Leon Tolstoy

A major challenge to be faced by the Free Trade Area of the Americas (FTAA) initiative will be the promotion of similar conditions of competition in the domestic markets of the member countries. Besides the disparities in terms of size and level of economic development, one additional contrast is that 22 countries in the region do not have competition policy institutions. According to a widespread view, the lack of these institutions is not a real problem since trade liberalization is powerful enough to impose market discipline in small economies. Furthermore, authors like Rodriguez and Coate (1996) have been questioning the relevance of an active antitrust policy in situations of unfinished reforms, which has been the case of most Latin American and Caribbean economies during the last 15 years. Instead of supporting market transparency, efficiency and welfare, new born antitrust agencies can easily be captured by special interests and become just another device for rent-seeking and monopoly practices.

This paper argues that the above opinions do not provide sustainable solutions for the FTAA because both of them are only partially true. There is no doubt that free trade is a key instrument to foster competition, but the evidence presented in section 2 shows that the sources of anticompetitive behavior are not associated with market size, but result from distortions that exist in any open economy. Moreover, as section 3 explains, international cartels, mergers and acquisitions through foreign direct investment and the growth strategies of transnational corporations may generate significant transfers of rents among countries and antitrust law is an effective mechanism for extinguishing these welfare losses. On the other hand, as section 4 indicates, capture is likely to be pervasive in every society that does not possess mechanisms for controlling special interests, but this problem affects all public policies, not just antitrust. To illustrate the first point, I will take selected aspects from the history of antitrust enforcement in the United States over the last quarter century, which is also useful to highlight the subtle relationship between antitrust and antidumping. To discuss transnational antitrust cases, I will use a simple analytical framework based on the concept of overlapping games, and to identify possible solutions for the FTAA challenge, I will briefly review the experiences of Australia and New Zealand, which are particularly relevant for Latin America, due to the economic reforms implemented by those countries in the recent past. Finally, section 5 summarizes the main conclusions.

Table 1 shows some figures on antitrust enforcement in the Western Hemisphere. The disparities in the number of cases by country are due to multiple factors. In some countries, like Costa Rica and Panama, the figures refer to the starting moments of the competition policy agency.² In others, like Brazil during 1996-97 and Jamaica during 1994-96, the authorities were busy in curbing certain traditional practices in their countries, and had opened simultaneous investigations against several industries, or the same industry in different parts of the country, on similar grounds. In Colombia, the merger review provisions are very stringent and compel the agency to carry out a large number of cases (see Jatar and Tineo, 1998), while in Argentina, Jamaica and Peru the laws do not regulate mergers and acquisitions (see OAS, 1997a).

Table 1:
Antitrust Cases in the Western Hemisphere

However, even if these peculiarities did not existed, the number of cases should not be expected to be proportional to the country's size. Table 2 includes 15 famous US cases, covering a period from the mid seventies to the early nineties. These cases were analyzed by prominent experts on antitrust and compiled in a book edited by Kwoka and White (1994). Only in three cases C Mobil's attempt to acquire Marathon Oil in 1981, the 1983 joint venture of General Motors and Toyota, and DuPont's growth strategy in the US titanium dioxide industry in the seventies C were the size and other features of the American market relevant issues. All the others could have happened in any small open economy. Some were local events, such as the joint venture of daily newspapers in Detroit, the merger of two hospitals in Virginia and the services rendered by another hospital in New Orleans. Other cases were related to the characteristics of the industry under investigation, and could have been even more serious in smaller economies, like the Coca Cola-Dr. Pepper merger, the computerized reservation systems owned by large airlines, or a price-fixing among manufacturers of gasoline additives (the ethyl case).

Table 2:
Asymmetric Information, Entry Barriers and Market Power in Selected US Antitrust Cases

The most interesting lesson to be drawn from the cases selected by Kwoka and White is the role played by asymmetric information, entry barriers and market power as sources of anticompetitive behavior. Jointly, entry barriers and market power were relevant issues in 12 cases, and asymmetric information was also present in half of those cases. Only in one case - a private litigation between GTE Sylvania and a small distributor of television sets in northern California - did neither asymmetric information nor entry barriers have any significant influence. Entry barrier was the single issue in two cases, the control of the National Collegiate Athletic Association (NCAA) over the broadcast rights to its members' football games and the dispute about the procedures used by the Jefferson Parish Hospital in the supply of anesthesia services.

In textbook descriptions of perfect competition, free entry, constant returns to scale and market transparency are key features. In this stylized world there is no room for antitrust. Every attempt to breach competition rules will be immediately noticed by the economic agents and duly punished by market forces. Conversely, any departure from those three assumptions will engender uneven competition conditions, either among the firms already established in the industry or between incumbents and entrants, although such "imperfections" do not necessarily imply welfare losses. Technical progress, for instance, reshapes periodically the profile of those variables across the economic system by creating entry barriers in some industries while destroying them in others; and by introducing new opportunities for economies of scale and scope which stimulate industrial concentration and, consequently, may strengthen the market power of the innovating firms. Indeed, every technological innovation implies a new form of asymmetric information since the innovating firms have better knowledge of the production frontier than their competitors. But technology also promotes transparency through the reduction of information costs and the diffusion of managerial standards.

This interplay between technical progress and competition poses an intricate challenge to the antitrust agency. As Baumol and Ordover explained: A... while monopoly is rightly recognized as an enemy of static efficiency, there are a number of reasons why it is suspected that its effects on intertemporal efficiency are not so clearly one-sided. Because both large firm size and the possession of market power can, in this view, be helpful to innovation and productivity growth, it is sometimes suggested that antitrust activity, as the enemy of market power and even of large firm size, can serve as an impediment to growth and, by enhancing its costs, as a source of intertemporal inefficiency. Furthermore, when antitrust rules create barriers to efficient interfirm cooperation in research and development and in the exploitation of the fruits of such activity, the adverse consequences from intertemporal efficiency are further exacerbated (1992, p. 83)."

The three sources of anticompetitive behavior can also be strengthened by government actions. Either when protecting the public interest through the regulation of natural monopolies, basic services, and other policies in the areas of environment and national security, or when promoting special interests through trade policies, procurement rules, subsidies and other forms of industrial assistance, the government may create asymmetric information, entry barriers and market power. For this reason, the scope of competition policy is not restricted to the control of business practices, but includes the assumption that the government is implementing policies that are consistently focused on the support of productive efficiency and consumer welfare.

In many situations, market power is engendered by a combination of imperfect information, entry barriers and increasing returns. As Stiglitz observed, "when imperfect information results in the demand curve becoming less than infinitely elastic, it implies that imperfect information confers a degree of monopoly power on the stores (1989, p.775)." However, if the demand remains elastic, no market power can be exercised, even in highly concentrated industries. In fact, one important advancement in antitrust enforcement in recent years has been precisely the adoption of this principle by the merger review procedures of a growing number of countries.

Table 3:
US Anticompetitive Cases by Sector, 1994-98

Table 3 shows the incidence of asymmetric information, entry barriers and market power in 233 cases of anticompetitive behavior filed by the Antitrust Division of the US Department of Justice between December 1994 and August 1998 (see list in the annex). This table is not as accurate as the previous one because here we do not have detailed studies of each case, like those edited by Kwoka and White, but just the summaries that were available at the Division's website as of the first week of August 1998. Thus the figures on entry barriers and market power are probably underestimated since many summaries do not include enough data on the characteristics of the sector under investigation. Besides, most cases refer either to private litigations or to bid rigging, price fixing and other forms of collusion, wherein undisclosed facts are normally the central issue. For this reason, and in contrast with table 2, asymmetric information appears to be so pervasive. However, the basic message is the same: anticompetitive behavior can happen in any sector of the economy and is not related to market size, but to its distortions.

If we compare the list of goods involved in those 223 cases with the 348 antidumping (AD) and countervailing duty measures (CVD) that were active in the United States as of December 1997 (see USITC, 1998, pp.183/192) a curious result emerges. Both lists have just one item in common, ferrosilicon, which is an alloying agent that improves the finished properties of steel products. There were five AD actions against exporters from Brazil (since March 1994), China (March 1993), Kazakstan (April 1993), Russia (June 1993) and Ukraine (April 1993), and three cases of price fixing among manufacturers of that good.³ For many years the steel industry has been the major focus of AD & CVD actions taken by the United States, but, apparently, such protection has not stimulated anticompetitive practices in the domestic market. Besides ferrosilicon, two products linked to that industry have been involved in antitrust investigations in the recent past, laminated tube-making equipment⁴ and steel drums⁵, but these products are not in the relevant market of any AD or CVD measure enacted by the United States.⁶

This evidence illustrates the subtle relationship between antidumping and antitrust. The conflicting goals of these policies are well recognized, but, at least in the United States, they do not affect the same industries. On the one hand, antidumping measures provide a relief to domestic producers from import competition, but do not seem to engender business strategies that would go beyond the limits allowed by the tariff surcharge. On the other hand, those firms that are able to venture into anticompetitive practices do not seem interested in spending resources in rent-seeking activities. Therefore, when the members of a free trade agreement decide to abolish AD & CVD actions among themselves, while harmonizing their competition policies, they are not indeed switching instruments, except for the rare events of predatory pricing. As section 4 shows, they are just making commitments that are natural outcomes of their trade agreement's stated objectives.

In sum, the three sources of anticompetitive behavior can be reinforced both by governmental decisions and the random action of technology, and may lead either to concerted or single-firm practices, but in all cases their ultimate consequence is to promote income redistribution inside the economy. Like most protective mechanisms, anticompetitive practices usually produce immediate and significant results. For instance, Higgins et al. (1996) have estimated that the international aluminum cartel created in 1994 was able to extract over $1 billion from US consumers in less than one year of transactions under that arrangement. Indeed, since 1914 the US antitrust law has provided that any person injured by anticompetitive practices is entitled to recover threefold the damages provoked by such practices (see Section 4 of the Clayton Act), but this rule is restricted, evidently, to domestic cases.

International cartels, mergers and acquisitions through foreign direct investment and the growth strategies of transnational corporations are the most frequent types of antitrust cases in which the process of income redistribution goes beyond the national borders. From the viewpoint of the national interests involved, these cases engender disputes among governments that are similar to those originated from trade policy measures. For this reason competition policy has been included on the negotiating agenda of the World Trade Organization (WTO), although governments are still far from reaching consensus on how to deal with this subject, as several authors have already pointed out (see, inter alia, Hoekman, 1997, McChesney, 1996, Tavares and Tineo, 1998). In contrast with trade policy instruments like tariffs, quotas and subsidies, competition policy issues cannot be settled through mercantilist negotiations, but depend upon the cooperation among national antitrust agencies in the enforcement of their respective domestic laws. As the next section argues, the most important part of this process is accomplished unilaterally, when the competition policy authority is prepared to act as the regulator of last resort in the economy.

Imagine that figure 1 describes the demand for imports of a sophisticated good x in country H (home country) and that x’s producers are members of an international oligopoly which has manufacturing facilities in many parts of the world, including country H. Initially, consumers in that country are importing f units of good x and the price level is b. Any arrangement that provokes a price shift from b to a would be interesting for the exporters from country F (foreign country) if the demand elasticity were less than 1, as in this case the growth of receipts measured by the rectangle abec is larger than the sale losses measured by the rectangle edfg. The inverse of the demand elasticity is the so called Lerner index of market power, and the more powerful the x’s producers are, the greater will be the transfer of monopoly rents from country H to country F. In country H’s domestic market, local manufacturers will be benefited by a similar process, since the demand elasticity for goods produced at home will follow the behavior observed for imports.

Figure 1: The international transfer of monopoly rents

Depending upon the effective market power of x’s producers, the price shift can be obtained through several arrangements. One possibility would be, for instance, an export restraint made by firms from country F, followed by a price increase in the domestic market of country H, which could be described as an informal counterpart of a VER (voluntary export restraint) agreement. Another way would be through transfer pricing among subsidiaries of transnational corporations established in both countries. A third arrangement would be through mergers and acquisitions among firms in either country which could lead to new conditions of competition in the supply of x. Each alternative will demand a particular form of cooperation between the antitrust agencies of each country. In the first case, the antitrust agency in country H will ask its counterpart in country F to initiate an investigation against the exporters of x. In the second case both agencies will probably carry out a joint investigation, while in the third case they could act independently, yet using similar criteria for reviewing the merger effects on their respective markets. These cooperative efforts can be described as an overlapping game, wherein the actions executed by the antitrust agencies are simultaneously limited by the enforcement power of their domestic instruments and the scope of their international agreements. The concept of overlapping or “two-level” games has been widely used in the research about international relations (see Putnam, 1988; Alt and Eichengreen, 1990; Gross and Helpman, 1995; Tavares de Araujo, 1995). It refers to a situation in which a particular player is engaged at the same time in games against distinct opponents, but the options available in one game are restricted by the commitments made in the other.

When dealing with transnational cases, the scope for cooperation among antitrust agencies is initially fixed by the enforcement capabilities allowed by their domestic laws. Mercosur and Nafta are good illustrations of this point. In December 1996, Mercosur countries signed an ambitious protocol setting out guidelines for a common competition policy in the region. The document addresses anticompetitive practices, the procedures for reviewing mergers and acquisitions and the efforts for harmonizing antitrust with other domestic policies. However, at least temporarily, the attainment of these goals will be limited by the current degree of heterogeneity in domestic legislation within Mercosur. Paraguay and Uruguay do not have any laws on this issue, while in Argentina and Brazil, although such legal instruments do exist, their design, their compliance rules and their general purposes differ substantially (see Tavares and Tineo, 1998). Among Nafta countries there is an interim pattern of cooperation that distinguishes the relations between Canada and the United States from the collective efforts for strengthening the Mexican competition policy institutions. In fact, chapter 15 of that agreement is a clear statement that there will be no regional competition policy while the Mexican Federal Competition Commission has not reached the enforcement capabilities of its American and Canadian counterparts.

But the commitment to cooperate establishes new standards for the domestic enforcement of competition principles. For instance, in the hypothetical situation described in figure 1, country F’s authorities would hardly have initiated an investigation against their exporting industry in the absence of an international antitrust agreement. Although the main reason for opening the investigation is the expected reciprocity from country H in symmetrical situations, country F may get additional benefits if the investigation finds domestic market distortions that otherwise would have remained unnoticed. Similarly, cooperation efforts may drive governments toward a more comprehensive approach on competition policy, allowing them to overcome difficult obstacles like the contradiction between antitrust and antidumping. An illuminating example of this process has been the experience of Australia and New Zealand during the recent past, as the next sections shows.

4. Economic reform and transparency: Australia and New Zealand

"Policy intervention was seen as a way of augmenting growth in diverse occupations. An import substitution strategy was a way to mobilize rents from the traditional exportable sector [...] which otherwise would have been capitalized into rural land values. Urban income earners were seen as the beneficiaries. After the Great Depression (1929-32), economic goals became more focused on full employment and the diversification of industry under the direction of government. A wide range of policies, including trade policy, were subordinated to meeting these ends (Lattimore and Wooding, 1996, p.316)."

Anyone familiar with Latin American economic history would bet that the above quotation refers either to Brazil, Mexico, Argentina or one of their neighbors. This is a classical description of the initial steps of the industrialization strategies followed by those countries throughout the twentieth century, from the collapse of the world trading system in the thirties to the debt crisis in the eighties. However, the country under analysis here is New Zealand, which, like Australia, also had opted for the same type of policy during that period, with similar results. Commenting on the Australian case, Bell (1993) noted that: "By the 1960s the tariff structure lacked any overall logic or economic rationale. Many tariffs were anomalous or fortuitous, and little effort was made to avoid over-protection or to promote efficient or economic production (p.28)."

Before the Uruguay Round (1986-93), Australia and New Zealand shared with Latin American countries a common attitude toward multilateral trade negotiations. Their goal was to improve export performance while keeping domestic markets closed. In November 1979, for instance, the Australian Trade Minister, made the following assessment of the Tokyo Round (1973-79): "With the exception of three items C namely tobacco, certain fancy cheese and an item relating to frozen poultry C the tariff rates are at or above current applied rates. This means that Australia has achieved a meaningful and advantageous settlement with the United States, EEC and Japan without reducing the current level of tariff protection on a single tariff item applicable to any manufacturing industry [...] This was, I believe C I am sure industry agrees with me C a commendable result (Rattigan et al., 1989, p.19)." A few weeks later, New Zealand's Prime Minister said: "It has been suggested that New Zealand should dismantle the system of import licensing which has operated for 40 years. I do not subscribe to that view. I have no intention of letting industries go to the wall for the sake of a theory (Lattimore and Wooding, 1996, p.326)."

One peculiarity of the Australian experience of import substitution industrialization was the creation of the Tariff Board in 1921. Its role was to advise the government on the costs and benefits of protection. Besides reviewing individual cases, that institution was supposed to conduct periodic studies on the macroeconomic consequences of the existing trade barriers. The first of these studies was the Brigden Report, which presented a comprehensive analysis of the Australian tariff structure in 1929 and stimulated several academic works during the following decades, including the 1957 classic article by Max Corden on "The Calculation of the Cost of Protection."

However, until the late sixties, the Board's activities engendered no public reaction against protectionism in the country. On the contrary, the general mood was that the welfare gains from industrial diversification would be greater than the protection costs. The tariff was perceived as a social investment whose present value could be weighed against the future benefits produced by economic development (see Corden, 1957). Moreover, in "Protection and Real Wages", one of the most celebrated papers in the history of economic thought, Stolper and Samuelson (1941) concluded that "... in Australia, where land may perhaps be said to be abundant relative to labour, protection might possibly raise the real income of labour (p. 73)." Despite their caveat that A... our argument provides no political ammunition for the protectionist", it really did, and import substitution policies remained popular for many decades, reinforcing the natural barriers already provided by geography and transportation costs.

In the seventies this conventional wisdom started to change. The Tariff Board was transformed into the Industries Assistance Commission (IAC), with a broader mandate to promote transparency in the economy and empowered with adequate instruments to assess the different impacts of public policies, including the creation of domestic entry barriers, uneven conditions of competition among firms in the same industry, and other market distortions. In its first annual report, for 1973/74, the IAC functions were defined as follows: "In summary, the Commission's role is to advise the Government on how individual industries, and industry in general, should be encouraged to develop in Australia. In providing this advice, it is required to have regard to the interests of the community as a whole, and relate its advice to the generally accepted economic and social objectives of the community. The Commission is concerned primarily with the long term development of industries, rather than with the fluctuations which may occur in their rate of development from one year to another, due to temporary changes in their business environment. The principles and objectives in the Industries Assistance Commission Act provide the general policy basis for the long term development of Australian industries (quoted in Rattigan et al., 1989, pp.98/99)."

To foster transparency, the IAC was supposed to keep Australian society informed on three basic topics: [a] the competition conditions in the different sectors of the economy; [b] the effectiveness of current public policies; and [c] the eventual conflicts between the use of public resources to support specific economic activities and the promotion of the community's welfare. Indeed, IAC's ultimate goal was to preserve the debate over what constitutes Australia's "national interest". Although IAC had no enforcement power, the government was required to be aware of the Commission's opinion when changing the level of protection to any industry, with the exception of antidumping and countervailing duties actions.

IAC's only task was to produce accurate information about economic policy on a timely basis, but this was enough to spur bitter animosity both inside the bureaucracy and the private sector. In certain moments, the Commission's roster of powerful enemies included not only leading politicians like J.D. Anthony and Ian Sinclair, trade ministers like James Cairns, but also the Metal Trades Industry Association (MTIA), which had about 6000 members responsible for more than 50% of the labor force in secondary industry (see Rattigan, 1986). According to the national director of MTIA in 1976, the real aim of IAC was to destroy the Australian industry: "We do not need the IAC, which is an excessively elaborate and expensive body of economic theorists, to tell us that most goods we make in Australia can be more cheaply imported by Australia ... What we need is to call a halt to the activities of the IAC in recommending the dismantling of sections of Australian industry. It is a folly of the greatest magnitude if we allow ourselves to be persuaded by a pure economic theory to close our factories because of our high cost structure." (Canberra Times. 24 July 1976; quoted in Rattigan, 1986, p.264)

The process of trade liberalization started in 1973 with an across-the-board tariff cut of 25%. The measure was not enacted for industrial policy reasons, but resulted from a large surplus on the country's balance of payments. Like in most Latin American economies, the process was long and marked by temporary reversals in some industries, specially textiles, clothing, footwear and motor vehicles. As table 4 shows, while the average rate of effective protection of the manufacturing industry suffered a steady decline during 1977-97, those four industries were able to remain away from the general trend. Between 1979 and 1985, the protection rates of textiles jumped from 47% to 74%, and from 140% to 243% in clothing. During 1977-85, footwear producers were bestowed with rates that varied from 121% to 250%, and car manufacturers got the range 67% C 137%. These rates began to decrease after 1985, but even in 1997, when the Australian manufacturing industry had an average rate of 6%, those four sectors were still securing two-digit rates. Tables 5 and 6 tell similar stories for New Zealand and Brazil.⁷