And now, another few stupid questions for economists, continuing on from last week. The subject of this week's stupid questions: comparative advantage.

To summarize briefly, comparative advantage is the idea that the world economy as a whole is best off when each country does what that country is best at doing, even if some other country could do it better. And furthermore, that this situation naturally comes about by free trade.

This notion was first put forward by David Ricardo, who came up with a numeric toy example of an economy with two goods and two countries: Portugal can produce either wine or wheat more efficiently than England, but the productivities differ. The details are like this:

	Cost (Man-hours per unit)			Total man-hours available
	Wine	Wheat
England	30	15		270
Portugal	15	10		180

And we also make a few assumptions, just to keep the math easy:

• There are no transport costs or tariffs.
• All wheat and all wine are interchangeable.
• The costs are static -- there is no progress in production, nor economies of scale.

And so forth. From there, a little simple arithmetic shows that if the countries don't trade, the economy as a whole produces 17 units of wheat and 11 of wine. But if England specializes exclusively in wheat, and buys wine from Portugal, the total output is 18 units of wheat, and 12 of wine. So even though Portugal can produce wheat more efficiently than England, it's still better off specializing in wine, and buying its wheat from the less productive English countryside.

And so, the argument goes, specialization in the good with the greatest comparative advantage is good for everybody. Even the toilers in English vineyards, who, according to the argument, should be just as happy to work on a wheat farm. (Part of the implicit background of economic arguments in this era -- or sometimes not so implicit -- is that market forces will act to keep workers' wages uniform and low. People who repeat the same arguments without qualification now, are arguing in effect that in the modern era, workers with IT jobs should take it with simple equanimity if they find that the Market, in its wisdom, has reallocated their labor to Wal-Mart).

Regardless, the effect on the scale of industries is clear: for comparative advantage to yield its benefits, the English vineyards, and Portugese wheat farms, have to get wiped out. And the operation of the market provides for that to happen. In Portugal, investing in vineyards yields greater returns than investing in wheat farms -- so that's where the money will go. In England, it's the reverse. (This happens through fluctuations in prices once goods start to travel; details here).

Or, to change the example slightly, suppose that Portugal is renamed "Nigeria", and its comparative advantage is now in oil. Once again, investing money in drilling oil wells will yield higher returns than just about anything else you could do with that money in Nigeria. With the result that oil drilling (or other extractive industries) would be seen to drive out investment in other kinds of industries in less-developed countries where it was found. As is found to occur -- a triumph for economic theory! Or is it?

Well, it's empirically true that oil-exporting third-world states are peculiarly slow to industrialize. There's even a name for the phenomenon -- the "oil curse". But, to judge from the economic literature, the cause is something of a mystery, and quite a few papers have been published searching for explanations. Here are pointers, for instance, to a few papers which attribute it to institutional factors -- the oil industry, for instance, is said to foster corruption. (One problem with that, as I pointed out here, is that China, the major current success story for ongoing third-world industrialization, is spectacularly corrupt, and industrialization there is going strong). These economists are puzzled, and I am puzzled by their puzzlement -- which brings me to my first stupid question for the week: Isn't the "oil curse" just comparative advantage at work?

More generally, let's consider the implications of this for a third-world "developing" economy which doesn't have any natural resources. In the absence of trade barriers, what should that country produce? The principle of comparative advantage would say that it should be whatever that country is best at producing. Or perhaps, in cases like this, it's more appropriate to say that the country's capital should go to whatever economic activity is the least comparatively disadvantaged, since countries like this don't do anything well. Either way, we can ask, what's that going to be?

Well, what do people there know how to do? They're able to do farming -- they have traditions in that going back millenia. And it doesn't necessarily require a whole lot of infrastructure beyond the ground itself, nor training; doesn't require workers to keep exact shift schedules; and so forth. If, on the other hand, you plop a factory in the middle of a country like that, you also need other facilities for that factory to succeed -- power and transportation infrastructure, and a trained work force -- which simply don't exist in the country, and which the factory is not likely to pay for on its own. So, our impoverished country has a very hard time making good use of a factory -- but it is able to get agricultural products out of a patch of dirt. Now, it may not do either of these things well, but it seems likely that its comparative advantage is going to be with agriculture.

And, in fact, you can see particular countries where things seem to be playing out just like that. All over Africa, for instance, people are dressing themselves in the cast-off T-shirts of Americans. In fact, Nigeria (where we'd expect things to be particularly skewed with the oil money) is now trying to ban imports of used clothing altogether, in order to preserve some kind of a local textile industry. Which might be a completely vain pursuit in the world of Comparative Advantage, just like the preservation of Ricardo's English vineyards, if the comparative advantage lies elsewhere.

What applies to textiles -- hardly high tech these days -- applies just as much to any other industry. African local industry is being starved out by first-world castoffs in everything from computers to mattresses to trucks. And the logic is the same in each case: the theory of comparative advantage seems to me to say that in the absence of trade barriers, industrialization cannot occur in a country if its comparative advantage lies in oil drilling or agriculture, because the lack of infrastructure dicates that you'll make more money running a farm in those countries, than putting up a factory there. And that's a good thing, because total world economic output is maximized by dedicating the continent to dirt farming -- in perpetuity.

The tricky words there are "in perpetuity". One of the assumptions of Ricardo's theory, as I mentioned above, is that the cost matrix stays the same -- that neither economies of scale, nor synergy, nor anything else is going to change it. It's a static theory. But an integral part of industrialization is doing things -- training the work force, building up transportation and energy delivery infrastructure -- which make it possible for a factory owner to get more bang for the buck. And these things take time. Longer than a decade. Longer than investors in the modern world are generally willing to wait.

To put it another way, a great deal comparative advantage in industrial goods comes from having a lot of industry, and the support structure for it, already in place. So, for countries that don't, what is to be done?

Well, one possibility is to put up trade barriers to force local people to invest in creating industries which would otherwise make no economic sense -- and in their share of a support structure which makes it cheaper to build the next factory. The idea here isn't that trade barriers are a positive good, to be maintained in perpetuity -- but rather, that for developing economies, they are a necessary evil, to be dropped when local industry can stand on its own. It's not a fairy tale to suppose that can happen. That's what South Korea did, for example, last century, pursuing mercantilist policies while building up its industry in the 1960s and 70s, and progressively lowering trade barriers after that. For that matter, it's more or less what the U.S. did in the nineteenth century. And it's what Mauritius, one of the few bright lights in the dismal African economic picture is doing right now. But this flies in the face of conventional "Washington consensus" advice to developing governments -- which is to drop all trade barriers immediately, and keep them down.

Another possibility is to have the government itself create and provide some of the support structure. To have it create electrical, water, and sewer systems, so that private industry doesn't have to shoulder that burden. But that again flies in the face of "Washington consensus" advice, which is to privatize, on the theory that "the competitive market" can provide more economical service. In practice, unfortunately, "privatization" all too often takes the form of first-world conglomerates getting monopolies, as in the case of Bolivian water utilities. The rationale was that the first-world monopolist, Bechtel, would serve more efficiently than the old public utility -- but in fact, it took advantage of its monopoly and the lack of competition to raise rates to levels that were literally killing Bolivians.

To recap: starting from a perfectly orthodox statement of the principle of comparative advantage, I've got the very heterodox position that in conditions of pure free trade, it's going to be extremely difficult for developing countries to industrialize. All of which goes way, way against economic conventional wisdom. So, the remaining dumb questions for the week: where are the countries where conventional wisdom has worked, and what's wrong with this argument?

Well, aside from the garbled version of the wine-and-wheat example that was originally posted up top. Repeat ten times: proofreading cannot be omitted. Proofreading cannot be omitted. Proofreading...