Over the present decade, a difficult task for Latin American governments has been to maintain a coherent stance throughout the process of economic reform. Stabilization, trade liberalization and privatization were meritorious goals of the new policies launched in the region since the late eighties. But in those cases where the government was unable to remove itself from special interests, the reforms turned into short lived monetary anchors, erratic trade policies and badly regulated private monopolies; and the end results were unemployment, increased social inequalities, decadent public services, low rates of economic growth and currency crisis.

Using the Schumpeterian notion of Creative Destruction, this paper discusses the peculiar role of competition policy - here defined in the narrow sense of antitrust - in the process of economic reform. As any public policy, antitrust is constantly submitted to several potential failures, especially those linked to regulatory capture. But it can render two important services to an open developing economy, which are a set of rules that guides the competition process toward efficiency and fairness, and a mechanism that may protect the national interest from anti-competitive practices originated elsewhere in the world economy. To set the stage for the discussion, section 2 briefly recapitulates the Schumpeterian approach to competition and indicates the core of the policy agenda derived from this approach, in which the antitrust authority is compelled to act as the regulator of last resort in the economy. Section 3 presents the domestic challenges faced by the newly created Latin American and Caribbean antitrust agencies, while section 4 deals with cross border issues at four levels: bilateral, sub-regional, hemispheric and multilateral. Section 5 concludes.

Nowadays, the most widely accepted economic approach to competition is the one proposed by Joseph Schumpeter, who defined competition as a dynamic process wherein firms strive to survive under an evolving set of rules that constantly produces winners and losers. In this process, the basic instrument that allows firms to be ahead of their competitors is the introduction of informational asymmetries. Depending upon the momentary set of rules, such asymmetries may result from three types of entrepreneurial activities, which are technological innovation, rent seeking and organized crime, as Baumol (1990) pointed out.

The above approach has a peculiar record in the history of economic thought. Schumpeter revised it several times throughout his professional life. His 1912 book on The Theory of Economic Development established the links between innovation and competition. His 1928 paper on The Instability of Capitalism highlighted the transient character of competition conditions. The influence exerted by innovations in the rhythm of economic activities was extensively documented in his Business Cycles (1939). Finally, in Capitalism, Socialism and Democracy (1942), the random frequency of technical progress and its interplay with competition patterns were brilliantly synthesized under the notion of creative destruction, which is a process of industrial mutation “that incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one.” (1976, p. 83, italics in the original)

Despite the heuristic power of these analytic tools, the Schumpeterian approach remained far from the mainstream of academic research for many decades. But this situation has changed rapidly since the late 1970s, due in part to the work done by Nelson and Winter (1982), and nowadays, “Schumpeter’s assertions inspired what has become the second largest body of empirical literature in the field of industrial organization, exceeded in volume only by the literature investigating the relationship between concentration and profitability.” (Cohen and Levin, 1989, p. 1060) The beginning of this new phase coincided with another important event: the debate engendered by the theory of contestable markets - created by Baumol, Panzar and Willig (1982) - which argues that industry structure is determined endogenously and simultaneously with the vectors of industry outputs and prices.

Schumpeterian competition and contestability theory provide a broad view of the issues related to industrial organization and a challenging agenda for public policy. According to this view, industrial growth results from the interaction among technology, market size and competition strategies. Efficiency can be compatible with any design of industry structure. In every industry, the available technologies will imply a certain degree of scope and scale economies, and a specific ratio between transaction costs and production costs, which in turn will define whether the most efficient structure is a diversified set of firms, an oligopoly or a monopoly. Due to the process of creative destruction, these industrial configurations are, in principle, temporary, including their entry barriers and the corresponding disciplinary power of potential competition.

The policy agenda derived from these theories contains two interconnected challenges. The former is to identify the situations that require intervention and the respective policy instruments to be applied. While the competition process generates technical change and economic growth, there is no guarantee that the public interest is being served, since entry barriers, asymmetric information and market power are natural ingredients of that process. So, the antitrust authorities are constantly on a border line position, where the reasons for intervening are as attractive as those for doing nothing. As Demsetz noted: “In a world in which information is costly and the future is uncertain, a firm that seizes an opportunity to better serve customers does so because it expects to enjoy some protection from rivals because of their ignorance of this opportunity or because of their inability to imitate quickly. […] To destroy such power when it arises may very well remove the incentive for progress. This is to be contrasted with a situation in which a high rate of return is obtained through a successful collusion to restrict output; here there is less danger to progress if the collusive agreement is penalized.” (1973, p.3)

This is a convincing point often made by the Chicago School. As Easterbrook argued in the same vein, "the hallmark of the Chicago approach to antitrust is skepticism. Doubt that we know the optimal organization of industries and markets. Doubt that government could use that knowledge, if it existed, to improve things, given the ubiquitous private adjustments that so often defeat public plans, so that by the time knowledge had been put to use the world has moved on." (1992, p.119) However, a dominant position acquired through a cumulative sequence of successful innovations can be long lasting, and if the antitrust authority is absent the firm may easily venture into abusive behavior whenever the opportunity arises. Hence, skepticism must be blended with a dose of cautious activism.

The latter policy challenge is to ensure that innovation will be the only available instrument for creating informational asymmetries, i.e., that rent seeking and organized crime have been effectively removed from the menu of competition strategies.² In fact, if this task were fully accomplished the former challenge would turn into a rare event, and competition policy would probably lose its relevance. Rent seeking opportunities may arise either from asymmetric information engendered by the competition process, whereby the public authorities are captured by special interests, or from policy priorities independently defined by the government. Similarly, the room for organized crime is directly related to the lack of market transparency and the amount of gaps in the regulatory framework. Therefore, antitrust authorities are supposed not only to be immune to regulatory capture but also to be strong enough to repress such practice elsewhere in the public sector whenever required, acting as a dependable regulator of last resort in the economy.

In this ambiguous environment, one usual prescription for antitrust action is the removal of entry barriers, especially those created by the government. According to Armentano, for instance, “abusive monopoly is always to be associated with governmental interference in production or exchange, and such situations do injure consumers, exclude sellers, and result in an inefficient misallocation of resources. But importantly for this discussion, such monopoly situations are legal, created and sanctioned by the political authority for its own purposes. All such legal restrictions on cooperation or rivalry should be repealed. Thus, ironically or intentionally, the bulk of the abusive monopoly in the business system has always been beyond the scope of antitrust law and antitrust policy.” (Armentano, 1996, p.3) Likewise, Singleton suggested that competition policy should focus its efforts, “first and foremost, on eliminating government-created entry barriers; second, on minimizing natural barriers; and third on prohibiting anti-competitive, entry deterrence by dominant firms.” (1997, p.4) Albeit important, removing entry barriers is just one topic among many others on the competition policy agenda and, evidently, is not a panacea. There are situations in which part of the problem is precisely the lack of entry barriers. As Rashid (1988) showed, quality has a clear tendency to deteriorate in industries with a large number of small firms and low entry barriers.

The promotion of economic development is a national priority in every Latin American or Caribbean country. The role of competition policy in this endeavor is to complement the governmental actions in the areas of education, science and technology, by creating a market environment in which firms can only survive if they are following the international rhythm of technical progress. The region’s experience with import substitution policies provides a good illustration of this point. A well known feature of these growth strategies was the lack of R&D investments by the private sector. Even in those countries that started ambitious public programs of science and technology in the sixties and seventies, such as Brazil, Colombia and Mexico, for instance, the private sector did not fulfil the government’s expectations.

Likewise, the innovative behavior of domestic firms can also be used as a benchmark to measure the success of a trade reform. If, after a certain period, import competition has only led to trade deficits, destruction of local supply, and no formation of endogenous sources of technical change, this will imply that the government has just replaced one group of inconsistent policies with another. Indeed, a normative prescription to be extracted from the analytical framework discussed in section 2 is that the provision of a coherent set of rules and incentives is a crucial assignment for the government in a world of volatile competition conditions.

The enactment of new laws and the creation of autonomous antitrust agencies in many Latin American countries during the nineties have been important initial steps toward that set of rules and incentives. These new institutions are now forging their public image and preparing themselves to act as regulators of last resort in those economies. At the present stage, their two principal challenges are, first, to introduce a clear-cut division of functions between the competition policy authority and the sectoral regulatory agencies; and, second, to curb rent seeking opportunities within a domestic scenario of unfinished economic reforms.³

This is a situation where the identification and eventual removal of entry barriers would imply a major improvement of the competition conditions. However, in contrast with the United States, Canada and the European economies, where any student can read an extensive theoretical and empirical literature on this subject (see Schmalensee and Willig, 1989; Geroski and Schwalbach, 1991; Sutton, 1991; Caves, 1998), I have been unable to find a single economy wide study on Latin American entry barriers. So, while the advice to promote market contestability became a platitude in the discussion about competition policy in developing countries, it has been useless thus far, at least in Latin America, because nobody has the relevant information about the existing entry barriers and their economic consequences.

The typologies suggested by Salop (1979) and Singleton (1997) are useful starting points for an assessment of these barriers. Salop distinguishes an innocent entry barrier, which is unintentionally erected as a side effect of successful innovations, from those strategic obstacles purposely invented to avoid potential competition; while Singleton highlights the importance of government-generated entry barriers. In policy oriented research, it is convenient to introduce two additional subdivisions: temporary versus long-lasting restrictions; and regulatory versus protectionist barriers. The former subdivision provides a time dimension for the competition policy agenda, and the latter separates the governmental measures that are imposed to protect the public interest from those actions that respond to special interests.

After identifying the relevant barriers across the economy, the next step is to analyze their consequences, which consists essentially of inquiring whether the affected sectors are following the international patterns of productivity, profitability and product differentiation. With the information gathered through this exercise the antitrust agency will be, at last, prepared to foster market contestability in some selected areas of the economy. The exercise may also include a cost benefit analysis of each entry barrier, indicating the firms, social groups and geographic regions affected by that restriction. This would allow the development of a competition advocacy program at a national level, mobilizing other public and private institutions in a collective effort to overcome the existing market distortions.

One stimulating example of the results that can be obtained through this type of exercise is the recent paper by Djankov and Hoekman (1998). Although not using the above mentioned typologies, they studied the current conditions of competition in the Slovak Republic by reviewing a set of indicators such as entry and exit, import competition, profitability, revealed comparative advantage, concentration indexes and size distribution of firms. They also presented a brief description of the policies implemented by the government since 1992 in the areas of antitrust, trade, foreign investment and privatization. This description highlights a fundamental aspect of the Slovak experience, which has been the role played by the competition office as regulator of last resort in the economy: “In 1995, the office issued over 200 comments on proposed and existing legislation and decrees; initiated 37 cases against government agencies (mostly provincial and municipal); reviewed 230 privatization deals; and investigated 141 cases dealing with potential anti-competitive practices. Of the latter, 39 dealt with horizontal practices (collusion, cartels, etc.), 77 involved allegations of the abuse of a dominant position, and 25 focused on proposed mergers. Most of these cases centered on the behavior of (public) utilities.” (Djankov and Hoekman, 1998, pp. 1111/2)

The common goal of these actions is to engender a coherent set of market incentives, avoiding those situations whereby the government fosters competition through one channel and creates market distortions through another, as has been so typical in Latin America for many decades. And the results have been well documented by Djankov and Hoekman: Nowadays, the conditions of competition across industries in Slovakia are, in most cases, similar to those prevailing in Belgium, while less than a decade ago the former country had a closed and highly concentrated industrial system. The average rate of profit is around 10 per cent, “with 15 per cent rarely being reach.” (1998, p. 1118)

In their concluding remarks, Djankov and Hoekman stress the importance of economic indicators as competition policy instruments: “Comparisons between industrial structures of roughly similar countries may be useful in establishing a benchmark for changes in the conditions of competition that prevail in a given economy. Clearly the value of such information would be enhanced the greater the number of countries for which data is compiled. Incorporation of descriptive statistics on concentration, import penetration, or price-cost margins for all WTO members in trade policy review reports would provide policymakers and analysts with a better sense of differences in market structure across countries, as well as information on the evolution of trends across countries than is currently the case.” (p. 1126) For Latin America and the Caribbean these data bases would be critical both to strengthen the existing antitrust agencies and, more importantly, to frame the public debate in those 22 countries of the region that do not yet have such institutions, as the next section indicates.

Among the members of the Free Trade Area of the Americas (FTAA) project the international aspects of competition policy are been treated nowadays at four overlapping levels: bilateral, sub-regional, hemispheric and multilateral. Bilateral cooperation among antitrust agencies have been pursued mainly by the United States, which signed agreements with Canada, Australia, Germany and the European Union. Typically these agreements cover the following procedures: [a] Mutual notification of enforcement activities that may affect the interests of the other country, including both anti-competitive practices and mergers. Notifications shall be sufficiently detailed to enable the notified Party to make an initial evaluation of the effect of the enforcement activity on its own important interests, and shall include the nature of the activities under investigation and the legal provisions concerned. Where possible, notifications shall include the names and locations of the persons involved. [b] Officials of either competition policy agency may visit the other country in the course of conducting investigations. [c] Either country may request that other initiate an investigation in its territory on anti-competitive practices that adversely affect the interests of the former country. [d] Mutual assistance in locating and securing evidence and witnesses in the territory of the other country. [e] Regular meetings to discuss policy changes and exchange information on economic sectors of common interest.

Bilateral agreements constitute a partial attempt to cope with the growing number of antitrust cases with cross border implications. In the United States, for example, about 30 per cent of the investigations carried out in 1997 by the Department of Justice involved multinational firms that often had operations in over 20 countries, while in 1993 only 10 per cent of the cases had international dimensions. In recent years, nearly 50 per cent of the merger cases reviewed by the Federal Trade Commission required contacts with foreign antitrust agencies (see U.S. Government, 1998, p. 4). Anti-competitive practices by multinational corporations usually imply significant damages to the public interest, as the two following cases illustrate. The citric acid cartel was controlled by four firms - Bayer, Hoffman-LaRoche, Jungbunzlauer and Archer Daniels Midland Co.. In an operation involving annual sales of about US$ 1.2 billion worldwide, these firms pled guilty for price fixing and sharing the sales of citric acid in the United States and elsewhere during the period July 1991- June 1995. As of March 1998, the U.S. criminal fines imposed on this conspiracy were over US$ 140 million and other private actions were pending (see U.S. Government, 1998, p. 7). In the case of aluminum, Higgins et al. (1996) have estimated that the international cartel created in 1994 was able to extract over $1 billion from US consumers in less than one year of transactions under that arrangement.

Besides the European Union and the Australia-New Zealand Closer Economic Relations Agreement (CER), bilateral agreements are the only existing binding mechanisms for dealing with cross border anti-competitive behavior (see Lloyd and Vautier, 1999). The ongoing discussions at other forums such as, inter alia, the WTO, FTAA, NAFTA, MERCOSUR and Andean Community may eventually lead to effective instruments in the near future, but they are not yet ready to be applied at this moment, as discussed below. However, bilateral agreements have two evident limitations: they do not curb anti-competitive actions originating in third countries and they ignore the interests of the rest of the world. As Falvey and Lloyd (1999) have argued, there are instances in which the maximization of global welfare is not compatible with the short run interests of the nation in which the anti-competitive conduct originates, and this conflict requires multilateral cooperation to be solved.

To start a debate on these issues, the WTO set up the Working Group on the Interaction between Trade and Competition Policy in December 1996. In the following two years about 170 governmental papers were presented to this working group, covering a substantive agenda that goes well beyond the relationship between trade and competition. Despite the active participation of practically all WTO members that have antitrust laws, this debate has been limited by two types of constraints. The former is that any meaningful multilateral treaty on competition rules would only be enforceable if all - or at least the vast majority of - WTO members had national laws covering the relevant topics included in that treaty. The latter is that the present design of the WTO is prepared to deal essentially with government conduct, whereas the major focus of competition policy is the behavior of firms. As Lloyd (1998) observed: “A multilateral system with the WTO acting as an international competition authority would need to investigate private actions in markets. Competition law in this form, by comparison with international trade law, is extremely intensive in its requirement of facts relating to the nature of competition, market share and so on. These vary from case to case and require detailed investigations. When the markets concerned in a case are international, this investigation would require information from different countries. There must be doubts concerning the ability of a remote centralized multilateral authority to understand behavior in markets, and in markets located in different countries to boot.” (pp. 1143-4)

Several suggestions have been made in recent years to improve this situation (see Scherer, 1994; Fox and Ordover, 1997; Jacquemin et al., 1998). The set up of a small and autonomous International Competition Policy Unit (ICPU) was proposed by Jacquemin et al. Located in Geneva and working in close coordination with the WTO, the ICPU would be focused on the long run convergence of international competition policy standards. The preliminary steps toward multilateral negotiations on this subject would include the strengthening of the network among antitrust authorities, which has been initiated by the WTO working group; substantive discussions on how to deal with the global effects of anti-competitive behavior; identification of the points of convergence and divergence among the existing national laws; and a program of technical assistance to help those countries that do not have competition policy at present.

While a multilateral system is not in place, regional trade arrangements may provide an interim solution, since the harmonization of the competition conditions inside the integration project is a natural priority for the member countries. Within the FTAA process NAFTA is the most advanced agreement in this direction, because all the members already have antitrust agencies. Although it does not include any formal commitment on policy harmonization, a natural development of the present situation would be the accession of Mexico into the 1995 US-Canada cooperation agreement. Moreover, as noted by Tavares and Tineo (1999), the convergence in the area of merger regulation is already well advanced. The Andean Community and MERCOSUR also have provisions on competition policy, but they still need some amendments to become operational, as pointed out by Jatar and Tineo (1998) and Tavares and Tineo (1998).

One ordinary issue in the inquiry about the international aspects of competition policy is the participation of small developing economies, which most often do not have antitrust agencies. In Latin America and the Caribbean some small countries such as Costa Rica, Jamaica, Panama and Peru already have these institutions, while others like the Dominican Republic, Trinidad & Tobago, Guatemala and El Salvador are preparing their respective laws. Considering the challenges impinged upon the antitrust authority by the Schumpeterian process of competition, the scant empirical knowledge on the contestability of markets in the region, and the precarious state of the contemporary policy instruments for dealing with cross border issues, one attractive option for the other small FTAA economies would be to follow the attitude of Hong Kong and Singapore, which decided against the adoption of antitrust laws, under the argument that free trade is sufficient to promote competition in their domestic markets (see Lloyd, 1998).

However, the foregoing discussion suggests that this option is not convenient, for three reasons. First, as argued in section 2, anti-competitive behavior is not related to the size of the economy but to some characteristics of the competition process, especially asymmetric information, entry barriers and market dominance. Indeed, as I have reported elsewhere (Tavares, 1998), in the American experience of antitrust enforcement, most cases could have happened in any small open economy, and, very likely, with greater damaging effects on the public interest. There is no doubt that trade liberalization is a powerful instrument, but it is not sufficient to eliminate all the relevant market distortions. Second, large flows of monopoly rents can be extracted from small countries by international cartels, mergers and acquisitions through foreign direct investment and the growth strategies of transnational corporations. The available evidence shows that this is not just a theoretical possibility, but a growing trend in the world economy. In the absence of antitrust institutions these facts seldom become objects of public scrutiny. Third, nearly all Latin American and Caribbean countries belong to sub-regional projects of economic integration which contain explicit commitments on the harmonization of the competition conditions. If these commitments were transformed into operational mechanisms they could provide a timely alternative for those countries that do not have antitrust laws. In the case of the Andean Community and MERCOSUR, after the necessary improvements in the Decision 285 and the 1996 Fortaleza Protocol, Bolivia, Ecuador, Paraguay and Uruguay could use these regional instruments as their antitrust agencies. Eventually, on a later stage, they might set up national agencies whenever the practical experience indicates that this would be the best option. If similar procedures were adopted by SIECA and CARICOM, 17 countries would benefit.⁴

5. Conclusion

In the case of Latin America and the Caribbean, the policy agenda derived from the Schumpeterian approach to competition includes both complex assignments and simple tasks, ready to be implemented, which would pave the way to the former goals. The enactment of a coherent set of market incentives that would lead to economic growth and less inequality in the region depends upon mature institutions that result from a collective learning process on the management of public resources. Removing trade barriers, selling state firms, and even fiscal reforms can be done over a couple of years, if there is political will. But setting up instruments to regulate a newly open economy is a cultural event that implies the development of subtle notions such as credibility, accountability and fairness. Without these notions the distinction among innovation, rent seeking and organized crime is meaningless, for example. And the same happens with the concept of regulator of last resort; and so on.

Two initial steps into this long term endeavor where identified in this paper: the research on market contestability and the improvement of sub-regional antitrust mechanisms. If economic indicators like those used by Djankov and Hoekman (1998) were available for all countries in the Western Hemisphere they would meet several domestic demands - from competition advocacy to technology policy - and would be instrumental for multiple foreign policy goals - from bilateral cooperation to WTO negotiations. Promoting sub-regional mechanisms would yield complementary results, by strengthening the existing antitrust agencies, by filling an institutional gap in 22 countries, and by highlighting the convergence among the different levels of international negotiations. Interestingly, in competition policy there is no room for debates such as regionalism versus multilateralism, or FTAA versus sub-regional arrangements. All of them are necessary under the present conditions of the world economy.