Having recently bashed Paul Krugman, and in the full expectation that I’ll have occasion to bash him again, let me interject that Krugman is not just a first rate economist; he is also, when he wants to be, a superb economic communicator, with a long paper trail to prove it.
Take, for example his essay on the widespread failure of intellectuals to grasp Ricardo‘s theory of comparative advantage (the basis of the case for free trade). Instead of simply bemoaning the problem like the rest of us, Krugman makes a valiant and useful attempt to identify its root causes.
He starts with an analogy I’m also fond of (I’m not sure which of us has been using it longer): The theory of comparative advantage is like the theory of evolution by natural selection—to those who understand it, it is simple and compelling; yet non-experts can find it remarkably difficult to grasp.
In The Big Questions, I argue that this analogy ultimately breaks down: The theory of evolution is compelling largely because of the evidence that supports it, while Ricardo’s theory is compelling largely because of the logic that supports it. It’s not too surprising that a first-rate physicst or literary critic could be unfamiliar with a body of evidence, but it’s a little more unsettling when that same physicist or literary critic can’t follow a simple chain of logic.
Krugman points out, though, that the chain of logic might be less simple than economists realize, resting as it does on several other points that we’ve learned to take for granted, and he makes a useful stab at enumerating these points and pondering how we can communicate them better. He observes also that while economists understand the importance of reasoning from models, they often fail to realize that not everybody thinks this way. Would-be communicators, take heed.
Sometimes I wish Krugman would take better heed himself. My main problem with so many of his New York Times columns lately is that they don’t seem to proceed from anything I can recognize as model-based reasoning. (Well, that plus what looks to me like a pigheaded obtuseness about the perils of centralized power.) But when he was good, he was very very good. Don’t believe me? Check out his web archive. Start here, say, or here or here. There’s stuff there I disagree with (especially some of the macro stuff) but mostly there’s excellence, and more than enough to make the world a smarter and a richer place.
Very gracious, Steve. Cheap shots don’t befit an intellect of your stature.
Did Krugman become oblivious to the lessons of economics (your words) after he started writing at the NYT, or is he still mindful of those lessons when it comes to free trade? does he not suffer from any cognitive dissonance after writing both “in praise of cheap labor” and in praise of the German model?
The theory of evolution is compelling largely because of the evidence that supports it
I beg to disagree: there is compelling evidence for evolution (i.e. against creationism), and against Lamarckian evolution. Also, there is overwhelming evidence for random mutation (RM), and of course for natural selection (NS) sensu stricto (which we might consider tautologous, i.e. necessarily true). But the theory that evolution proceeds exclusively by RM and NS is compelling because of Occam’s razor, rather than direct evidence to support it.
[…] the theory of evolution by natural selection — to those who understand it, it is simple and compelling; yet non-experts can find it remarkably difficult to grasp.
Perhaps you are right, but I would claim that what people find difficult is working out the _implications_ of RM and NS. For instance, how many people understand that eugenics (i.e. artificial selection) is not a logical consequence of natural selection? Not even Galton, a brilliant man and Darwin’s cousin, could figure that out! (As I hinted before, I think that there is a misunderstanding of Darwin also in The Armchair Economist.)
Persistent errors are an interesting subject, but I’ve never really figured out a good way of dealing with them (suppressed rage doesn’t seem to work very well). It’s not just amateurs who make the errors either– here’s a few physics examples that I’ve seen among professionals– “F=ma is the definition of force”, “Two-slit wave function interference involves two electrons”, “The electrical energy that comes out of a socket is carried by electrons”. No, no, and no.
I was just wondering if Prof. Krugman would be as gracious to his critics as you seem to be.
[I]t’s a little more unsettling when that same physicist or literary critic can’t follow a simple chain of logic.
Unsettling, but sadly not surprising.
“My main problem with so many of his New York Times columns lately is that they don’t seem to proceed from anything I can recognize as model-based reasoning.”
Krugman often presents reasonably simple models on his blog when he wants to elaborate on points made in his columns. Perhaps the problem is that you have no respect for those particular models? Fair enough, but it’s up to you to make the case against them. I notice that the three articles you link to, approvingly, all relate to international trade. He has also written articles explaining why he still likes the IS-LM model as well as “the World’s Smallest Macromodel” which he learned from Robert Hall. He laments the fact that many freshwater economists simply don’t seem to know this stuff. For my part, when I read Krugman’s columns I often see in my mind’s eye diagrams which I studied decades ago as an undergraduate. Maybe they aren’t your kind of models but they are models nonetheless.
MattF:
“The electrical energy that comes out of a socket is carried by electrons”. No, […]
Well, the phrasing is ambiguous, but if electrons do not move around (i.e. if the plug has very high input impedance … or is not plugged in), then no energy “comes out” of the socket, so I’d give it a pass.
Also, I am not sure why you won’t accept F = ma as a definition of force. But I grant that 2-slit interference does not involve 2 electrons; nor 2 photons, either.
I’ll admit I was feeling grumpy this morning, here’s my explanations, for what they’re worth–
The electrical energy that comes out of a socket is carried to you (at the speed of light) from the power plant by the electric (and magnetic) fields– you have to have (at least) finite conductivity in the wires to support traveling wave fields, but the forces on the electrons in the wires are small and generally dissipative. Variations in electric currents follow the variation the electric field.
The F=ma business is somewhat more controversial, but I regard force as undefined (a ‘push’ or a ‘pull’, if that helps), and F=ma is the equation of motion– given the forces, it’s the second-order differential equation that determines the motion– connecting kinematics with dynamics.
There’s a bit of an explanation for the Krugman NYT column that Landsburg bashes. See here:
http://krugman.blogs.nytimes.com/2009/11/13/its-the-stupidity-economy/
Basically Krugman’s policy preferences are (1) inflate the economy, (2) fiscal expansion, and (3) job subsidies and work-sharing, in that order. He thinks that (1) and (2) are politically infeasible so we’re left with (3) as the third-best alternative. I guess the question is whether (3) is a poor alternative that might still do some good, or whether it’s simply bad economics. I just wish he’d use his stature and his column to continue to push for (1) and (2) if he really thinks they’re better ideas.
But yes, Krugman has some really good essays out there, including the Slate ones. I really like his explanation of Japan’s currency crisis by analogy with a “recession” in a Washington babysitting co-op:
http://www.slate.com/id/1937/
Economists are always frustrated that others (even intellectuals) can’t accept the compelling logic of Ricardo’s theory of comparative advantage: Comparative Advantage. I find it ironic that Krugman would think comparison with natural selection is appropriate.
I am pretty sure that I do understand Ricardo’s argument, and I’m also pretty sure that I understand why most non-economists instinctively reject it. The essential logic of comparative advantage is that both trading parties do better if they specialize since they can thus each get more goods. The problem with this idea is that the logic of our evolution teaches us that the struggle for existence is a nearly zero sum game.
If the wolf and the coyote cooperate in the hunt [or trade rabbits for deer] then they may each get more game, but they also hasten the day when the supply of wolves and coyotes is larger than the available supply of game can feed. The wolf knows (at some instinctive level) that it’s better for his long term progeny for him to lose some game in order to deprive the coyote of the chance to compete with him.
Similarly, the trade between China and the US brought the US a great bundle of cheap stuff (TV’s etc) and built China from a backward, impoverished country into an economic superpower – good for both in the short run. It also turned China from a minor competitor to a formidable rival in the struggle for energy and, ultimately, existence. From the standpoint of strategic evolutionary competition, that tradeoff was a disaster for the US.
In practice, most economists seem to act as if Darwin had never existed. They don’t understand why most people are skeptical of comparative advantage because they aren’t taking into account that those people’s brains were wired by evolution. When we live in a world where the only way nations interact is through trade, or where the struggle for existence has been repealed, Ricardo will get pride of place. Until then, I’ll stick with Darwin.
Basically Krugman’s policy preferences are (1) inflate the economy, (2) fiscal expansion, and (3) job subsidies and work-sharing, in that order. He thinks that (1) and (2) are politically infeasible so we’re left with (3) as the third-best alternative.
Here is more evidence that top economists have an uncanny ability to hold on to logically incompatible ideas: Krugman’s preference for items (1) and (2) is blatantly at odds with his admiration for the German model.
the logic of our evolution teaches us that the struggle for existence is a nearly zero sum game.
Not at all: if a plant develops a more efficient way of converting sunlight to ATP, then nobody loses. OK, if the plant is in a tropical forest, then there will be competition for ground and sunlight, but those are second-order effects: they are no more intrinsic to Darwinian evolution than the competition between China and the USA for energy is intrinsic to Ricardian free trade.
As for cooperation between wolves and coyotes:
[a] wolves primarily compete with other wolves; competition with coyotes is a relatively minor concern for either species;
[b] some wolves started cooperating with humans some thousand years ago; they have done much better than other wolves, haven’t they?
But it is true that most economists do not get Darwin (so to speak). First, because there is nothing like random mutation in neoclassical models afaik. Second, and consequently, because neoclassical economics is about reaching an optimum, while evolution is open-ended: whenever an ecological optimum is reached, some mutation will come up which will upset that balance in a bout of creative destruction.
if a plant develops a more efficient way of converting sunlight to ATP, then nobody loses
No, for the reasons you state but inexplicably ignore. The competition for resources is the essence of Darwinian evolution.
And wolves and coyotes don’t peacefully coexist. When wolves take over an environment, they kill out the coyotes. We have so many coyotes today because people have killed out the wolves.
Ever since humans became the apex predator, the greatest threat to human life has been other humans (through war, disease, murder or starvation). If that’s the creative destruction you want, that’s what you get. I would prefer to use technology and population control to avoid them.
No, for the reasons you state but inexplicably ignore.
Could you be more specific?
The competition for resources is the essence of Darwinian evolution.
Last time I checked, the “essence” of Darwinism is random mutation and natural selection. Natural selection can come from competition, but it also comes from other sources of death or infertility.
And wolves and coyotes don’t peacefully coexist
I never said they do.
Ever since humans became the apex predator, the greatest threat to human life has been other humans (through war, disease, murder or starvation).
In disease, the killers are germs, not other humans. In starvation, no biological agent is the killer (except for Communist regimes, sometime).
I do not get what you are driving at in your last paragraph; nor do I see how it relates either to my previous comment, or to your claim that the “struggle for existence” is a nearly zero-sum game.
Snorri,
To address your points: The more efficient photosynthesizer and its progeny will tend to crowd out similar plants.
Darwin never heard of genes or mutations, but he did know about selective breeding of horses, dogs, and racing pigeons. His genius was to recognize that the struggle for existence, aka competition for resources, could act as a “natural” version of the selector.
Disease affecting humans germs are spread almost exclusively by other humans. The arrival of white men and their germs in the Americas set off the greatest genocide we know of, killing off perhaps 200,000,000 people.
My last paragraph is commentary, and just claims that any hopes of achieving something like universal prosperity depend far more on controlling population than on details of trade or economic organization.
C.I.P.: Your interpretation of Darwinism looks not so much wrong as garbled; but let’s leave that aside. The following is more to the point:
The arrival of white men and their germs in the Americas set off the greatest genocide we know of, killing off perhaps 200,000,000 people.
[1] Epidemics is not genocide.
[2] It’s hard to believe (to put it mildly) that a population of 200 millions could live in the Americas when very little of the land was cultivated, the Agricultural Revolution had not yet taken place, there were no fertilizers, and Norman Borlaug was not yet born.
WRT points [1] and [2], see also:
http://users.erols.com/mwhite28/warstat0.htm#America
[3] Now let’s forget all of the above. Let’s assume that there were 200 or even 300 million people in the Americas in 1492. There are now over 900 million people in the Americas, so the total population at least tripled.
Further, the prosperity of these 900 millions is much higher: Angus Maddison (The World Economy, 2001) reports life expectancy in 1820 of 39 years for the USA, 27 years for Brazil. Today, nearly every African country, and every country outside Africa, has life expectancy of more than 39 years.
My last paragraph is commentary, and just claims that any hopes of achieving something like universal prosperity depend far more on controlling population than on details of trade or economic organization.
Well, the Americas have much larger population AND much higher prosperity than they used to have, so your assertion is proven wrong by history; and not only in the Americas, but in every continent.
This real-life example also demonstrates the point that I was trying to make with the example of more efficient photosynthesis.
But I agree that population cannot continue to grow forever. Whether it needs to be controlled, is another matter.
The problem with the economic models is that they are models of equilibrium state. That’s fine for establishing basic principles, for providing a playground where you can learn an intuition for the rules; but it’s not necessarily a good way to model the real world. Starting point, sure; but how far beyond the starting point must one go for reasonable accuracy? I think that’s the catching point.
Consider weather: While we learn the basic physics (statics) in equilibrium models, the entire essence of weather is that it is *not* an equilibrium state: Wind is precisely what happens when there is no equilibrium in air pressure (on the kilometer scale). Any intuitions you develop from statics–from equilibrium pictures of mechanics–are effectively worthless when applied to weather. (Well, the basic principles of statics are correct, of course; but the pictures you see in the statics textbook bear not the slightest resemblance to planetary meteorology.)
What about for world economics? Equilibrium states are what one might have in long-term static periods, but probably aren’t terribly apt descriptions of anything in the modern world. So arguments based in equilibrium models aren’t going to be accepted at face value, even if the models are accurate descriptions of basic principles. The “secondary effects” are likely to be seen as dominating the picture (whether that’s actually true or not).
Is global economics more like statics or more like meteorology? I’m not in any position to say. But I strongly suspect that the feeling among those who argue economics from political and philosophical perspectives is that the meteorological analogy is the accurate one; and that’s why they feel justified in eschewing models in favor of political persuasiveness.
Steve (Harris) I don’t really think your analogy is all that water tight. The weather (especially wind) is a dynamic equilibrium.
Breaking it down: each particle in the air exherts a force (which includes direction) on all its neighbours, and we observe wind as the cumilitive effect of all those trillions of interactions.
Mathemtaticians use abstractions called vector fields to model things like this, and have come up with the (improbably named) hairy ball theorum, which basically says there must always be a source (say hurricane) of wind soemwhere on our planet.
There’s wind because force is a vector (it has magnitude and direction) and our planet is roughly spherical. So it exists in a dynamical (ever changing) equilibirum. Which seems a lot like an economy taken as a whole.
Tom, no, wind is definitely not in any kind of equilibrium: That is why it is always changing. (The Coriolis-derived trade winds are maybe an exception over a period of weeks.)
The hairy ball theorem states there is no vector field on a sphere which is everywhere non-zero (Steve Landsburg and I learned this in the same topology class from Prof. Lulavecis, over three decades ago). What this says, in particular, is that there is always some point where there is no wind; in fact, there is always some point from which the wind is coming and always some point to which the wind is flowing, and there is no wind at either that source point or that sink point. But this tells us nothing about equilibrium conditions, i.e., what the global vector field looks like in a state of non-change.
“Ever-changing equilibrium” is a self-contradictory phrase; the entire point of equilibrium is that things have settled down to a state of non-change. That’s what makes it simple to study statics and why that is that is the first subject in elementary physics; but dynamics is the other side of mechanics and must be studied next. My strong suspicion is that there is a rough analogy with economic models, and that the Ricardian model is akin to statics, not dynamics. Of course cross-field analogies are no more than rough metaphors.
“But the theory that evolution proceeds exclusively by RM and NS is compelling because of Occam’s razor, rather than direct evidence to support it.”
No, it’s compelling because it underpins the chief religious faith of our age amongst intellectuals, namely, reductionist materialism.
I think what I meant was more of a dyanmic equilibrium akin to thermodyanamic equilibrium – where the state of the system could be any one of many in a configuration space, which are exlored over time. The state of the system will change but the long term characterisics will be simmilar, statistically speaking (a bit like walking up a ‘down’ elevator).
Whereas we can’t say something about which configuration the system will go to next, we can have knowledge about the long term state of the system (once it has reached that equilibrium).
I admt the wind alanogy was a bit poor now, in retrospect.
Tom,
I think what you’re suggesting is something along the lines of an attractor: The Ricardian model is simplistic but represents a sort of ideal which, while not matching any real system, none the less marks a spot in the space of all models, which real systems tend not to stray far from.
Is that, in fact, true? Or do the “secondary effects” tend to keep real systems significantly far from the Ricardian model? I don’t know. My suspicion is that the real world has so many significant secondary effects, that the pressures pushing towards the Ricardian model amount to only a fairish fraction of all pressures, and sustained departure from the ideal is not at all unlikely. But that’s nothing more than a guess, and I have little to support it.
Steven,
I’ve just carefully read the Krugman essay you pointed at (it ate my whole morning). I recognize myself in his remarks about lefty intellectuals. I’m well-educated and reasonably bright, but with no training in economics. I’ve read popular arguments for and against free trade in many contexts, and always felt strongly opposed to free trade based on those readings.
The idea that the argument in its favor is as simple, elegant and right as the argument in favor of evolution is disturbing. I’d hate to be on the side of the creationists in any conversation.
After reading Krugman’s article though, I still do not understand Ricardo’s mysterious idea. I’d like to learn about it, but I’m a journalist and might need the baby talk version Krugman bemoans being reduced to when talking to editors.
Is there a good, clear, simple version of the free trade argument available? Or is this as clear as it gets?
Sierra: First, I’m sure that the argument for free trade really is as simple, elegant and right—and as thoroughly convincing to people who understand it—as the argument in favor of evolution. There are economists who question it, just as there are biologists who question evolution, but in both cases they constitute insignificant fractions of their professions. And just as evolution has won over both the theists and the atheists in the biological world, so free trade has won over both the lefties and the righties in the economics world.
But that’s an answer to a question you’re not asking. You’re asking: What is the argument? As a teacher and a writer, I’ve spent decades trying to figure out the best way to answer that question. You can find my latest attempt in Chapter 5 of The Big Questions, but I will summarize it here. (You can also find several excellent attempts on Paul Krugman’s web page.)
First, though: As Krugman has so eloquently pointed out, this argument rests on a lot of auxiliary assumptions that virtually all economists take for granted (and for excellent reasons), but that we sometimes forget are not obvious to people who are not economists. So, I hope I’m not taking too much for granted by stripping the argument down so much.
I hope this makes sense, though, because this really is the essence of the whole thing.
John sells blankets to Mary for $10 each. Now because of an opening to free trade, Chinese blankets become available for $2 each. This has three effects:
A) Mary saves $8 per blanket.
B) John has two choices: He can match the Chinese price, in which case free trade costs him $8 per blanket. Or he can get out of the blanket business and do something else (either work at McDonald’s or go back to school or be unemployed). Whatever he chooses can’t be worse than matching the Chinese price, because if it were worse, he wouldn’t choose it. So the WORST case scenario for John is that he matches the Chinese price and loses $8 per blanket.
C) Frieda, who was never willing to pay $10 for a blanket in the first place, buys a Chinese blanket for $2 and goes to bed warm tonight.
A) is an $8 gain to Mary; B) is *at most* an $8 loss to John; therefore the good from A) cancels the bad from B). At the same time, the good from C) is pure gravy. So the upside of free trade must outweigh the bad.
That’s the economic argument. It seems to me that there are also compelling *moral* arguments for free trade, but of course there’s no reason to assume that economists know more than anyone else about morality. So at least in this comment, I’ll stick to the economic case.
Steve (Landsburg): the Iowa Car Crop argument is also pretty convincing. I find that I need at least two distinct logical arguments to be convinced.
Sierra: being a non-economist myself, I found that I had to apply economic principles to what I read in the newspapers, for at least a couple of years, to really understand economic principles. The Economist used to be good for that: it has declined, but this blog more than makes up for it.
Steve and Sierra: important as the Ricardian principle is, I think that economic history makes an even stronger case for free trade. I guess what I am saying is: The theory of comparative advantage is compelling largely because of the evidence that supports it.
OK. So while free trade might be bad for some individuals (like John in your example), it’s going to benefit at least the wealthy country engaging in free trade. John might end up unemployed, but overall the economy benefits, so most of the former blanket purveyors will find other work in industries their country does have a “comparative advantage” in.
If that is the basic theory, I get it now, and did not before. Thank you.
I too am tempted to dive into the moral implications of this system where the “competitive advantage” a poorer country has over a wealthier one comes from cheaper labor and looser environmental regulations, but as you say that’s straying pretty far from economic principles.
it’s going to benefit at least the wealthy country engaging in free trade.
Sierra: Exactly the same argument applies in both countries. The conclusion of the argument is that wealth increases in both of the countries that are trading.
Steven,
One of the arguments I’ve often seen made in opposition to free trade is that while international trade often increases the amount of money in a less-developed nation, it can damage the quality of life for the people living there.
Here’s an example: you might have a small community that has subsisted for many years on small farming and handcrafts that are traded locally. The people are poor, but getting along OK.
Then a foreign-owned mine opens up, and offers higher wages than anyone is making farming in their backyard. A generation of workers goes to the mine, and in the process loses much of the farming and crafting skill their parents and grandparents used.
When the local resource is exhausted and the mine closes, the people are left impoverished and unskilled; a worse situation than if they had never traded outside their local region.
Overall, the wealth of the country has gone up, and the mine probably made a few people in that country rich. But the workers are worse off than before, and don’t have a clear path to accessing the overall improved wealth of their nation.
I can see how free trade makes more money for both countries, but I’m not convinced (yet) that more money always makes life better.