¹ Leading examples of this approach are Vernon, Raymond, The Economic and Political Consequences of Multinational Enterprise: An Anthology (Boston: Harvard Graduate School of Business Administration, 1972)Google Scholar;Smith, David N. and Wells, Louis T., Negotiating Third World Mineral Agreements: Promises as Prologue (Cambridge, Mass.: Ballinger, 1975)Google Scholar;Curry, Robert and Rothschild, Donald, “On Economic Bargaining Between African Governments and Multinational Companies,” Journal of Modern African Studies 12 (02 1974): 173–89CrossRefGoogle Scholar;Turner, Louis, Multinational Companies and the Third World (New York: Allen Lane, 1973)Google Scholar;Zartman, I. William, The Politics of Trade Negotiations Between Africa and the European Economic Community: The Weak Confront the Strong (Princeton, N. J.: Princeton University Press, 1971)Google Scholar;Lietaer, Bernard A., Europe + Latin America + The Multinationals: A Positive Sum Game for the Exchange of Raw Materials and Technology in the 1980s (New York: Praeger, 1980)Google Scholar;Moran, Theodore H., Economic Nationalism and the Politics of International Dependence: Copper in Chile (Princeton, N.J.: Princeton University Press, 1974)Google Scholar.

² For further examples and an extended discussion, seeCobbe, James, Governments and Mining Companies in Developing Countries (Boulder, Col.: Westview Press, 1979)Google Scholar.

³ The qualified statement is necessary since within a bargaining framework that recognizes more than two actors and complex multiple objectives for some of the actors, ownership could plausibly weaken the owner. See Cobbe, , Governments and Mining Companies, and Sideri, S. and Johns, S., eds., Mining for Development in the Third World: Multinational Corporations, State Enterprises and the International Economy (New York: Pergamon Press, 1980)Google Scholar. For an example of an acquisition of formal ownership apparently weakening the position of government, seeLibby, R. T. and Woakes, M. E., “Nationalization and the Displacement of Development Policy in Zambia,” African Studies Review 23 (04 1980): 33–50CrossRefGoogle Scholar.

⁴ See in the context of planningKillick, Tony, “The Possibilities of Development Planning,” Oxford Economic Papers 28 (07 1976): 161–84CrossRefGoogle Scholar. This is not intended to suggest that the more sensitive practitioners using bargaining frameworks do not recognize the importance of competing interest groups and constituencies within host countries.

⁵ Sklar identifies two different ethical traditions in Marxist thought—the historicist and the humanist. The latter, which we are attempting here to follow, engenders critical assessments of social institutions from the standpoint of their human impact.Sklar, R. L., “Political Science and National Integration—A Radical Approach,” Journal of Modern African Studies 5 (03 1967), p. 9CrossRefGoogle Scholar. For examples of discussions of the role of the state in postcolonial states and their bureaucracies, seeAlavi, Hamza, “The State in Post-Colonial Societies—Pakistan and Bangla Desh,” New Left Review no. 74 (07/08 1972)Google Scholar;Saul, John, “The State in Post-Colonial Societies—Tanzania,” Socialist Register (London, 1974)Google Scholar; andLeys, Colin, “The ‘Overdeveloped’ Post Colonial State: A Re-evaluation” Review of African Political Economy no. 5 (01–04 1976)Google Scholar. For a more general discussion of the subject, seePoulantzas, NicosPolitical Power and Social Classes (London: New Left Books, 1973)Google Scholar; andMiliband, Ralph, “Poulantzas and the Capitalist State,” New Left Review no. 82 (11/12 1972). See alsoGoogle ScholarDuvall, R. D. and Freeman, J. R., “The State and Dependent Capitalism,” International Studies Quarterly 25 (03 1981): 99–118CrossRefGoogle Scholar.

⁶ “Comprador” denotes groups who serve as intermediaries between foreign traders and indigenous markets The usage's origins are usually alleged to date from the early institutional arrangements under which the Portuguese were permitted to trade with China. Other examples include slave merchants in West Africa. A good general description of middlemen and compradors in Third World societies is found inFanon, F., The Wretched of the Earth (New York: Grove Press, 1963)Google Scholar; the classic analysis of comprador governments in Third World countries isBaran, Paul A., The Political Economy of Growth (New York: Monthly Review Press, 1957)Google Scholar. For examples of discussions of state compradors in the postcolonial setting, seeTurner, Terisa, “Multinational Corporations and the Instability of the Nigerian State,” Review of African Political Economy 5 (01/04 1976)Google Scholar;Bennoune, Mahfoud, “The Political Economy of Mauretania: Imperialism and Class Struggle,” Review of African Political Economy 12 (05/08 1978)Google Scholar; andBecker, David G., “The New Bourgeoisie and the Limits of Dependency: The Social and Political Impact of the Mining Industry in Peru since 1968,” Ph.D. dissertation, UCLA, 1980Google Scholar.

⁷ Cf. the following statement from a foreigner in Chile: “In the past five years or so, several very promising agreements have been reached between private [mining] corporations and governments of developing countries, which indicate real progress is being achieved toward establishing an improved business relationship between private and public interests. It is ironic that in some respects, it is becoming easier for American corporations to work with foreign governments than with their own government.”Hunt, John P., book review in Economic Geology 75 (06/07 1980), p. 620Google Scholar.

⁸ Note that the political economy approach may well employ bargaining theory arguments; in-deed “the bargaining approach… is a useful tool in obtaining a preliminary picture of what is at stake, gross power relations and hopefully in revealing puzzling inconsistencies” in the words of one advocate of what we call a political economy approach;Burdette, Marcia M., “Nationalization in Zambia: A Critique of Bargaining Theory,” Canadian Journal of African Studies 11 (1977), p. 496Google Scholar. For an alternative critique, that the bargaining model is only appropriate to special cases, seePenrose, Edith, “Ownership and Control: Multinational Firms in Less Developed Coun-tries,” in Helleiner, G. K., ed. A World Divided: The Less Developed Countries in the International Economy (Cambridge: Cambridge University Press, 1976)Google Scholar.

⁹ Foreign equity ownership may be reduced but the foreign partner may continue to provide management, recruitment, technical, personnel, marketing and other services under conditions that remain attractive to the foreign company. European mining companies, at least, have made explicit that it is sudden changes in the operating conditions, not questions of legal ownership, that concern them in the Third World: “The risk that an investment in a mining project may be expropriated is one with which mining companies have been familiar for a long time. In speaking of the remarkable instability of operating conditions, the companies make it clear that this is not their primary concern … ‘The possibility of revolution or sudden political change leading to expropriation has long been recognized by overseas mining companies. What is new is the manner in which host governments can exploit a mining company's necessary vulnerability … to impose major changes in the operating conditions…’ The focus of the companies' concern is not, therefore, title to property or their tenure of mining rights.”Faber, M. and Brown, R., “Changing the Rules of the Game: Political Risk, Instability, and Fair Play in Mineral Concession Contracts,” Third World Quarterly 2 (01 1980): 105–6CrossRefGoogle Scholar . The interior quotation is from “Raw Materials and Political Risk,” submission by European mining companies to the Commission of the European Communities, 1976.

¹⁰ Deception may be necessary since the true intentions of a comprador class may not be tolerable to the general populace. This deception typically takes the form of establishing structures that achieve the desired ends while purporting to do something quite different. In the U.S., this is allegedly common in such areas as government transfer programs and trade protection. Comprador states might do the same thing when they “nationalize” an extractive industry and then award a service contract to the parent of the nationalized firm.Tullock, Gordon, “The Rhetoric and Reality of Redistribution,” Southern Economic Journal 47 (04 1981), p. 904CrossRefGoogle Scholar.

¹¹ Faber, and Brown, , “Changing the Rules,” p. 106Google Scholar.

¹² Jodice, David A., “Sources of Change in Third World Regimes for Foreign Direct Investment, 1968–76,” International Organization 34 (Spring 1980): 177–206CrossRefGoogle Scholar.

¹³ Ibid., p. 177. Note that although Jodice's use of expropriation is sanctioned by Webster's, in common use “expropriation” is taken to imply a loss of value to the expropriated, so we will use “formal nationalization” instead, since Jodice's usage does not require any loss on the part of the expropriated.

¹⁴ Ibid., p. 180.

¹⁵ Ibid., p. 179.

¹⁶ E.g., the formal nationalization of the major Zambian copper companies on 1 January 1970; see Cobbe, Governments and Mining Companies, chap. 8; and the Saudi government's reluctant decision to buy out the 40% interest of its four American industry partners (i.e., Exxon, Mobil, Texaco, and Standard of California) to take full ownership control of Aramco. Since 1973, the companies had been pressing the Saudis to buy out their equity in Aramco obviously to minimize the risk to their exposed equity position. The takeover gives Saudi Arabia 100% ownership of the crude oil concession rights, refining, and production facilities of Aramco. However, this change has had absolutely no effect upon the American firms because they have retained their service contracts to manage the industry and they have exclusive rights to market Aramco's oil.

¹⁷ See Smith and Wells, Negotiating Third World Mineral Agreements, particularly chap. 2.

¹⁸ We can observe if formal nationalization takes place or not, even though we cannot measure or observe the perceived benefit of the event. Within Jodice's framework, therefore, Probit or Logit analysis would appear to be more appropriate techniques for investigating formal nationalization than ordinary least squares. The problems with standard regression procedures arise because the dependent variable that we can accurately observe takes on only two values, one (say, if formal nationalization occurs) and zero (say, if it does not). Ordinary least squares will predict values for the dependent variable meaninglessly outside the zero-to-one range, and the test statistics and R² will be incorrect or not meaningful. Probit and Logit analyses make adjustments to correct for these problems. See, e.g.,Judge, G. G., Griffiths, W. E., Hill, R. C., and Lee, T. C., The Theory and Practice of Econometrics (New York: Wiley, 1980), chap. 14Google Scholar.

¹⁹ Jodice, , “Sources of Change,” p. 189Google Scholar.

²⁰ Development Assistance Directorate, Stock of Private Direct Investments by D.A.C. Countries in Developing Countries, End 1967 (Paris: OECD, 1972)Google Scholar.

²¹ Jodice, in his article (p. 186) givesU.S. Department of Commerce, U.S. Direct Investment Abroad, 1966, Part II: Investment Position, Financial and Operating Data (Washington, D.C.: U.S. Government Printing Office, 1971)Google Scholar as the source for the data on average size in 1966–67 of U.S. foreign affiliates by country. In fact, the source is Part I of that study;U.S. Department of Commerce, U.S. Direct Investment Abroad, 1966, Part I: Balance of Payments Data (Washington, D.C.: U.S. Government Printing Office, 1970): 177–82Google Scholar, which gives an estimate (with some suppressed data that tend to underestimate the number of firms) of the number of U.S. affiliates by country and sector, and the OECD study (fn. 20 above), as described in the text.

²² The Development Assistance Committee (DAC) member countries of the OECD are Australia, Austria, Belgium, Canada, Denmark, France, Germany, Italy, Japan, Netherlands, Norway, Portugal, Sweden, Switzerland, the United Kingdom, and the United States.

²³ For a description of the data sources in this project, seeKobrin, Stephen J., “Foreign Enterprise and Forced Divestment in the LDCs,” International Organization 34 (Winter 1980): 65–88CrossRefGoogle Scholar.

²⁴ Jodice, , “Sources of Change,” p. 187Google Scholar.

²⁵ David A. Jodice, private communication to R. T. Libby. The three are Senegal, Morocco, and the Philippines.

²⁶ Zambia, 100.0 in Jodice, 101.83 by our calculation; Uganda, 45.12 Jodice, 47.35 us; Guinea, 20.29 Jodice, 21.41 us. For reasons unknown to us, the computer printout showing the data actually used by Jodice for his statistical analysis gives Guinea a zero value.

²⁷ David A. Jodice, private communication to R. T. Libby, 12 September 1980.

²⁸ The explanation for the discrepancy in the Togo case appears to be that Jodice based his calculation of Togo's coefficient on zero foreign affiliates in the country's mining sector whereas the Department of Commerce reports that there was in fact one U.S. foreign affiliate in Togo's mining sector. Use of a regional average rather than a country-specific one may also contribute to the discrepancy for the Congo. We have not been able to find an explanation for the discrepancies between the data shown in Jodice's article and those in the computer printouts he has supplied to us. The six countries concerned, the coefficient values shown in the article, the values we recalculated, and the values shown in the computer printout are as follows:

It is possible that the discrepancies are in some way connected with Jodice's definition of the petroleum sector to include only production, since in some of these countries the nationalizations that took place affected only refining, transportation, and/or marketing. However, it is clear from the OECD data (which distinguish these four activities) that the definition was not consistently applied; in our recalculation process, we used OECD data for the entire petroleum sector, including refining, marketing, and transport, and in most cases—including countries such as Jamaica where there was no investment in production—our recalculated data replicate Jodice's very closely. The coefficient for Nicaragua in the printout is wholly inexplicable to us; according to the source for the data on firms nationalized, there were no nationalizations in the mining or petroleum sectors in Nicaragua in 1968–1976; Stephen J. Kobrin, private communication to R. T. Libby, 4 December 1980.

²⁹ E.g.,Tait, Alan A., Gratz, Wilfrid L. M., and Eichengreen, Barry J., “International Comparisons of Taxation for Selected Developing Countries, 1972–76,” IMF Staff Papers 26 (03 1979): 123–56CrossRefGoogle Scholar, and the many references therein.

³⁰ All three anonymous referees of this paper very sensibly suggested we should present data on this point. It remains a topic for further research. We do, however, note a further statistical problem. Jodice's variables “state capacity” (i.e., government revenue as % of 1967 GDP) and “modernity” (i.e., 1967 GDP/capita) are highly correlated; in the full sample, the correlation coefficient is above 0.55, only exceeded by the correlation between PE and “state capacity” in the entire 6 x 6 correlation coefficient matrix. This implies a multicollinearity problem between “modernity” and “state capacity,” which was not addressed in the multivariate analysis.

³¹ Mikesell, Raymond F., New Patterns of World Mineral Development (Washington, D.C.: British-North American Committee, 1979), p. 40Google Scholar.

³² French aid to Zäire in recent years has been interpreted by some as connected with a French desire to supplant the Belgians as the major collaborator in the exploitation of Zaïrian minerals. For an exposition of the idea of competition between industrial countries, seeTanzer, Michael, The Race for Resources: Continuing Struggles over Minerals and Fuels (New York: Monthly Review Press, 1980)Google Scholar.

³³ We also note that we have grave doubts about the proxies used, “collective protest” and “internal war.” We do not question the care and sophistication that went into their construction; but having lived in developing countries, we are extremely dubious about the quality of the underlying conflict data because of the frequency of political control of information media and the paucity of independent sources of information in developing countries.

³⁴ We also ran the multiple regression program using our recalculated values of the PE coefficient as the dependent variable. We did this for the whole sample; for Saudi Arabia, Kuwait, and Zambia excluded (B); and for Ghana, Congo, and all countries with zero PE scores excluded (C). The simple correlation coefficients and multiple R² for the three runs were as follows:

It is immediately evident that, using our recalculated data for the dependent variable, the explanatory power of the regression is greatly reduced, the pattern of signs on the variables is neither consistent with a priori expectations nor invariant to reasonable changes in the sample, and that no firm conclusions can be based on the statistical analysis.