Krugman’s Alternative Theory of Trade

To many young economics students, Paul Krugman’s legacy is defined, in large part, by his blog. If you have not had a chance to read his academic work, your view of him is going to be based on Krugman the pundit and Krugman the economist who advocates for fiscal stimulus, the minimum wage — an economist who is turning towards Old Keynesian, or Post Keynesianism, whatever the differences may be. But, modern Krugman is probably not what will be remembered within 15–30 years. Rather, what he is known for by economists, and what he will be known for in the future, is his work in trade and international monetary theory (specifically, exchange rates and capital flows). This is what he won the Nobel Memorial Prize for, in 2008. Krugman introduced a formal model of a new trade theory, an alternative to the theory of comparative advantage.

This post is an attempt to communicate the core of Krugman’s theory, for the layman. I will rely mainly on three of Krugman’s original articles on the subject: Krugman (1979), Krugman (1980), and Krugman (1981). There is also Krugman (1985), but the three earlier papers are shorter and go straight to the point, so I recommend interested readers to read those.1 I am also using my favorite textbook, Krugman, Obstfeld, and Melitz, International Economics.

Prior to the 1980s, most trade theorists thought about international trade within the Ricardian framework of comparative advantage. The theory states that, assuming heterogeneous agents and opportunity costs, a person can specialize in producing the good of lowest opportunity cost to them and trade for other products (produced by other people) and be better off than if there were no trade at all, and each person manufactured everything they want on their own. If someone else can make you t-shirts at a lesser opportunity cost than you, you can buy the t-shirt at that cost, and use your own time towards something more productive. You specialize in products others’ demand, which you can sell to them at at least the cost of production — and your relative costs are the lowest, so that is where your competitive advantage is.2

Ricardian trade theory continued to be developed throughout the 19th and 20th centuries, and one of the directions later economists took Ricardian trade theory in is worth mentioning. In the early 20th century, trade theorists began working towards what is now known as the Heckscher–Ohlin theory. Ohlin would go on to win the Nobel Memorial Prize, in 1977. The main insight the model gives is that countries will tend to specialize in goods that are relatively intensive in the inputs (factors of production) that country is relatively abundant in. Thus, the model looks at differences in factor endowments as a cause of international trade. If the U.S. is relatively abundant in capital and Mexico is relatively abundant in labor, it means that the ratio of labor to capital is lower in the U.S. than it is in Mexico. If labor is cheaper in Mexico, Mexican industry is likely to use a greater labor to capital ratio in their production than U.S. industry. Mexico will also tend to produce more of their labor-intensive goods, because labor is relatively inexpensive (to capital). The U.S. will export capital-intensive goods to Mexico, and Mexico labor-intensive goods to the U.S.

But, there are empirical problems to Ricardian trade theory,

  • Against the predictions of the Heckscher–Ohlin theory, Wassily Leontief, in a 1953 article, published data showing that U.S. exports are less capital-intensive than its imports.
  • Other evidence shows that the degree to which countries specialize is exaggerated in the models, and that intra-trade industry makes up a significant chunk of international exchange that is not accounted for by standard, Ricardian, trade theory.
  • After the Second World War, and before the 1990s, it was found that growth in international trade was not leading to the distributional changes that Ricardian theory predicts. In fact, trade was found to be, in large part, neutral to income distribution.

Economists had been toying around with the relationship between economies of scale and trade, but it wasn’t until Krugman that we had a simple formal model. Krugman also took the original insights and developed them further. He did all this by focusing on internal returns to scale, and by adopting a recent modelling innovation in Dixit and Stiglitz (1977), making it easier to model monopolistic competition.

Assuming a situation where are all agents have identical comparative costs, technologies, and tastes, and there is only one factor, there are none of the standard reasons for trade. But, we assume that there are internal economies of scale. Internal economies of scale occur as long as the average cost per unit of output falls as total output increases. The easiest reasons to cite for internal economies are high fixed costs, where more output allows the firm to spread this fixed cost. If there are internal economies of scale, markets are not perfectly competitive. Instead, there will be less firms, and each firm will produce more. Each firm will also have an incentive to differentiate their product from those of their competitors — if they are close/imperfect substitutes —, to compete for profits. The total number of firms can be said to be determined by average cost and price. As long as price is higher than average cost, it might pay for new firms to enter the market to compete; but, when price equals average cost, profits won’t be high enough for new firms to recover their fixed cost investment.

The size of an economy matters for its well-being. All else equal, larger economies — economies with more people — are wealthier. This is because larger economies will have higher demand, will have more inputs, and therefore more output. More output allows firms to exploit greater internal economies of scale, which in turn lowers average cost. Prices fall, real wages increase, the number of firms will increase, and therefore product diversity will increase (the italicized consequences are welfare-enhancing).

International trade creates similar benefits as population growth. If trade between, say, the U.S. and China suddenly emerged, the market each firm faces would grow. There will be less total firms (if the two countries were isolated, the sum of their firms would be greater than the total number of firms in an integrated market), and each surviving firm would produce more, but all consumers in both countries would be able to buy from a greater range of firms. That is, the diversity of the products offer would increase. The price per unit would also fall, because of the exploitation of further internal economies of scale. Thus, even if none of the standard reasons for trade (comparative advantage) existed,  trade would still occur, to exploit the benefits of internal economies of scale.

One implication is that if there are barriers to trade, factors of production will tend to move to countries where there are economies of scale in industries relatively intensive in a given factor (input). For example, if we assume that the only factor is labor, barriers to trade would induce foreign labor to move to the country with the largest market. Remember, larger markets mean more product diversity and higher real wages, both of which are incentives to immigrants. As immigrants arrive, the market grows further, and real wages and product diversity will increase. Sending states, in turn, become poorer, as product diversity and real wages fall.

Countries will, all else equal, export the goods where domestic demand is highest. It will behoove firms to localize production in markets where demand for that type of product is highest. This is because these firms will be able to exploit greater internal economies of scale than anywhere else. Thus, under conditions of internal economies, countries will tend to export the good they produce more of. In a world of no transaction costs, differences in local demand for a product will induce the country with the greatest internal economies to specialize in that product. In a world of transaction costs — where there are added costs to trade —, specialization will be more limited, because these costs reduce the profitability of exporting. Also, the extent of internal economies will also decide the extent of specialization; the less the opportunities for internal economies, the less a country will specialize in a type of good. Thus, with costs to trade and limits to economies of scale, what we expect is intra-industry trade, as each country produces multiple types of good and trade these between each other, even goods of the same type. But, generally speaking, the country with the larger home market for a given good will be a net exporter of that good, because of economies of scale (and out of interest in minimizing transaction costs).

Finally, the type of trade between two nations has much to do with differences in factor endowments (the type of inputs which are relatively abundant). If two countries are similarly endowed, then trade will tend to be of the intra-industry type. As factor endowments become more unique, the type of trade predicted by the Heckscher–Ohlin model will prevail. The implications for changes in the distribution of income as a result of trade is that if endowments are the same, trade is Pareto optimal. If factor endowments differ, how much they differ will decide relative gains from trade and changes in income distribution. Namely, the more unique a country’s factor endowment, the more the relatively scarce factor will lose from trade and the relatively abundant factor will gain. The scarce factor loses, because with international trade, the price of that product in that country falls (as it faces competition from foreign producers, who have lower costs, because they are in countries that have a relative abundance in that factor). Whether trade is Pareto optimal depends on whether the welfare increase from product differentiation is large enough to make up for the relative loss of income for the scarce factor.

The internal economies of scale argument Krugman formalized allows economists to explain aspects of international trade that were previously not explainable by Ricardian comparative advantage. If there are internal economies of scale — firms are monopolistically competitive —, markets will be supplied by a certain quantity of firms (less than the number in perfectly competitive markets), each producing a greater amount of output than its perfectly competitive analogue. In these cases, even if there are no differences in relative costs, tastes, or technology, there will be gains from trade in the form of lower prices and greater product diversity. Whereas standard Ricardian theory applies when there are differences between agents, economies of scale explain trade when agents are similar. It is an alternative approach to the theories of the division of labor and trade.

All economists borrow from their predecessors and their peers, so Krugman’s theory is by no means entirely original to him. In fact, he cites a number of trade theorists who dabbled with economies of scale prior to him: Herbert Grubel, Bertil Ohlin, Irving Kravis, Bela Balassa, et cetera. But, Krugman was able to formalize the theory in a relatively simple model (more simple than alternative approaches to trade with economies of scale). This allowed him to explore the implications of internal economies in greater detail, and with much more precision. This allowed him to persuade the majority of his peers, whereas previously Ricardian theory had continued to dominate alternatives. This is what rightfully earned Krugman his Nobel Memorial Prize.



  1. You don’t really need to know how to do the math to get the main points in any of those Krugman articles. But, I still recommend reading over them and thinking about them, because sometimes you can get the intuition behind the model without really knowing how to solve it. You will get more out of all those, however, if you can follow the math.
  2. This is an example of a situation in which the whole is greater than the sum of its parts.

8 thoughts on “Krugman’s Alternative Theory of Trade

  1. ckmurray

    Nice post. Economies of scale seem so important to most economic analysis, not just international trade.

    What’s been your experience with Krugman’s trade theory making into mainstream teaching?

    1. JCatalan

      I was taught New Trade Theory in my trade theory class. Of course, we used Krugman’s textbook. But, otherwise, I would have never heard of New Trade Theory at school, as an undergrad. But, I was taught comparative advantage in Econ 101.

      Personally, I think modeling economies of scale and monopolistic competition in the context of other problems would go a long ways in increasing our understanding. But, wasn’t there a marked increase in the use of economies of scale after the 1970s? I’m thinking of New Growth Theory. But, these ventures weren’t always fruitful.

      1. Sébastien Broos

        Indeed, modeling economies of scale and monopolistic competition in a general framework would be amazing. Check this for some recent advances : Evgeny Zhelobodko & Sergey Kokovin & Mathieu Parenti & Jacques‐François Thisse, 2012. “Monopolistic Competition: Beyond the Constant Elasticity of Substitution,” Econometrica

    2. Fernando

      I’m an undergrad, but I learned about Krugman’s theory in my International Economics class (which is mandatory to take).

  2. Pingback: Kirzner, the Entrepreneur, and the Market Process | Economic Thought

  3. Pingback: Do More Immigrants Mean Lower Wages? | Economic Thought

  4. Jaqueline

    Hi, the article is really amazing. I am currently studying Krugman´s modelling and EoS and it helped a lot. Do you happen to know how the elasticity of demand is an issue for the modell?
    That would be really helpful for me to know.


Leave a Reply