27II  C THE CAUSES OF TRADECTHE CAUSES OF TRADEFrom an economic perspective, the case for freer trade rests on the existence of gains from trade and most economists typically agree that there are gains from trade. In recent years, however, free trade has increasingly come under fire and it is not uncommon to hear trade sceptics say that economists’ arguments in favour of free trade and in particular comparative advantage may have been valid at the time of Ricardo (in the early 19thcentury) but that they are no longer valid in today’s globalized world. This section critically assesses the relevance of economic theories of international trade in today’s global trading environment. Most trade models are designed to answer two closely related questions: what goods do countries trade and why. While the main focus of this section is on the causes of trade, the discussion often touches upon the question of the patterns of trade.This assessment of the relevance of trade theories is based on an overview of the theoretical models as well as of the empirical literature. This section begins by examining how robust the theories are and how far they can be generalized. This is an important part of the discussion – in particular, when the traditional approach is considered. This is because the traditional case for gains from trade is largely theoretical. In fact, it could even be argued, as Leamer and Levinsohn (1995) do, that “though obviously important and theoretically robust, the existence of gains from exchange is fundamentally a premise of economics, not a testable implication of a particular model”. Bearing this in mind, this section also reviews empirical work that tests trade theories and that attempts to estimate the relative importance of different types of gains from trade. The idea that there are gains from trade is the central proposition of normative trade theory.1 The gains-from-trade theorem states that if a country can trade at any price ratio other than its domestic prices, it will be better off than in autarky – or self-sufficiency.2 More generally, the basic gains from trade propositions are that:3 i) free trade is better than autarky; ii) restricted trade (i.e. trade restricted by trade barriers) is better than autarky; and, iii) for a small country (i.e. a country too small to influence world prices) free trade is better than restricted trade. Samuelson (1939) showed that there are potential gains from trade for small countries provided world prices diverge from autarky prices. Kemp (1962) showed that restricted trade is better than no trade. He also extended the argument to the large country case, proving that free trade is potentially superior to autarky, in the case when there are many commodities and factors and with variable factor supplies. As noted by Deardorff (2005a), most treatments of the gains from trade say that if trade could potentially benefit all members of a country’s population (assuming their preferences and income were identical), it is regarded as benefiting the country because some form of income redistribution among the country’s consumers is assumed to be feasible. Beyond the feasibility of income redistribution in the form of lump-sum transfers (which is necessary to avoid market distortions associated with taxes), these results are based on a number of other key assumptions, notably constant returns to scale,4perfect competition,5 no other market distortions, such as externalities,6 and the flexibility in the prices of factors of production (principally capital and labour) that ensure full employment. While the main message of the gains-from-trade theory remains valid when some of those assumptions are relaxed (for example, feasibility of lump-sum transfers), attempts to relax others (such as constant returns to scale) introduce significant complexities (Corden, 1984).These basic propositions about the gains from trade, however, are not the end of the story. First, as pointed out by Corden (1984), the divergence between autarky and free trade prices is only an approximate explanation of the gains from trade. A full explanation of those gains should link them to the causes of trade – that is, to the elements that give rise to divergence between autarky and free trade prices. Those elements are the ones that lie behind the sources of comparative advantage. They would include differences in technology or differences in endowments. Second, economic theory points at other forms of gains from trade that are not linked to differences between countries. In particular, countries trade to achieve economies of scale in production7 or to have access to a broader variety of goods. Also, if the opening-up of trade reduces or eliminates monopoly power or enhances productivity, there will be gains from trade additional to the usual ones. Finally, trade may have positive growth effects.28WORLD TR ADE REPORT 2008This section covers the traditional gains from trade and their underlying causes, the gains from trade highlighted in the more recent trade theories, and the dynamic gains from trade. Each sub-section starts with a brief presentation of a theory focusing on these specific gains from trade. The robustness of the theories to changes in their main assumptions is examined. Finally, the empirical evidence concerning the proposed rationales for international trade is reviewed. Before considering the simplified theoretical frameworks (models) which focus on any particular source of gains from trade, it is important to emphasize that patterns of international trade typically reflect the interaction of several different causes. International trade theories and specific applications of the theories (models) should not be seen as mutually exclusive. This is of particular importance when trying to assess their relevance. The validity of a particular theory should be assessed on the basis of its capacity to explain trade in its limited domain. North-South trade might be explained by models which link trade patterns to differences between countries, while a model of monopolistic competition may best characterize trade between similar countries.1. THE TRADITIONAL APPROACH: GAINS FROM SPECIALIZATIONUntil recently, most trade models explained the commodity pattern of trade in terms of the law of comparative advantage.8Before turning to particular models, such as the Ricardian model or the Heckscher-Ohlin model, which focus on particular product