens surrender their populist economic regulations in return for a forthright wealth transfer from the US rich to other US social classes?
I have difficulty evaluating criticisms of populist economic regulation without knowing what alternative world we’re looking at.