Bryan Caplan  

Return to Solipsism

PRINT
Workers Usurp the Means of Pro... Parents and Peer Pressure Revi...
In my "Solipsist's Guide to Comparative Advantage," I showed how, for all practical purposes, trade actually raises worker productivity.  Notice, though, that my example shows the productivity effects of trade if you and I have different absolute advantages: you're a better farmer, I'm a better steel-worker.  What about the case where you're better at both jobs?  Such as:

 

Wheat

Steel

You

10

5

Me

1

2

For convenience, assume again that the wheat:steel price ratio is 1:1.

Then on the solipsistic assumption that I'm merely your tool, trade revises your productivity as follows:

Wheat

Steel

You

10

10

On the opposite assumption that you're merely my tool, trade revises my productivity thusly:

Wheat

Steel

Me

2

2

Putting both perspectives together:  

 

Without Trade

With Trade

 

Wheat

Steel

Wheat

Steel

You

10

5

10

10

Me

1

2

2

2

Thanks to comparative advantage, trade allows you to effectively double your steel-making productivity, and allows me to effectively double my wheat-growing productivity.  And this is true despite the fact that I have absolutely nothing to teach you.


Comments and Sharing





COMMENTS (13 to date)
Ironman writes:

Brian wrote:

Thanks to comparative advantage, trade allows you to effectively double your steel-making productivity, and allows me to effectively double my wheat-growing productivity. And this is true despite the fact that I have absolutely nothing to teach you.

That's okay. By the rules of solipsism, you don't really exist, anyway.

Vladimir writes:

The problem with comparative advantage arguments of this sort is that they are clearly valid mathematically, but the underlying simple mathematical model hides lots of questionable assumptions.

Therefore, the best answer to these models that are simplified to the point of extreme to make a strong point is to take an extreme real-life example that makes a strong point in the other direction. So, here is my question: why didn't the majority of horses benefit from comparative advantage when motor transport was invented, but ended up as dog food and glue instead?

The answer, of course, is that when a new party appears that enjoys absolute advantage over you, the comparative advantage says absolutely nothing about your individual position in the new equilibrium. You may be richer, or you may be poorer -- perhaps even to the point of falling below subsistence, as happened with most horses several decades ago. Moreover, the comparative advantage argument says nothing about what proportion of people will win, and what lose in the new equilibrium, even if the total wins are necessarily greater than the total losses. One can easily construct examples where arbitrarily large majorities end up as losers.

Now of course, this is not by itself a decisive argument against free trade or anything else. But it is a decisive argument against propagandistic touting of comparative advantage that sneakily suggests that trade tautologically leaves everyone better off. Bryan's comparative advantage posts contain no arguments that couldn't also be applied to a horse a century ago, telling him that soon he would be subject to absolute advantage by motor transport, but no worries, comparative advantage will make all parties better off.

(There are many other sneaky assumptions in these simplistic comparative advantage stories that may or may not apply in any particular real-world case. But this problem is probably the simplest one to point out.)

RPLong writes:

Between this post and the last one, I have never seen so many objections raised to the Law of Comparative Advantage in all my life.

Reading Human Action, I wondered why Mises spent so much time discussing the ins and outs of Ricardo's law... But I guess I know why now.

It's obvious to me, but I guess it is less obvious if you come from a slightly different background than I.

Nick C. writes:

How can one party trade for goods that the other party isn't capable of producing? If in a unit of time I can produce 1 unit of wheat and 2 units of steel, then specialization only allows me to produce 4 units of steel in the same amount of time. You might offer me 10 units of wheat in exchange for 10 units of steel, but I don't have the productive capacity to fill the order without working 2.5 times as much as you. What if instead of sitting on your hands, you decide to produce 50 units of wheat? After we trade, you'll have 40 wheat and 10 steel, but I'll only have the 10 wheat to show for it and I'll still have to work another 0.5 time to produce my own original demand for 2 steel. I'm simply not capable of fulfilling your demand for steel once you start working as much as me.

Of course, if you value wheat as much as you value steel, then you wouldn't spend your free time producing that kind of surplus wheat; at some point, you'd probably spend the balance of your time producing your own steel to supplement what you buy from me.

So, going back to a unit time, I can only produce 4 steel, and I'm willing to trade 2 steel for 2 wheat. Since you're buying steel from me, you may decide to reallocate slightly more of your productive resources to wheat production, so you can produce 12 wheat and 4 steel. After the dust settles, you're left with 10 wheat and 6 steel. By trading with me, you have one more unit of steel than you would've had if you had produced it yourself. Similarly, I have one more unit of wheat than I would've had. By engaging in trade, either party is still better off than before, even though you're short of your ideal preference of 10 wheat and 10 steel.

I suppose what I'm asking is wouldn't there be some upper limit on the productivity gains from trade?

Nick C. writes:

Vladimir, maybe while the uncompetitive individual is hanging out in the pasture with the horses, he can contemplate why he doesn't have a comparative advantage in something. Unlike horses, most people have a range of talents that can be utilized for their advantage.

If he is squeezed out of one market, then the resultant productivity gains offer the opportunity to satisfy new demands. Perhaps his comparative advantage is art. With low wheat and steel productivity, the other players couldn't buy his art if they wanted to. As they become more productive, they can begin to consider buying more luxury goods. Thus, our individual can still benefit if he can successfully compete in this new market. If he still can't earn a living, then perhaps he should consider taking the time to improve his skillset so as to be able to compete.

Vladimir writes:

Nick,

Remember that after the invention of motor vehicles, horses still had comparative advantage in transport! Any creature that could be put to any imaginable productive use enjoys a comparative advantage in something. The question is, how much can you actually earn by exercising your best comparative advantage? The horse example illustrates the fact that once a party enjoying absolute advantage over you appears on the market, your earning power can fall, and in the most extreme case it can fall arbitrarily low - even though your loss is offset by someone else's greater gain.

The goal of this example is to show the disingenuous nature of the commonly made comparative advantage arguments, which typically make a strong underhanded implication that regardless of any absolute advantages, everyone has to be better off in the new equilibrium. (This falsity is often even claimed explicitly, although of course Bryan is too careful a thinker to commit such an error.) The number of people who will end up worse off can be anywhere from zero to an arbitrarily large majority, and the comparative advantage argument by itself implies nothing at all about this number - merely that the gains of the winners exceed the losses of the losers.

Note that I'm not arguing for or against any concrete policy in the present context. I am only pointing out that if you want to convince a person that they should support free trade, a valid and honest argument must say, "You'll be better off because of (1) comparative advantage, and (2) the following reason why you won't end up as a loser in the new equilibrium: [...]". Arguing only the former part and leaving the latter as an unjustified underhanded implication can be only fallacious or dishonest.

(Plus, as I said earlier, there are also many shaky assumptions that can make even the basic logic of comparative advantage inapplicable in real-world cases. But the problem I pointed out is striking because it is present even if we accept the simple comparative advantage story as an accurate model of reality without any reservations.)

agnostic writes:

"Unlike horses, most people have a range of talents that can be utilized for their advantage."

Not when they've spent their life following economists' advice to specialize so much -- those other potential talents have never and can no longer materialize (if there's a critical window like learning a language, too large upfront costs of effort & time like a 10,000-hour apprenticeship, etc.), or they have by now atrophied.

In either case, even if you wanted to switch, will you find an open niche or instead a lot of others who have spent their lives specializing in the tasks you're now interested in? Likely the latter. So you'll get shut out there too, or barely scrape by.

It's odd that libertarians haven't seen the connection between hyper-specialization and the rise of the welfare state. Specialization makes you fragile to negative shocks (creative destruction, changes in fashion, natural shocks, whatever). Generalization makes you robust to them -- unless they all happen at the same time, one shock will ding you a bit, but you still have so much to fall back on.

Thus, in a generalist society no one will feel sorry for you if you can't find work as a seasonal thresher -- you know how to do so many other things, just like everyone else around here, so why should we support you? And perceiving this total lack of sympathy with their pleas, such unfortunate people will not bother agitating to implement such policies.

In a specialized society, others will feel sorry for your bad luck because you have little or nothing to fall back on. And they don't either, so it's easy for them to realize "Hey, that could've been me" and truly sympathize with your pleas. So, citizens will agitate for and receive "social safety net" policies.

It follows that the more narrowly people specialize, the greater the social safety net programs will grow -- both in terms of what areas of life they cover (first education, then old age, then health care, then...) and how big of a cushion they provide in a given area (first a little help with basic nutrition and antibiotics, then operations, then life-long supplies of drugs to combat your obesity and diabetes, then yoga and massages, then...).

This is all straightforward, so the libertarians' championing of both hyper-specialization and a limited or non-existent welfare state is another case of education making you denser. The liberal and socialist economists are at least consistent -- let's all specialize a lot, but let's also have a big welfare state to protect against the inevitable fragility that results.

Nick C. writes:

Vladimir,

I'm not sure that horses even have a comparative advantage over motor vehicles in transport. Cars are specifically designed to excel at one thing: transportation. A car can't easily be used to herd cattle, and we'll never see a race car win the Kentucky Derby, but they do seem to have an advantage over horses in their ability to get people and their stuff from here to there.

I see your point, though, and I can concede that it's possible that some subset may be worse off. If a new competitor enjoys both an absolute and comparative advantage over you, then others may be less likely to engage in trade with you, for example. This isn't necessarily a problem unless your skillset is sufficiently narrow and you can't easily adapt (e.g., an older person who is very good at making widgets, but is considerably worse at everything else, and will be dead before he sees an offsetting gain from lower prices).

Let's say we have two players, X & Y, in a market consisting of wheat, steel, and cooking (you've gotta eat). X's productivity is 2 wheat, 1 steel, and 1 cooking; Y's productivity is 5 wheat, 10 steel, and 1 cooking. X has a comparative advantage in wheat production. If these two engage in trade, then X produces 4 wheat, and Y produces 4 wheat and 12 steel; after trade, X gains 1 steel, and Y gains 1 wheat. (X might decide to specialize in cooking instead; by producing 2 wheat and 2 cooking, then Y can produce 5 wheat and 20 steel, so the potential gains for both are greater given Y's large opportunity cost for cooking.)

Enter X's competitor, Z, who can produce 10 wheat, 5 steel, and 2 cooking. He has an absolute and comparative advantage over X in every sector. If he similarly trades with Y, then Z can produce 20 wheat and 2 cooking, while Y can produce 5 wheat and 20 steel. If Z charges 2.5 wheat for his cooking, then after trade both parties have 12.5 wheat, 10 steel, and they're fed. Again both are better off than before, but what about X? Does this imply that he's necessarily worse off since Z is better at everything?

If X specializes, then he can produce 3 cooking. Since both Y and Z have high opportunity costs for cooking, by mutual trade they would be able to produce 30 steel and 30 wheat respectively. Once Y and Z trade with each other, they each have 15 wheat and 15 steel before they pay the cook. If X charges 2 wheat and 2 steel, then he's no worse off than if he had just traded with Y directly! Y and Z would certainly be willing to pay such a low price because 13 wheat and 13 steel a piece is more than they could produce otherwise. In fact, they might even be willing to pay up to 2.5 wheat and 5 steel a piece, so X can potentially earn up to 5 wheat and 10 steel--much more than he could earn by trading in a protectionist market.

I'm sorry that illustration was a little long-winded, but I wanted to show that comparative advantage creates new opportunities that weren't possible before, so it isn't the case that comparative advantage necessarily causes some people to be worse off either. In other words, having an unproductive skillset isn't by itself sufficient to cause you to end up as a loser. Your skillset would have to be so extraordinarily unproductive as to overcome the opportunity costs of your potential trading partners. At some point one may begin to wonder exactly how you managed to get along before.

Though it may well be possible that there are losers under free trade, I'd argue that the conditions under which this would actually occur are sufficiently narrow as to be outliers. Most of us have skills that are also useful in other sectors, so the individual that is exceptionally skilled in one niche and practically unskilled in everything else would be the exception rather than the rule.

Vladimir writes:

Nick,

The problem with your example (and most other simple comparative advantage examples) is that it doesn't take into account the scenario where Z's absolute advantage causes the prices of whatever X has to offer, in terms of other goods, to fall drastically, thus leaving him pauperized (even though the total amount of wealth in the world goes up).

In reality, you are capable of doing only certain productive things, and you depend on selling those things to be able to afford other stuff - and a competitor with absolute advantage can cause the prices of everything you have to offer to plummet. Now clearly, some trade is always better than everyone being self-sufficient, and the comparative advantage arguments emphasize this point, but not the more nuanced point that among two different trade scenarios, a switch to a scenario with more free trade can, at least in principle, screw over arbitrarily large numbers of people.

Take this example. An island is inhabited by farmers and fishermen. Everyone needs a loaf of bread and a fish each day to get all the necessary nutrients. The farmers own all the farmland, and the fishermen can only grow tiny amounts of grain in their backyards. At the same time, the fishermen own all the wood supplies for building boats, so that the farmers can only catch a tiny number of fish with primitive sticks. But in his own line of work, everyone produces twice as much bread or fish as the subsistence amount, so trade takes place - one loaf for one fish - and everyone eats happily. Comparative advantage vindicated!

But then suddenly fishermen make contact with a neighboring larger island whose inhabitants have much better land but very little wood for boats, so they're willing to trade far more bread for fish than the farmers on the first island. Now the fishermen can get two loaves for each fish, and clearly, they have no reason to accept the one for one trade with their local farmers any more. So the outcome is that the farmers from the first island can now afford only half a fish with their daily loaf (or some other combination that is below subsistence), and soon die from starvation or nutrient deficiency. Yet an economist touting comparative advantage would insist that they should look forward to the opening of trade with the larger island!

Of course, can always argue that what happens in the real world is more similar to your example than mine. I might even agree that it's usually so. But this requires a separate argument that will grapple with the actual complex situation in our world; these simple thought experiments can't by themselves prove anything either way.

Nick C. writes:

Vladimir,

What would prevent the fishermen from hiring the now-unemployed farmers? Just because they own the capital required for fishing doesn't mean they have to perform the labor themselves.

Suppose a farmer, in his desperation, offers his labor in exchange for one fish and one loaf of bread. Since he can catch two fish a day, he keeps one and gives the other to the fisherman whose boat he's using. With negligible effort, the fisherman trades the fish for two loaves of bread, and gives one loaf to the farmer as salary (the farmer wouldn't perform the work any other way). He still needs to eat, but luckily he owns the wood supplies. Once he performs the one-time task of constructing a new boat, he is able to work half as hard for the same results (one fish and one loaf of bread).

He may also decide to work a full day, giving him a fish and three loaves of bread. He can use this profit to hire two more farmers, which gives him two more fish for negligible effort. He only needs to sell one fish to meet his obligation of a loaf of bread a piece to the farmers, so he can keep the other for himself. Now he has enough employees to provide him with a fish and a loaf of bread for negligible effort. Finally he's free to enjoy leisure time, and he's able to pursue entirely unproductive pastimes if he wants to.

Certainly there will be some fishermen that are too greedy (and shortsighted?) to agree to such an arrangement, but fortunately it only takes a few fishermen to hire many farmers. Once other fishermen are convinced of the benefits, they'll want in on the action, and the entrepreneurship continues until the farmers are at full employment and the fishermen can't hire anyone else if they wanted to.

Even in your example, it's not clear that some subset of people would necessarily be worse off with more free trade. I can see why they wouldn't enthusiastically embrace free trade (maybe the farmers actually enjoy their work and, given a choice, would want to farm instead of fish if the end result is the same), but it doesn't follow that they would definitely be worse off either.

"Of course, can always argue that what happens in the real world is more similar to your example than mine. I might even agree that it's usually so. But this requires a separate argument that will grapple with the actual complex situation in our world; these simple thought experiments can't by themselves prove anything either way."

I think both examples are useful for understanding the advantages and potential drawbacks of comparative advantage. In either example, there is the possibility in principle that one set of people would choose not to trade with another set given the entry of a stronger competitor. For instance, if no fisherman hires any farmer, then the farmers lose. However, this outcome only occurs if the fishermen are satisfied with their arrangement with the other island, they lack ambition to improve their lives further, and they lack compassion to help their neighbors in need when it's well within their ability to do so.

I think such a conclusion conflicts with human nature in two ways: in general people are ambitious enough that they seek to maximize their output with minimal input, and a significant subset of people will do what they can to help those in need once their own needs are taken care of. Examples abound of both ambitious entrepreneurs seeking new opportunities, and of generous benefactors that use their personal surpluses to improve the lot in life of the less fortunate, so I think the burden is probably on the opponent of free trade to justify why such behaviors would have a limited impact under a free trade scenario. In a sense, they have to justify in what plausible world a fisherman would be willing to accept a lower salary and let his neighbor starve, instead of enjoying a higher salary and putting his neighbor to work at the same time!

(Note also that only free trade has the potential to improve everyone's life, while protectionism doesn't. Under free trade, though there is a chance that some number of people might be worse off, it's certainly possible--if not probable--that practically everyone on both islands wins. However, if the first island adopts protectionism to protect domestic farmers, then presumably the people on the second island die with nobody to sell them fish. So, which scenario is better: the one that might leave some people worse off but has the potential to improve everyone's life, or the one that definitely leaves some people worse off in the name of protecting the interests of a preferred subset?)

Vladimir writes:

Nick,

Yes, of course we can always add additional assumptions to these toy models to make the final conclusion go one way or another. For example, we can assume that by the time a new tree is mature enough to use it for a new boat, the farmer can starve to death many times over, because it takes years, and neither his ability to cash-in (fish-in?) the surplus from his farm nor the wages he can earn working as a servant for the fishermen are sufficient to provide subsistence in the meantime. Or numerous other scenarios that leave a bleak outcome assuming rational self-interested behavior on behalf of all parties, without anyone lacking ambition, foresight, etc.

The point I'm trying to underline is that when you argue what the effects of a particular real-world instance of liberalizing trade or migration will be, and you use a toy model in your argument, you have to justify why your believe that your toy model captures reality - since others may come up with other toy models that suggest very different conclusions. Yet economists often present arguments about comparative advantage as if the toy example was by itself an airtight proof that all will be great.

Moreover, I am aware of various potential and actual benefits of trade, and I'm in general agreement with your arguments why they are likely (though this is by no means the whole story). However, discussions like these usually fail to capture many highly relevant complexities of the real world. In particular, they are usually based on a comparison between perfectly free markets versus markets perfectly free except for some specific trade barriers.

But in reality, it is not so - and even if we accept that a perfectly free market would be ideal, it still doesn't follow that if we can identify, say, ten different interventionist policies, it follows that eliminating three out of those ten would be beneficial. On the contrary, it might be a disaster, since the negative effect of regulations and interventionist policies is not a monotonous function of their number, and some of them might be canceling the ill effects of others. Failure to understand this is a very common fallacy among libertarians, and I think this applies to at least some instances of trade liberalization too.


Two Things writes:

What about Negative Productivity?

For example, thieves may exhibit negative productivity--they destroy more value than they produce.

Bryan Caplan set the premise each player has independent productivities for different goods, e.g., Wheat and Steel. Here I shall replace-- without loss of generality-- "Wheat" and "Steel" with "Flamenco-Dancing" and "Roof-Repairs". As before, for simplicity, the price of one unit of each good is $1.


WITHOUT Trade

_______Flamenco-Dancing______Roof-Repairs
You_________1____________________5
Gypsy_______5___________________-5

WITH Trade

_______Flamenco-Dancing______Roof-Repairs
You_________5___________________-1
Gypsy_______5____________________5


As this simple example shows, trade between productive and counter-productive people does not always benefit both.

A more sophisticated model which accounted for elasticities of demand would show that negative-productivity traders can do more damage than this simple model illustrates.

Kenneth A. Regas writes:

In support of Vladimir, and then some:

Who are these farmers, cooks, fishermen, etc. of the toy models - in real life? They aren't actually practitioners of specific trades because people can change jobs. They are sectors of the society ranked by a kind of overall productivity. If you are smart, healthy, reliable, industrious, etc., then you are productive. Take these elements away and you are less productive.

When free trade is entered into with impoverished foreigners, whether by importing them in person or just their products, the sectors of the society with whom the newcomers compete see a reduction in the prices (wages) they can charge. Supply and demand. That's a real reduction in purchasing power that cannot be escaped by retraining, etc. This is true whether the foreigners are engineers or janitors. I live in Southern California, which has an abundant supply of illegal low-end labor at very favorable wages. What do you think are the employment prospects here for the mentally retarded?

In Ricardo terms, when X and Y are trading, the entry of Z into their economy need not be beneficial to both. The only thing for certain is that the total production of X, Y, and Z together increases. I argue that in today's world, the least capable of my countrymen suffer a real decline in their standard of living when Z of the toy model consists of improverished foreigners.

It is a value judgment whether this outcome is favorable. I don't like it.

Ken

Comments for this entry have been closed
Return to top