Rev. Mar. 9, 2018 Why Countries Trade: The Theory of Comparative Advantage Increasing Trade We live in an increasingly global economy. World merchandise exports as a share of world gross domestic product (GDP) have almost tripled since 1950, when they were 5% to 6%. Throughout most of the post-world War II period, trade has grown faster than output, and economies have become more open and integrated, especially since the 1990s. Despite the focus on globalization issues in recent years, not all aspects of globalization are unprecedented. As a matter of fact, before World War I, the world experienced significant international trade flows as well. Even so, trade liberalizations are not uncontroversial. Especially since the financial crisis of 2007, there has been renewed skepticism about the gains from trade liberalizations in the public domain, and a call for more protectionism, especially in the United States. Questions have arisen as to whether international trade is responsible for (most of) the rising income inequality, how it relates to the degradation of the environment, or whether trade undercuts labor standards. These concerns notwithstanding, economists often tend to defend free international trade with the theory of comparative advantage. Central to their defense is the belief that the gains from international trade are such that, at least in principle and in the long run, all parties involved can be better off. In this note, the basic theory of comparative advantage is presented with an example that drives home its basic logic when there are technological differences between countries. In addition, we look at some extensions that specify other sources of comparative advantage, as well as some of the imitations of the theories. Comparative Advantage When Nobel Laureate Paul Samuelson was asked to name one of the most important concepts in economics, he answered, comparative advantage. It is a surprisingly common-sense idea, but it is often misunderstood. Comparative advantage tells us why we purchase virtually all goods that we consume in our daily lives instead of producing them ourselves. For instance, most of us buy clothes from stores and cars from car dealerships instead of making them ourselves. In other words, we specialize in what we do, and the jobs that we have, in many instances, reflect what we are best at doing. Michael Jordan became a basketball player and not a house painter because he was so much better at playing basketball than at painting houses. It is immaterial whether Michael Jordan is potentially even better at painting houses than a regular house painter. The point is that he is, relative to house painting, so much better at basketball than the house painter. The same logic applies to countries. Advanced countries as well as developing countries should specialize. Advanced countries should produce more high-tech products than they consume, and trade (export) these for the relatively standard textile products that developing countries produce. When countries specialize and trade, This technical note was prepared by Peter Debaere, Associate Professor of Business Administration. Copyright 2016 by the University of Virginia Darden School Foundation, Charlottesville, VA. All rights reserved. To order copies, send an e-mail to sales@dardenbusinesspublishing.com. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means electronic, mechanical, photocopying, recording, or otherwise without the permission of the Darden School Foundation. Our goal is to publish materials of the highest quality, so please submit any errata to editorial@dardenbusinesspublishing.com.
Page 2 world production is organized efficiently. With international trade, the pie is bigger (more goods are produced), and everyone can be better off. Ultimately, this increased efficiency of world production is what generates the gains from trade. Technology as a Source of Comparative Advantage: An Example Take a world with only two countries, Spain and Holland. Both countries produce and consume two goods: wine and cheese. Figure 1 shows the production possibility frontier of each country. It indicates how much each country can produce if it optimally uses all its labor force. We assume that Spain has 180 workers and Holland has 18 workers. For simplicity, we assume that only labor is needed to produce wine and cheese, and countries are not equally productive in either sector. If Spain used all its labor to produce wine, it could make 18 gallons of wine. If all its labor was active in the cheese sector, it could produce 12 pounds of cheese. As one can see, Holland s production possibility frontier is different from that of Spain. It is much steeper. From the data, we can infer that it takes 10 workers (180/18) to produce one gallon of wine in Spain and 15 workers (180/12) to produce a pound of cheese. Holland is overall much more productive. Only 3 workers are needed to produce a gallon of wine (18/6) and 1.5 workers per pound of cheese (18/12). Figure 1. Production possibility frontier = consumption possibilities before trade. Source: Created by author. The key difference between Spain and Holland is that at any level of production the opportunity cost of producing wine in terms of cheese is not the same. (Another way to say this is that the relative price of one good in terms of the other is unequal.) Spain foregoes 2/3 of a pound of cheese (12/18) for every additional gallon of wine it produces this is the opportunity cost of wine in Spain (the opportunity cost of cheese is the inverse of that: 3/2). Alternatively, Holland foregoes two pounds of cheese for each gallon of wine it wants to produce. A country is said to have a comparative advantage in the good that has the lowest opportunity cost in terms of the other good. Clearly, the opportunity cost of wine in terms of cheese is lower in Spain than in Holland, so Spain is much more efficient in producing wine compared with cheese than Holland. We say that Spain has a comparative advantage in producing wine and Holland a comparative advantage in cheese. When neither country is trading, its consumption options are limited to the downward-sloping line in Figure 1, called the country s production possibility frontier. Countries can consume only what they produce.
Page 3 Note that the specific point of consumption on the production possibility frontier will be determined by how much consumers care about wine versus cheese. Without trade, the prices of cheese and wine will be different in Spain and Holland. If preferences are similar in both countries, we expect Spain to consume more of the relatively cheap (compared to cheese) wine and less of the relatively expensive cheese than in Holland, where cheese should be relatively cheap and wine relatively expensive. A key insight of David Ricardo, who developed the theory of comparative advantage, was realizing that differences in opportunity costs and relative prices determine the pattern of trade. When both countries trade, and there are no transportation costs, they face the same prices on international markets, and these prices differ from the domestic prices before trade. Consumers realize they can be better off if they export the good with the comparative advantage (i.e., relatively cheap before trade) and import the other good (i.e., relatively expensive before trade). In particular, Spain realizes that Holland is not so good at producing wine (in terms of cheese) as it is. Therefore, even if the Dutch pay less for wine (relative to cheese) than they had paid before trade, they would still be paying more than the Spanish had paid for wine before trade. In the extreme, if the Dutch were willing to trade at their domestic price before trade, the Spanish could get up to two pounds of cheese for each gallon of wine they exported. As a matter of fact, so long as the Dutch are willing to pay more than the initial domestic price of wine in terms of cheese (2/3), trade would make Spain better off. (The reverse is true for Holland.) This key insight induces specialization of production: Both countries will produce more of the good they have a comparative advantage in (and export it), and they will produce less of the other good (and instead import it). With complete specialization (in this example, Holland produces only cheese, and Spain produces only wine), it is easy to see that both countries will be better off with trade. For one, the total world production of cheese and wine will be maximized, so the pie is larger. Take a look at Figure 2 for Spain. In case Spain only produces wine, it produces 18 gallons of wine and no cheese (18, 0). It then sells its total wine production (18, 0) at an international price (in terms of cheese) that is higher than the 2/3 pound of cheese per gallon of wine that it got at home before trade. This makes the dashed line that goes through (18, 0), which represents the consumption possibilities in Figure 2, steeper than the production possibility frontier. What is clear from Figure 2 is that international trade implies that the total set of consumption options is more attractive than before when Spain s consumption was limited to the points on the initial production possibility line. The same is true for Holland. Note that the international price of wine in terms of cheese has to lie in between the respective prices of the two countries before they traded. Figure 2. Consumption possibilities versus production possibility frontier with trade. Source: Created by author.
Page 4 A few things are worth noting: Both countries are better with trade. Because there is only one production factor, the model does not tell us whether some people in Spain or Holland could potentially suffer from trade. The whole analysis is driven by comparative advantage. Holland is more productive in making both cheese and wine. Still, it is advantageous for Holland to trade. Needless to say, because Holland is more productive, the wage level will be higher in Holland. A country always has a comparative advantage in something. One can generalize the model we analyzed to multiple countries and multiple goods. 1 Other Factors behind Comparative Advantage: Relative Factor Supplies At the heart of comparative advantage is the notion that across countries the relative price of producing one good in terms of the other is different before countries trade. In Spain, producing wine in terms of cheese is much cheaper than in Holland, which is why Spain, when given the chance, will specialize in the production of wine and consequently export wine and import cheese. What then determines a country s comparative advantage? In this example, we simply assumed there were technological differences that made labor more or less productive in making wine or cheese in either country. Technological differences are an important determinant of comparative advantage. Other causes such as differences in the availability of production factors (e.g., labor or capital), which are often referred to as factor endowments can also play a key role. The theory that describes these factors is known as the Heckscher-Ohlin theory of international trade. To emphasize the difference from the technological explanation of trade, it assumes in its simplest form that there are no technological differences between countries. So why would some countries be able to produce goods more cheaply than other countries? The core idea is relatively straightforward. Again, consider two countries. One country (e.g., Indonesia) has lots of labor (L) and not too much capital (K) relative to labor. The second country (e.g., the United States) has much capital but, relative to that amount of capital, little labor. In other words the K/L ratio of the United States is larger than that of Indonesia, or, as one often says, the United States is capital abundant and Indonesia labor abundant. As before, there are two sectors, say, textiles and computers, and you need a great deal of capital per worker to produce one computer, and not so much capital per worker to produce one piece of cloth. In other words, production in computers is relatively capital intensive, and textile production is relatively labor intensive (i.e., K c /L c > K t > L t ). For simplicity, we assume that the United States and Indonesia have the same technology to produce either computers or textiles. Without trade, you would expect the wage in Indonesia to be lower compared with the cost of/return on capital than in the United States. Why? So much less capital per worker is available in Indonesia than in the United States. Therefore, it would be relatively easy and cheap to produce textiles in Indonesia, given that producing textiles requires a lot of (cheap) labor and not so much (expensive) capital. The reverse will be true in the United States. Because of the abundance of capital, capital will be relatively cheap compared to labor, which is why producing computers, which requires lots of capital and not so much labor, will be relatively cheap in the United States. This Heckscher-Ohlin setup generates a few key predictions. 1 Rüdiger Dornbusch, Stanley Fischer, and Paul Samuelson extend the model to an infinite number of goods in Comparative Advantage, Trade, and Payments in a Ricardian Model with a Continuum of Goods, American Economic Review 67 (December 1977): 823 39. Jonathan Eaton and Samuel Kortum extend the model to more than two countries in Technology, Geography, and Trade, Econometrica 70, no. 5 (September 2002): 1,741 79.
Page 5 1. The pattern of trade: Capital-abundant countries export capital-intensive goods, and labor-abundant countries will export labor-intensive goods. Indeed, because Indonesia can produce textiles more cheaply in terms of computers compared with the United States, Indonesia will produce more textiles than it needs domestically, and it will import all or some of the computers it needs. 2. The factor prices: An increase (drop) in the price of the labor-intensive good increases (decreases) the wage (which is the reward of the factor that is intensively used), relative to the return on capital. Similarly, an increase (drop) in the price of the capital-intensive good increases (decreases) the return on capital relative to the wage. The intuition is straightforward. As Indonesia starts exporting cheap textiles, it experiences an increasing demand for its textiles from the United States. This, in turn, increases the demand for the factor that is intensively used (labor), which, in turn, bids up the wage for textile workers. This is bad news for the other factor: capital, in this case. More workers will be attracted by the higher wages in the textile sector and leave the computer sector. At the same time, with the computer sector shrinking, more capital, then, is needed in the textile sector and will be released, which will bid down the reward for capital. This prediction, sometimes referred to as the Stolper-Samuelson result, is incredibly powerful and politically explosive. It implies that the gains from trade are not equally shared within a country. There can be winners and as well as losers. Think about what a reduction in U.S. tariffs on the imports of textiles (labor-intensive products) into the United States would mean for wages and the return on capital in the United States. You may assume the tariff reduction lowers the price for textiles in the United States. Think about why firms tend to be supportive of such tariff reductions, whereas workers tend to be opposed to them. It is important to note that societies have ways of alleviating the negative impact of trade for the ones who lose, for example by income redistribution (taxation). In other words, the answer to fighting the distributional impact of international trade is not necessarily protectionism that will nullify specialization and the gains from trade. One should also keep in mind that there are many other factors beyond trade liberalizations (such as changes in demand associated with technological change) that can create winners and losers. 3. The structure of the economy: An increase (decrease) in a country s total supply of the factor that is intensively used in producing a good will increase (decrease) the output (and exports) of that good and decrease (increase) the output (and exports) of that good. In other words, as Indonesia accumulates more capital, all else being equal, it will start producing more computers and fewer textiles. This result, often referred to as the Rybczynski effect, helps us understand, for example, the different patterns of trade between developed and developing countries and how these may evolve over time. Think about Japan s rapid capital accumulation since World War II, which was driven by its high savings (and investment) rate and how that affected Japan s production and trade pattern over time. Strengths and Limits of Comparative Advantage The comparative advantage theory is very good at explaining trade between countries that are different in terms of technology and/or in terms of endowments. In other words, it is very well fit to explain trade between the North and the South (i.e., between developed and developing countries). It also explains intersectoral trade well: One country exports goods from a particular sector (wine) and imports goods from a different sector (cheese). Although these are clear strengths of the theory, at the same time, these characteristics expose a clear weakness: The bulk of the world s volume of trade occurs among developed countries that, in terms of technology or relative factor supplies, are similar (especially when compared with the big gap between developed and developing countries). Also, much of the trade among rich countries is intrasector trade (i.e., trade of particular varieties within the same sector). For example, the United States imports BMWs and exports
Page 6 Fords. New trade theory, for which Paul Krugman received the Nobel Prize, addresses intrasector trade among similar countries. To do so, it goes beyond the assumptions of perfect competition and relatively standardized goods (with firms as price takers, and enough entry and exit), which are central to comparative advantage arguments, and it involves imperfect competition, differentiated products, and increasing returns. Beyond this, comparative advantage theories focus on a country s overall determinants of trade, without specifying any. It was not until Melitz 2 that firms have played an explicit role in international trade theories. Note that the Heckscher-Ohlin model that we described in terms of capital and labor can be applied to any pair of factors (i.e., one could compare high-skilled versus low-skilled workers and study how trade affects the gap between low-skilled versus high-skilled wages). For example, how would a liberalization (lowering of tariffs) in low-skilled, labor-intensive sectors such as textiles affect the difference between what high-skilled and lowskilled workers earn? The theories of comparative advantage are typically theories that explain the long run. They implicitly assume that workers (and or capital) can freely change sectors without any cost in pursuit of higher returns. Of course, we all know that there are short-term costs associated with changing sectors, and these tend not to be taken into account. It is worth emphasizing that societies can alleviate the costs that trade liberalizations impose, for example, by providing unemployment insurance or opportunities for education and retraining programs that make it easier for workers who lose their jobs to switch occupations. Finally, note that the main results of the comparative advantage theories can be generalized to include more countries and more goods. Obviously, the predictions are weaker under more general conditions. Also, while most economists associate comparative advantage especially with technological differences between countries or differences in the supply of production factors, sometimes the notion is extended to include other characteristics of countries such as infrastructure or institutions and policies, to the extent that these affect the production in certain sectors differently. 2 Marc J. Melitz, The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity, Econometrica 71, no. 6 (2003): 1,695 725.