Funding from the following foundations made possible the larger research project of which this paper is a part: The Tinker Foundation, the Social Science Research Council, the Carnegie Endowment for International Peace, and the Doherty Foundation. We are particularly indebted to Linda Lim for comments and criticism on an earlier draft.

¹ Prebisch, Raùl, The Economic Development of Latin America and Its Principal Problems (New York: United Nations, 1950)Google Scholar.

² The most thorough general statement in the argument is in Little, Ian, Scitovsky, Tibor, and Scott, Maurice, Industry and Trade in Some Developing Countries; A Comparative Study (London: Oxford University Press, 1970)Google Scholar. See also, Balassa, Bela, “Growth Strategies in Semi-Industrial Countries,” Quarterly Journal of Economics 84 (1970): 24–42CrossRefGoogle Scholar; and Schydlowsky, Daniel M., “Latin American Trade Policies in the 1970s: A Prospective Appraisal,” Quarterly Journal of Economics 86 (1972): 263–289CrossRefGoogle Scholar.

³ For a spirited rebuttal of the exhaustion argument, see Hirschman, Albert O., “The Political Economy of Import-Substituting Industrialization in Latin America,” Quarterly Journal of Economics 82 (1968): 1–32CrossRefGoogle Scholar.

⁴ Schydlowsky, Daniel, loc. cit.: 270–271Google Scholar.

⁵ Balassa, Bela, loc. cit.: 28–37Google Scholar.

⁶ For discussions of the Mexican case that focus exclusively on relative factor prices, see Reyes, Saùl Trejo, “El Sector Externo en la Economia Mexicana: Crecimiento Optimo y Politica de Exportaciones,” El Trimestre Economico, 166 (1975): 399–427Google Scholar; and Boatler, Robert W., “Trade Theory Predictions and the Growth of Mexico's Manufactured Exports,” Economic Development and Cultural Change 21 (1974–1975): 491–506Google Scholar.

⁷ Schydlowsky, , loc. cit.: 279Google Scholar.

⁸ Ibid.: 275–276.

⁹ de la Torre, Jose, “Foreign Investment and Export Dependency,” Economic Development and Cultural Change 23 (1974–1975): 137–138CrossRefGoogle Scholar.

¹⁰ For a distinguished exception, see Albert O. Hirschman, loc. cit.

¹¹ For one discussion of the merits of import substitution and export promotion on distributional grounds, see Villarreal, Rene, El Desequilibrio Externo en la Industrializacidn de Mexico (1929–1975); Un Enfoque Estructuralista (Mexico: Fondo de Cultura Economica, 1976), pp. 90–103Google Scholar.

¹² Boatler, Robert, loc. cit.: 506Google Scholar.

¹³ In 1973 the two leading companies, GM and Ford, accounted for over 40 percent of total world automobile sales, and the largest eight for about 85 percent. The largest eight, in order, were GM, Ford, Chrysler, Fiat, Volkswagen, Toyota, Nissan, and Renault. “New Strategies for a World Auto Market,” Business Week 24 November 1973: 38. Six of these firms operate in Mexico with wholly foreign-owned subsidiaries (GM, Ford, Chrysler, Volkswagen, and Nissan) or with licensees (DINA-Renault); and only one Mexican firm is affiliated with a transnational firm not among these eight (VAMSA, a minority-owned subsidiary of American Motors).

¹⁴ Jenkins, Rhys Owen, Dependent Industrialization in Latin A merica; The A utomotive Industry in Argentina, Chile and Mexico (New York: Praeger Publishers, 1977), p. 20 ffGoogle Scholar.

¹⁵ On the role of risk in creating this pattern, see White, Lawrence J., The Automobile Industry Since 1945 (Cambridge: Harvard University Press, 1971)CrossRefGoogle Scholar.

¹⁶ “Decreto que prohibe la importaci6n de motores para automdviles y camiones, asi como de conjuntos mecanicos armados para su uso o ensamble, a partir del 10. de septiembre de 1964,” Diario Oficial (25 August 1962). For a detailed account of how this decree came about, see Douglas Bennett and Kenneth Sharpe, “Agenda Setting and Bargaining Power in TWC-Host Government Conflicts: The Case of the Mexican Automobile Industry, Comparative Politics, forthcoming.

¹⁷ The three firms that disappeared were Representaciones Delta, which planned to manufacture DKWs and perhaps Mercedes; Reo, which planned to produce Toyotas; and Fabrica Nacional de Automoviles (FANASA) which sold only 2000–3000 Borgwards before closing in 1970. Neither Representaciones Delta nor Reo produced any vehicles under the terms of the 1962 Decree.

¹⁸ On the 1969 policy, see “Compensation de Importaciones de Partes Automovilisticas,” Comercio Exterior (1969): 864Google Scholar. For the 1972 Decree, see “Decreto que Fija las Bases para el Desarrollo de la Industria Automotriz,” Diario Oficial (24 October 1972).

¹⁹ “Decreto Para el Fomentode la Industria Automotriz,” Diario Oficial (20 June 1977): 2–7.

²⁰ A study done by Ford and submitted to the government showed that increasing the level of local content to 90 percent would raise prices 36 percent; Ford Motor Company, “Estudio Sobre Contenido Nacional y Generacion de Divisas por Exportaciones,” mimeo, May 1976.

²¹ The Ford study contended that the industry would have to invest $2.50 for every $1.00 of yearly savings in foreign exchange between 1980 and 1985 if increased local content were the option chosen, whereas if compensatory exports were chosen, that same yield in foreign exchange savings could be achieved with $0.60 of investment.

²² One major firm, Fabricas AutoMex, was not opposed to increasing local content. Prior to 1970 this firm was majority Mexican owned; the minority share was Chrysler's. In part, because of Chrysler's pricing policy to its majority Mexican owned subsidiary, Fabricas AutoMex had a difficult time competing with the 100 percent foreign owned subsidiaries of Ford and General Motors. In its efforts to improve its market position, Fabricas AutoMex proposed merging with the two other surviving Mexican owned firms (DINA and VAM). With government production quotas guaranteeing the new firm almost half of the market, and the economies of scale that would come through the merger, Fabricas AutoMex argued that increasing domestic content would be economically feasible and could be done with a minimal rise in consumer prices. The merger, however, fell through, in part because of opposition from the other companies. This intercorporation conflict, which enhanced the bargaining power of the Mexican government, will be discussed in a forthcoming paper.

²³ On this agreement, see Wonnacott, Paul, “Canadian Automotive Protection: Content Provisions, the Bladen Plan, and Recent Tariff Changes,” Canadian Journal of Economics and Political Society 31, 1 (02 1965): 98–115CrossRefGoogle Scholar. On how the pact has fared since see, “Don't Shoot, They're Our Allies,” Forbes (15 May 1973): 135–6; and Fosdick, Richard J., “U.S.-Canada Pact Re-examined,” Automotive Industries (15 06 1975): 14–15Google Scholar.

²⁴ Actually there were quotas before 1962, but these served primarily to limit the total volume of automotive imports.

²⁵ This clause was included in the Decree because of the urging of Ford which had already developed a special kind of tooling for export to other Ford plants.

²⁶ How were the transnational firms in the terminal industry induced to accept this new scheme? That is an important and difficult question, one that concerns the balance of bargaining power between these transnational firms and the Mexican government. We mean to analyze the origins of this 1969 Decree in some detail in a later publication, but it suffices here to note that competitive maneuvering among the firms, particularly involving Fabricas AutoMex (Chrysler) and Ford in which each was trying to capture an enlarged share of the market, served to enhance the bargaining position of the Mexican government and make possible this important shift in policy.

²⁷ At first, even this was not clear. One high government official told us that several of the terminal firms explored the idea of exporting unrelated products—one even considered exporting potatoes—to compensate for their imports.

²⁸ Since the only part the terminal industry could produce (under the 1962 Decree) was the motor (and this was restricted to machining), the terminal firms would have to look to the Mexican parts industry if it wanted to export parts other than motors. In practice, however, many were able to integrate into the parts industry, the most fully integrated operation being that of Volkswagen.

²⁹ Under the terms of the 1972 Decree, the automobile firms were required to compensate a certain percentage of their import bill with exports, but they could also earn an ‘extra-quota’ of cars that could be sold in Mexico for exporting automotive products. During the years of the shortfall, most of the automobile firms asked for and were granted an extra-quota on the basis of certain exports, even though these firms had not met the export obligations for their basic quotas. Clearly, the export shortfall is greater if one considers the total export obligation of the firms as including export commitments made to gain extra-quota than if one asks only if the total exports of the firms compensated for the required percentage of automotive imports (leaving these extra-quota commitments out of consideration).

³⁰ Torre, José de la, “Foreign Investment and Export Dependency,” Economic Development and Cultural Change 23 (1974–1975): 138Google Scholar.

³¹ Fajnzylber, Fernando W., and Tarrago, Trinidad Martinez, “Las Empresas Transnacionales; Expansión a Nivil Mundial y Proyección en la Industria Mexicana,” version preliminar, mimeo, México 1975, p. 538Google Scholar. Their study also showed that TNCs were more likely than national firms to be substantial exporters—35 percent of TNCs but only 26 percent of national firms export more than M$500,000 per year—but that export activities were largely “irrelevant” for the overwhelming majority of TNCs in Mexico. 72 percent of the TNCs in Mexico export less than percent of their output, and 89 percent export less than 3 percent.

³² It follows that if government policies have (and have had) a strong hand in influencing whether transnational corporations will export manufactured goods from developing countries, then it does not make sense to ask in a general sort of way what contribution transnational corporations make to increasing exports from developing countries. Asking the question in this manner ascribes responsibility solely to the TNCs. Instead, we need to ask to what extent can different kinds of government policies in different manufacturing sectors be effective in inducing TNCs to export. This would argue as well against a procedure such as Boatler's in which he tries to explain the unexpectedly high capital intensity of Mexico's manufactured exports, but begins by ruling out of consideration explanations that turn on government policy (“Trade Theory Predictions and the Growth of Mexico's Manufactured Exports,” Economic Development and Cultural Change 23 (1974–1975): 492, note 3)Google Scholar.

³³ Of course many of the Mexican parts firms are subsidiaries of U.S. auto parts firms, having been brought into Mexico with the encouragement of the major auto firms when manufacturing operations commenced in the early 1960s. When this is the case, the auto firms may arrange with their parts suppliers to import parts from these suppliers' subsidiaries in Mexico. Here again, however, transfer pricing arrangements intrude; and procurement decisions are based on more than considerations of cost and quality.

³⁴ There are some examples in Mexico of export manufacture of such parts for the after market. A Monterrey firm makes replacement body stampings for Mercury Cougars for export to the United States, but only because it has been allowed to obtain the necessary stamping dies. Volkswagen de Mexico supplies replacement fender and hood stampings for its ‘Beetle’ to the U.S. after market, a decision that was rational within the firm's global operations with the phase out of that model in the United States, and with the stamping presses and dies close at hand in Mexico.

³⁵ The sanction of quota reductions could be used if only one or two of the terminal firms failed to meet their export commitments. The quotas of the other firms could be correspondingly increased and aggregate levels of production, employment, parts purchases, etc. would be maintained. Given the character of the world automobile industry, it is likely that a short-fall in exports by one firm would be matched by all, as occurred in Mexico in 1974–75 (unless one of the smaller firms in the world industry—American Motors, for example—began to lose its competitive viability). And it is also important to note that while aggregate production might be maintained, there would be distributional consequences, and not just for the terminal firm involved (i.e. for workers, parts suppliers, etc.) See the next section.

³⁶ Fred Bergsten has suggested that any such ‘performance requirements’ placed by LDCs on TNCs may provoke such reprisals. “Coming Investment Wars?” Foreign Affairs (October 1974): 135–52. While the economic consequences may be roughly equivalent—lost jobs, etc.—the greater visibility of a policy that requires the home country to accept imports of products it once made for itself would make more likely the political reaction that could provoke the counter-measures.

³⁷ Moreover, the diadvantaging of these firms through export requirements decreases the possibility of using these two government-owned firms as instruments to regulate the wholly foreign owned firms.

³⁸ According to Business Trends, a Mexican business annual, the auxiliary industry in 1975 consisted of “560 plants (about 280 of which produce only automotive industry equipment)”, Business Trends (1975): 215.

³⁹ By law, all auto parts firms must be at least majority Mexican owned (wholly foreign-owned subsidiaries are not allowed).

⁴⁰ Fernando Fajnzylber and Trinidad Martinez Tarrago, “Las Empresas Transnacionales: Expansión a Nivel Mundial y Proyección en la Industria Mexicana.”

⁴¹ See, for example, Little, Scitovsky, and Scott, p. 11. Product cycle theory as developed by Raymond Vernon and his associates, of course, depicts these as integrated parts of the same process. See, for example, Vernon, Raymond, Sovereignty at Bay: The Multinational Spread of U.S. Enterprises (New York: Basic Books, 1971)Google Scholar; and Wells, Louis T. Jr, ed., The Product Life Cycle and International Trade (Boston: Harvard University, Division of Research, Graduate School of Business Administration, 1972)Google Scholar.

⁴² There are, of course, exceptions. Some auto parts manufacturing has been initiated primarily for export. Chrysler, for example, has begun manufacturing condensers for automobile air-conditioners for export. This kind of manufacture-for-export is almost strictly confined, however, to the parts manufactured by the terminal firms themselves. The general point holds for the supplier firms.

⁴³ The study was one submitted to the Ministry of Patrimony and Industrial Promotion in March 1977 by the Auto Parts section of the National Chamber of Manufacturing Industries. The organization's estimate was that the terminal firms could increase their purchases from auto parts firms (substituting for imports) by 27 percent without any new investments by the auto parts firms, and by an additional 21 percent with minimal new investments.