Considering trade policy: Then and now

Abstract

In this essay, I begin with the pioneering contribution of James Brander's doctoral dissertation to our understanding of international trade under oligopolistic competition. I then mention some of our early joint work showing that oligopoly creates a profit-shifting role for import tariffs and export subsidies. Our work also includes the role of subsidies to R&D/investment for international profit shifting in an oligopoly setting in which firms make sequential decisions with respect to R&D and exports. Trade policy incorporating considerations of rent- or profit-shifting is often referred to as “strategic trade policy.” After a very brief consideration of some criticisms concerning the generality of strategic trade policy results, I turn to a few thoughts on the recent move toward a much more protectionist US trade policy. I argue that a consideration of profit-shifting effects and subsequent investment and location decisions by large oligopolistic firms inside or outside the United States is essential for an understanding of the overall gains and losses from US trade policy.

Prior to the late 1970s and early 1980s, research in international trade was dominated by Ricardian (i.e., “classical”) and Heckscher–Ohlin (i.e., “neoclassical”) models that emphasized the importance of comparative advantage in a world of perfect competition in which individual firms were too small to matter. There was some early consideration of monopoly and, following Krugman (1979), substantial interest in monopolistic competition.1

James (“Jim”) Brander's 1979 PhD dissertation from Stanford University, “The division of labor and intra-industry trade,” includes, to the best of my knowledge, the first formal model of oligopoly and trade. Jim, was motivated by the large and growing phenomenon of intra-industry trade documented by Grubel and Lloyd (1975). Intra-industry trade is the export and import by a country of the same good, or very similar goods, as, for example, in the auto industry. Such trade is not explained by comparative advantage, which predicts that imports and exports are very different. A chapter in the dissertation became Brander (1981), “Intra-industry trade in identical commodities.” Brander and Krugman (1983), “A ‘reciprocal dumping’ model of international trade,” is a second, better-motivated, extended and more general version of the same model.

In the fall of 1979, Jim became an assistant professor at Queen's University in Kingston, Ontario, Canada, where I was a visitor for the year. We started working together, which continued when I moved to Boston College the next year. Our first international trade paper together (Brander and Spencer 1981) considered the domestic welfare effects of a tariff applied to a foreign monopolist engaged in preventing entry by domestic firms. Our next trade paper, Brander and Spencer (1984), “Tariff protection and imperfect competition,” was prepared for a conference in 1982 and later published in a book of conference papers. This paper was the first to demonstrate that an import tariff can raise domestic welfare by shifting profit from a foreign to a domestic rival under oligopolistic competition. It seemed to us that the prevalence of oligopoly could help explain the much more widespread use of import tariffs than would be expected on the basis of the traditional terms-of-trade argument or the infant-industry argument for a tariff. The then dominant models of trade and trade policy based on perfect competition provided very important insights, but they had limited relevance for the understanding of many actual trade flows and actual trade policies.

Our paper, “International R&D rivalry and industrial strategy,” published as Spencer and Brander (1983), extended the scope of rent- or profit-shifting policies to include subsidization of R&D or capital investment in an oligopolistic export-oriented industry. The paper pioneered the use of staged decisions, or sequential rationality, in models combining oligopoly and international trade. In our context, sequential rationality involved governments as well as firms seeing through to the effect of their actions on the equilibrium outcome of subsequent stages of oligopolistic competition. Important previous work included Prescott and Visscher (1976). In that paper, each firm sequentially chooses its location in product space on the basis of correct expectations of the way its decisions influence the decisions of firms yet to locate. Edward C. Prescott was the supervisor of my 1979 PhD from Carnegie Mellon University.

By 1980, the industrial organization literature included analysis of the role of irreversible investments by an established firm in preventing entry. In “Strategic commitment with R&D,” published as Brander and Spencer (1983), we extended this literature to an oligopoly setting in which firms simultaneously choose their investments so as to affect the equilibrium outcome of a subsequent output game. In two highly cited papers, Brander and Lewis (1986, 1988), Jim and Tracy Lewis demonstrated the importance of sequential rationality for the relationship between financial decisions and product market decisions by oligopoly firms.

One issue not addressed in “International R&D rivalry and industrial strategy” was the link between the possible use of trade policy for rent- or profit-shifting purposes and standard trade models. We were strongly encouraged to write such a paper by Ronald W. Jones, a prominent trade economist known for his contributions to Heckscher–Ohlin trade theory. We responded with what became our most cited paper, Brander and Spencer (1985), “Export subsidies and international market share rivalry.” This paper considers a subsidy to the exports of a domestic firm engaged in Cournot competition with a foreign rival in a third country market. The profit shifted from the foreign to the domestic firm is shown to exceed the cost of the subsidy despite the standard adverse effect of the subsidy on the terms of trade.

Subsequent to our early work, there were many substantial contributions by a range of authors to the field of strategic trade policy. One response was criticism of the idea that strategic trade policy could or should be relevant for actual policy choices. A main concern was the sensitivity of the choice of an export subsidy or an export tax to the nature of oligopoly, whether Cournot or Bertrand. By contrast, the subsidization of R&D or capital investment in an export-oriented industry is much more robust as a profit-shifting policy. As shown by Bagwell and Staiger (1994), the Spencer and Brander (1983) result based on Cournot competition also applies to Bertrand competition. However, the industry must be profitable for profits to be shifted. Subsidization of declining industries is not uncommon, and is not a winning strategy. As shown in Brander and Spencer (1984), a tariff applied to final-good imports is also relatively robust as a profit-shifting policy.

Monopolistic competition is the workhorse of most current theory and empirical work in international trade. However, it has been established that very large and profitable firms now dominate US exports in a wide range of industries. This finding has led to a resurgence of interest in the incorporation of oligopoly into empirical trade models.2 The simplifications achieved by abstraction from the effects of trade policy on profits and on oligopolistic interactions are now being criticized as an oversimplification of reality.

Most trade economists, indeed, most economists, have traditionally been strong supporters of trade liberalization to enhance the benefits of globalization, which include access to a wider variety of goods and services, lower prices and higher living standards for many. Correspondingly, the United States had a focus on fostering international treaties to enhance trade.

The first Trump Administration in the United States changed all that. On taking office in 2017, President Trump signed an executive order removing the United States from the Trans-Pacific Partnership (TPP) and declared an end to the era of multinational trade agreements. In 2018, this act was followed by an escalating series of US tariffs covering about US$380 billion in mainly Chinese imports. The Trump tariffs against China largely continued under the Biden Administration that commenced in 2021.

The Biden Administration also instituted a range of trade and industrial policies with a number of stated objectives, such as environmental protection and US self-reliance with respect to what are viewed as “critical products,” including the safeguarding of US technology. In contrast with earlier times when picking winners was viewed as doomed to failure, the United States began to heavily target electric vehicles (EVs) with subsidies. Biden also imposed a prohibitive import tariff of 100% against EVs from China. These EVs are at rock-bottom prices as a result of China's long-standing promotion of their EV industry.

The rent-shifting effects of policies targeting the EV industry are complex. At first glance, profit shifting from foreign firms may seem to be relatively unimportant because the 2023 tax credits for US content in investment and production are available regardless of firm ownership. Nevertheless, US ownership is likely to dominate, and perhaps more importantly, the subsidies are viewed as creating “good jobs.” Good jobs could potentially allow US workers to earn substantial rents due to above-normal wages. Strategic trade policy is a beggar thy neighbour policy, and firms not producing in the United States would suffer from lower exports due to US content requirements.

Partly in response to the US subsidy measures, other governments, including in Europe and Canada, are now also promoting EV production. When many countries subsidize the same product, prices and profits net of subsidies tend to fall, making it highly unlikely that the profit shifted from foreign firms would be sufficient to cover the cost of a subsidy. Other objectives, such as a slowing of climate change or maintaining access to critical supply chains in a crisis, are necessary if such policies are to be justified.

The dramatic departure of US trade policy away from globalization in the Trump and Biden administrations can be partly explained by the rise of China as a world military power, particularly Chinese threats against Taiwan. Longer-standing concerns about Chinese government policies that are viewed as unfairly promoting their exports also play a role. From an economic point of the view, the Trump tariffs on Chinese imports were a clear failure: much of the tariff was passed on to US consumers in higher prices, there was extensive retaliation by China and no material improvement in the very large US trade deficit with China. A significant proportion (about 40%) of US imports from China are intermediate goods; tariffs on these goods raise the costs of US-located firms and reduce US competitiveness in export markets. These negative effects have been somewhat mitigated by the shift of supply chains from China to other low-cost countries such as Mexico and Vietnam. A broadening of tariffs to include these low-cost countries might motivate more production of intermediate goods directly in the United States and further worsen US competitiveness.

Even if strategic objectives around trade policy are not the main motivation for the increase in protectionism by the United States, the dominance of large oligopolistic firms in US trade means that effects on profits, as well as subsequent investment and location decisions by these firms, can be crucial for longer-term welfare in the countries affected. With respect to trade between Canada, Mexico and the United States, intra-industry trade in an oligopoly context, which was first studied by James Brander, plays a central role, particularly in the auto industry. The implications of US tariff and subsidy policies for decisions by firms as to the location of auto plants and production across the three countries is of major significance for economic welfare in both Canada and Mexico.