Michael Pettis@michaelxpettis
David Ricardo showed in 1817 that if each country produces according to its comparative advantage, global output is maximized. Many analysts argue today that China's trade surplus reflects its comparative advantage in producing everything, and so its trade surplus is good for global growth.
But that's not true, and that is certainly not what Ricardo showed. China has a competitive advantage, not a comparative advantage, largely for the same reason that it has such weak consumption – large direct and indirect transfers from the household sector, in the form of an undervalued currency, cheap credit, restricted labor, overspending on infrastructure, land and other subsidies, etc., subsidize production across the board at the expense of household consumption.
The point is that these lower production costs give China a competitive advantage in producing most things, but this not the same as a comparative advantage. The former means you are able to produce more cheaply than your trade partners. The latter means that the relative "cheapness" with which you produce some goods is greater than the relative "cheapness" with which you produce other goods, so that you can only have a comparative advantage in roughly half the goods you produce. This is because comparative advantage is about relative costs, not absolute costs, and while you can have lower absolute costs in most or even all things, by definition you cannot have lower relative costs in more than half of what you produce.
Ricardo's example shows this very clearly. In his model. Portugal produced both textiles and wine more cheaply than England, which meant that Portugal had a competitive advantage in all goods, and England a competitive disadvantage in all goods.
But Ricardo did not argue that the world would benefit if Portugal produced both wine and textiles, with England producing neither and acquiring them by running trade deficits with Portugal. Instead he showed that because the relative "cheapness" with which Portugal produced wine was greater than the relative "cheapness" with which it produced textiles, Portugal only had a comparative advantage in producing wine, and England had a comparative advantage in producing textiles.
He showed that if Portugal only produced wine, and exported some of it to England to buy textiles, and if England only produced textiles, and exported some of it to Portugal to buy wine, trade would be balanced and total output would be maximized even with Portugal's competitive advantage in both.
If you do the math in Ricardo's model, you quickly see that global output is maximized only when global trade is balanced. For global production to be maximized, countries should not be net exporters of all goods in which they have a competitive advantage. They should be net exporters only of those goods in which they have a comparative advantage, and they should use the proceeds of those export revenues to import those goods in which they have a comparative disadvantage.
Once you allow trade to become persistently unbalanced, you run into the problem that Keynes identified in the 1930s – trade imbalances allow countries that have become more competitive by suppressing domestic demand to export weak domestic demand to the rest of the world. When that happens, either total global production is reduced and unemployment rises, or debt must rise in the deficit country to make up for the weak demand in the surplus country and to prevent unemployment from rising.
Economists often cite David Ricardo's model of comparative advantage as one of the few, great, non-trivial models in economics, and so it is worrying that so few academic economists understand the math behind the model. Ricardo's whole point was to make the unintuitive point that competitive advantage is not the same as comparative advantage, and that the world benefits from balanced trade even when one country can produce everything more cheaply.
This becomes more obvious when you realize that just by changing the value of the currency you can shift competitive advantage from one country to another, whereas comparative advantage is structural, and does not shift so easily. The main point, which surprisingly few academic economists understand, is that Ricardo's model of comparative advantage is a model of balanced trade.
