Monday, December 08, 2008

Krugman's Nobel Lecture

If you have an hour to present to the Nobel committee, you lay out your Big Idea. Krugman's presentation here is pretty good. He goes over the puzzle, his solution, and why it is interesting.

Basically, international trade is predicated on comparative advantage, so that regardless of total productivity, if the relative productivity of nations in particular goods is different, there are gains from trade. That is, if I can make a $10 worth of wine in 1 hour, and $20 worth of cloth in 1 hour, and you make $5 worth of wine or cloth in 1 hour, it is best that I make all the cloth, and then buy all the wine from you, and you buy cloth from me. The essence of our relative productivity, presumably, if from differences in resources. But over time trade stopped being 'North-South', that is, from developed countries in the Northern Hemisphere to non-developed countries in the Southern Hemispher (trading hi-tech goods for commodities), to more trade between developed countries. This was a puzzle to standard models, that would expect specialization in more fundamental items, like high-tech and low-tech. This is because technology should spread out over time, so Germany and France should have fewer comparative advantages relative to each other, compared to say Germany and Nigeria.

Krugman points out that if there are increasing returns to scale geographically (with an area within a country), a high tech sector in computers may get more and more productive, while an auto manufacturer may also get more productive, though both are 'high tech' in some sense. Thus, increasing returns to scale are important in explaining the growth of various economic sectors.

This is a pretty good idea. It is simple, but that is not a knock. Indeed, many great economic ideas are simple, such as comparative advantage, the Invisible Hand, search theory, Nash equilibria, or ARCH models. That they can be refined mathematically is merely a plus for insiders, but I think the power of simple ideas highlights that there is probably too much mathematics in economics, because the level at which one can debate and discuss the returns to scale argument is really not aided by set theory, (s,S) inventory modeling, or dynamic programming. The number of really good economic insights--true and important--that are only discernible after learning graduate level mathematics, statistics or economics, is a pretty small set (Arrow's Impossibility Theorem, the Revelation Principle).



Krugman then shows that such economies of scale seem to be diminishing. He presents a graph that shows new US auto plants are outside the geographies where there were increasing returns to scale at one time in this area (eg, Detroit), suggesting that there are no longer increasing returns to scale in certain geographies. The new plants are in the south. What is happening to increasing returns to scale? In my mind, it's clearly an attempt to get away from the UAW, the auto unions, which have made productivity growth near impossible with their onerous work rules and penchant for turning any ephemeral productivity increase into parri passu wage increases. So new manufacturers make sure they are far enough away to make unions less likely. Thus at one level, this Nobel Prize winning economist is still pretty much at 40,000 feet in explaining trends happening on the ground. A good idea, but it put academic economics in perspective.

10 comments:

Anonymous said...

funny, i was just on the subject this week. the way i understand it, it makes sense to trade the wine for cloth only if the price is higher than your opportunity cost. so maybe this ricardo stuff is not all what is meant to be? maybe the south just stopped trading cuz it was not worth it?

Eric Falkenstein said...

Ricardo's logic is airtight--given its assumptions, the implication is that trade is beneficial in spite of one side have higher productivity in everything. People who disagree, can point to other issues, like increasing returns to scale, or something, but it still makes sense for Bill Gates to hire someone to monitor his home network, even though he could do it himself better, because he is more valuable doing other things.

Anonymous said...

yes, but what about the other guy's (or the South) opportunity cost? that's my beef.

from wikipedia: i make 100t of food and 100t of clothes. you make 400t of food and 200t of clothes. in order for you to focus on food and trade to me, you require at least 1t of clothes for 2t of food.

Eric Falkenstein said...

Well, its a subtle argument, involving algebra, and is usually a centerpiece of intermediate International Trade course. textbooks like Krugman's go over it quite well. It's well referenced on the web, but if you don't believe it, that's ok by me.

Anonymous said...

don't know what to believe that's why i'm picking your brain. been taking it to face value all this time. i'm going to ask krug on his blog.

Plamen said...

Anonymous: the assumption is (if I understand correctly your problem's setup) you can produce 100 units of food or 100 units of cloth, or any combination in between (say 50-50), and you are indifferent among those combinations. So, your "indifference" price is 1 unit of food for 1 unit of cloth, either way. Mine is 2 units of food for 1 unit of clothes. If we choose any price between those two, we both gain. For example, at 1.5 units of food for one unit of cloth, you can produce 100 cloth, and I can produce 150 food 125 cloth. Then you can trade all your cloth to me for my 150 food (price 1.5). This way you end up with 150 cloth, which is more than the 100 you can produce alone; I end up with 225 cloth, which is more that the 100 I can produce alone. We have both gained from trading.

Plamen said...

Blah, the last 100 should be 200 - apologies.

Plamen said...

Man, I bollocksed it up - let's redo the ending:
"This way you end up with 150 cloth, which is more than the 100 you can produce alone; I end up with 225 cloth, which is more that the 100 I can produce alone. We have both gained from trading."
should instead read
"This way you end up with 150 food, which is more than the 100 you can produce alone; I end up with 225 cloth, which is more that the 200 I can produce alone. We have both gained from trading."

Anonymous said...

that's exactly the problem. in real life we don't "choose any price between those two". price is whatever the price is. maybe lower or higher than the two. in which case the gains of trade happen only for bill gates, ergo my argument. but that doesn't seem to bother economists too much. hey. maybe i can get a nobel myself.

Plamen said...

Anonymous, simple resolution. First, if you walk through the details, it's clear that the parties can ALWAYS engage in mutually beneficial trade, except if their conversion prices are exactly equal. If the price is not inside the range, then no trade occurs, because indeed only one side gains - but that can happen only if one side unreasonably tries to push it outside of the range, and thus decides to punish the other, while suffering too at the same time. The range is sometimes called "negotiation space" - where inside it a price ends up depends on the negotiating skill and power of the trading parties.

Of course, the theory assumes trade between free and informed parties; in reality, of course, that is not always so.

The point stands though - in the vast majority of cases, free trade can benefit both sides. Furthermore, it is even beneficial to trade with others if they impose tariffs and quotas on you, and not impose your own. Most frequently, they are hurting their national well-being in order to protect a "privileged" industry.