The author would like to thank David Jodice of Harvard University and Jan Hack and Ken Rodman of MIT for research assistance. The UN Centre for Transnational Corporations, the Overseas Private Investment Corporation, and New York University Project on Multinational Enterprise provided funding. Jean Boddewyn, Robert Hawkins, Richard Robinson, Ray Vernon, Gerald West, and Burns Weston, as well as two anonymous reviewers for International Organization commented critically on earlier drafts.

¹ Vernon has defined a multinational enterprise as “… a cluster of corporations of different nationalities that are joined together by a parent company through bonds of common ownership, that respond to a common strategy, and draw on a common pool of financial and human resources.” See Raymond, Vernon, “Multinational Enterprise and National Economic Goals,” in Transnational Relations and World Politics, Keohane, Robert O. and Nye, Joseph S., eds. (Cambridge: Harvard University Press, 1971), p. 344.Google Scholar

² C. Fred, Bergsten, “Coming Investment Wars?”, Foreign Affairs 53 (1974): 142.Google Scholar

³ Kobrin, Stephen J., “Political-Economic Factors Underlying the Propensity to Expropriate Foreign Enterprise” (mimeo, Sloan School of Management, M.I.T., 1979).Google Scholar

⁴ See, Ingram, George M., Expropriation of U.S. Property in South America (New York: Praeger Publishers, 1974)Google Scholar and Moran, Theodore H., Multinational Corporations and the Politics of Dependence (Princeton: Princeton University Press, 1974).Google Scholar

⁵ See Bradley, David G., “Managing Against Expropriation,” Harvard Business Review (07–08 1977): 75–83Google Scholar; Hawkins, Robert G., Norman, Mintz, and Michael, Provissiero, “Government Takeovers of U.S. Foreign Affiliates,” Journal of International Business Studies 7 (1976): 3–16CrossRefGoogle Scholar; Harrald, Knudsen, Expropriation of Foreign Private Investments in Latin America (Bergen: Universitetstorlaget, 1974)Google Scholar; Root, Franklin R., “The Expropriation Experience of American Companies,” Business Horizons XI (04 1968): 69–74CrossRefGoogle Scholar; and J. Frederick, Truitt, Expropriation of Private Foreign Investment (Bloomington, Ind.: Graduate School of Business, Division of Research, Indiana University, 1974).Google Scholar

⁶ The nucleus of the data base (875 of 1535 firms) was compiled by the United Nations Economic and Social Council. See, Economic and Social Council, “Permanent Sovereignty Over Natural Resources” Annex, , “A Profile of Recent Cases of Nationalizations or Takeovers of Foreign Enterprise” (New York: United Nations A/9716, 1974).Google Scholar

⁷ Truitt, , op. cit., p. 9Google Scholar. Expropriation is generally accepted as a sovereign right given both a public purpose and prompt, adequate, and effective compensation. Taking of property without the latter becomes confiscation. The difficulties inherent in establishing when a purpose is public (e.g., to establish private local control of a key sector) or what constitutes prompt, adequate, or effective compensation are legion and have led to a vast legal literature on the subject. Gillian, White, Nationalization of Foreign Property (London: Stevens and Sons, Ltd., 1961)Google Scholar is a basic reference in this area. For a more recent view see Nicholas, Doman, “New Developments in the Field of Nationalization,” New York University Journal of International Law and Politics 3 (Winter 1970): 306–322Google Scholar; Henry, Landau, “Economic and Political Nationalism and Private Foreign Investments,” Denver Journal of International Law and Policy 2 (1972): 169–178Google Scholar; Lillich, R. and Burns, Weston, International Claims: Their Settlement by Lump Sum Agreements (Charlottesville: University Press of Virginia, 1975). While questions of law are certainly important, at least one author has concluded that, “In the contrived clash between the contractual rights and security of private investments and the proclaimed invulnerable sovereign right of the capital importing states to seize possession and transfer title of foreign property to the state, traditional international law offers little economic relief to the former owner.” Doman, op. cit., p. 322.Google Scholar

⁸ See, Weigel, Dale R. and Burns, Weston, “Valuation Upon the Deprivation of Foreign Enterprise: A Policy Oriented Approach to the Problem of Compensation Under International Law,” in The Valuation of Nationalized Property in International Law, Lillich, Richard B., ed. (Charlottesville: The University Press of Virginia, 1972), p. 3Google Scholar. Weston prefers wealth deprivation which “ … describes the public or publicly sanctioned imposition of a wealth loss (or blocking of a wealth gain). … ” Similarly, Hahlo, Smith and Wright suggest extinguishment as a general term. Hahlo, H., Smith, Graham J. and Wright, Richard W., Nationalism and the Multinational Enterprise: Legal, Managerial and Economic Aspects (Dobbs Ferry: Oceana Publishers, paper-back ed., 1973), p. 227.Google Scholar

⁹ Truitt, , op. cit.Google Scholar

Weston, Burns H., “Constructive Takings Under International Law: A Modest Foray into the Problem of ‘Creeping Expropriation,’” Virginia Journal of International Law 16(1975): 103–175. See particularly page 105.Google Scholar

¹⁰ A crucial question is why a government would choose the latter. In the second phase of this study, it is suggested that host country technological, managerial, and administrative capabilities are important determinants of the choice between regulation and expropriation. See Kobrin, , op. cit.Google Scholar

¹¹ See Economic and Social Council, op. cit. Hawkins, , Mintz, , and Provissiero, , op. citGoogle Scholar. and Bureau of Intelligence and Research, “Nationalization, Expropriation and Other Takings of United States and Certain Other Foreign Property” (Washington: Department of State, 1971).Google Scholar

¹² See Robinson, Richard D., International Business Management, Second ed. (Hinsdale, Illinois: The Dryden Press, 1978), p. 412 for an excellent discussion of these terms.Google Scholar

¹³ An example would be the expropriation of the Venezuelan subsidiary of Dow-Corning in 1975 in retaliation for that company's cooperating with the kidnappers of one of its managers, contrary to the government's instructions.

¹⁴ Raymond, Vernon, Sovereignty At Bay (New York: Basic Books, 1971), p. 47Google Scholar. For a review of the concept of the obsolescing bargain see, C. Fred, Bergsten, Thomas, Horst and Moran, Theodore H., American Multinationals and American Interests (Washington: The Brookings Institution, 1978)Google Scholar; Smith, David N. and Wells, Louis T. Jr., Negotiating Third World Mineral Agreements: Promises as Prologue (Cambridge: Ballinger, 1975)Google Scholar and Raymond, Vernon, Storm Over the Multinationals (Cambridge: Harvard University Press, 1977).Google Scholar

¹⁵ Nye, Joseph S. Jr., “Multinational Corporations in World Politics,” Foreign Affairs 53 (1974): p. 154.CrossRefGoogle Scholar

¹⁶ Moran, , op. cit., p. 164.Google Scholar

¹⁷ Bergsten, , Horst, and Moran, , op. cit.Google Scholar

¹⁸ As noted above, it is also quite obviously a function of a number of host country characteristics such as relative development and human resource capabilities. The second phase of the study posits that the propensity to force divestment is a function of pressures on the host government to exercise control over economic actors, perceptions of foreign dominance, and local administrative, technical and managerial capabilities. See, Kobrin, , op. cit.Google Scholar

¹⁹ These problems are obviously exacerbated by the structure of most natural resource-based industries which, with the exception of some agricultural products, take the form of vertically integrated transnational oligopolies. See, Moran, , op. cit., for a good discussion of the relationship between industry structure and national control in the copper industry.Google Scholar

²⁰ To analyze the relative importance of FDI, either in an entire economy or in a single sector, requires using the country as a unit of analysis.

²¹ FDI theory posits that foreign investors need some sort of advantage vis-à-vis local competitors to offset their disadvantages of distance and unfamiliarity. These advantages, which may flow from technology, vertical integration or the ability to promote branded products, need to be such that they can be contained, and thus provide sources of rent over time. Thus, given the inherent oligopolistic nature of FDI, barriers to entry may remain after the sources of advantages (e.g. technology or managerial skills) have—or should have—dissipated. See, Kindleberger, Charles P., American Business Abroad (New Haven: Yale University Press, 1969) for a good introduction to this topic.Google Scholar

²² With the exception of the petroleum sector, industry was defined on the basis of a three-digit SIC code.

²³ Truitt, , op. cit., p. 55, suggested the “occasion of taking” as the basic unit of analysis.Google Scholar

²⁴ Economic and Social Council, op. cit., Annex 2.

²⁵ The variables included in the data base will be discussed in detail below. Missing data was minimal for the host country, year of taking, sector and number of firms per act. Coverage was considerably less complete for data on the mode and reason for taking (15 and 51 percent of cases were missing), ownership structure (27 percent), disposition after taking (18 percent) and the nationality of the investor (14 percent). Thus, findings involving the second set of variables must be taken as considerably more tentative than those for the first.

²⁶ Most of the “returns” were in Indonesia, Argentina, and Chile after the deposition of the Sukarno, Peron, and Allende regimes respectively.

²⁷ This is an approximation since the total number of firms taken was not available for a few of the acts. 1521 firms are known to have been taken in 497 acts. For the fourteen acts in which the number of firms is unknown, we have assumed (conservatively) the median and mode of one firm/act.

²⁸ The estimate was made by adjusting the 7778 subsidiaries of U.S. firms reported in the Department of Commerce's benchmark study of direct investment abroad in 1966 for the U.S. share of total FDI in 1967 and the increase in FDI in the LDCs from 1967 to 1976. Sources are, OECD, Stock of Private Foreign Investment by D.A. C. Countries in Developing Countries, Year-End 1967 (Paris: OECD, 1972)Google Scholar; United Nations, Multinational Corporations in World Development (New York: United Nations, 1973)Google Scholar; United Nations, Transnational Corporations in World Development: A Re-examination (New York: United Nations, 1978)Google Scholar. A comparison with previous estimates reinforces the tentativeness of findings. Williams found that the book value of expropriations in the LDCs from 1967–72 amounted to 7.8 percent of end of period stocks plus expropriated assets. (His estimate for 1956–1972 is 18.8 percent; however, the socialization of the Cuban economy is included.) Hufbauer and Briggs estimate gross expropriations of U.S. property in the LDCs from 1965–70 at 3.2 percent of the total value of U.S. FDI. See Hufbauer, G. C. and Briggs, P. H., “Expropriation Losses and Tax Policy,” Harvard International Law Journal 16 (Summer 1975): 533–564Google Scholar, and Williams, M. L., “Extent and Significance of Nationalization of Foreign Owned Assets in Developing Countries, 1956–1972,” Oxford Economic Papers (1975): 260–273.CrossRefGoogle Scholar

²⁹ At this point we cannot judge whether the apparent decline in takings in 1976 (38 versus 83 in 1975) reflects a real phenomenon or simply a lag in reporting by the secondary sources used to compile the data base.

³⁰ See Bergsten, , op. cit., p. 136.Google Scholar

³¹ For example, see Weekley, James K., “Expropriation of U.S. Multinational Investments,” MSV Business Topics 25 (Winter 1977): 27–36.Google Scholar

³² Op. cit., p. 11.

³³ Algeria, Angola, Chile, Ethiopia, Indonesia, Mozambique, Peru, Tanzania.

³⁴ See Bradley, , op. cit., p. 76Google Scholar; Ingram, , op. cit., p. 81Google Scholar; Knudsen, , op. cit., p. 213Google Scholar; Ray, Dennis M., “The Causes of Expropriation of American Property Abroad,” Stanford Journal of International Studies 11 (Spring 1976): 139Google Scholar; Truitt, , op. cit., p. 14.Google Scholar

³⁵ OECD, op. cit.

³⁶ Although both the OECD and the Commerce data are disaggregated by both host country and sector, one still must assume that all firms are evenly distributed about the average firm size when one adjusts for the increase in U.S. FDI from 1966 and 1967. The problem becomes more acute as one must assume the average firm size by sector is constant across all investor countries when expanding the estimate of number of firms to all OECD countries using U.S. share data based upon book value. Furthermore, as almost 75 percent of all acts occurred after 1970, we would prefer to compare the sectoral distribution of takings with that of the stock of FDI in the early 1970s. The available data (which is considerably less detailed than the 1967 OECD study) suggests a slight shift into manufacturing and petroleum and out of mining between 1967 and 1972 (U.N., 1978, op. cit., p. 260). Based on both OECD and U.S. data, we suspect that the sectoral distribution of FDI in 1967 understates the importance of manufacturing (slightly), and banking and insurance (more considerably), and overstates the importance of the extractive sector (moderately) vis-à-vis the actual situation in the early 1970s.

³⁷ There is also obviously a considerable variation in the value of assets taken per act by sector. According to the U.S. Department of Commerce's 1966 benchmark study, the average U.S. petroleum and mining firm was 5–6 times as large as that in manufacturing. (Reported in U.N. 1973, op. cit., p. 141.)

³⁸ Hawkins, , Mintz, , and Provissiero, , op. cit., analyzing 170 foreign takeovers of U.S. firms from 1946–1960, reported the following sectoral distribution:Google Scholar

³⁹ Hawkins, , Mintz, , and Provissiero, ,Google Scholaribid., report that utilities accounted for 50 percent of all expropriations of U.S. firms over 1946–1960. There were also a significant number of insurance expropriations before 1960, indeed, well before World War II.

⁴⁰ The difference does not reflect differences in the distribution of FDI by class of taker. Manufacturing and trade accounted for approximately 16 percent of all FDI in the five countries classed as heavy takers as of year-end 1967.

⁴¹ The five heavy takers—Algeria, Chile, Ethiopia, Peru and Tanzania—each took firms in between 6 and 8 sectors in a relatively concentrated time period.

⁴² Hawkins, , Mintz, , and Provissiero, , op. cit.Google Scholar, and Truitt, , op. cit.Google Scholar

⁴³ Hawkins, , Mintz, , and Provissiero, reported an increase in the manufacturing sector's share from 27 percent in 1961–1971 to 40 percent in 1972–1973 and a decline in the extractive sector's share from 50 percent in 1961–1966 to 39 percent in 1967–1971 and 37 percent in 1972–1973. The fact that Hawkins' data base included only U.S. firms does not account for the difference. A tabulation of only those acts in which one or more U.S. firms are involved shows the manufacturing sector accounted for 24 percent of all of those takings in 1960–1969 and 26 percent in 1970–1976. Op. cit., p. 9.Google Scholar

⁴⁴ According to Business Week's 1977 Survey of R & D spending, the Food and Beverages and Tobacco industries spend 0.5 percent of sales on R & D versus an all-industry average of 1.9 percent (Business Week 3 July 1978, p. 58).

⁴⁵ According to Business Week's 1977 survey, the drug industry spends 4.9 percent of sales on R & D and the chemical industry 2.5 percent compared to an all-industry average of 1.9 percent (Business Week 3 July 1978, p. 58).

⁴⁶ Surprisingly, the findings are not changed when one disaggregates by class of taker. In fact, there are few significant differences between the distribution of manufacturing acts taken by mass and selective expropriators. Specifically comparing heavy “takers” (twenty-one or more acts over the seventeen-year period) with the remainder of the countries one finds significant differences in only three industries. All of the tobacco and machinery acts were taken by countries classed as more selective (twenty or less acts) and all of the acts in the pulp and paper industry occurred in countries classed as heavy “takers.”

⁴⁷ For example, Stopford and Wells note “… many governments seem to assume that they can more easily influence a local partner than a foreign owner.” See, Stopford, John M. and Wells, Louis T. Jr. Managing the Multinational Enterprise (New York: Basic Books, 1971), p. 166.Google Scholar

⁴⁸ In the original coding process each value of equity and ownership structure (i.e., majority J.V.) was regarded as a separate variable and assigned a value equal to the percentage of firms in the act taking that form. However, analysis revealed that the vast majority of acts were homogeneous, i.e., the value of each variable was either 0 or 100 percent in a minimum of 96 percent of the acts. Thus, two categorical variables were created assigning each act to a response if the value of the appropriate variable was equal to 75 percent or more and no “competing” variable equalled 74 percent. Only 2.5 percent of the acts were lost through ambiguity.

⁴⁹ As the coverage of equity was greater than that of ownership, the totals (i.e., the row and column marginals) may at times be greater than the sum of the parts (i.e., the row or column cells). See Table 6.

⁵⁰ See Bradley, , op. cit.Google Scholar and Truitt, , op. cit.Google Scholar

⁵¹ Curhan, Joan P., Davidson, William H., and Rajan, Suri, Tracing the Multinationals (Cambridge: Ballinger Publishing, 1977).Google Scholar

⁵² It should be noted that contract negotiations coded emcompass only a subset of all such renegotiations: those leading to forced divestment of ownership.

⁵³ Interventions (and contract renegotiations) are more likely to be associated with selective takeovers than either nationalizations or forced sales. While we only have data for 47 percent of acts and results are thus very tentative, a cross tabulation of the type and reason for taking reveals that 42 percent of interventions (and 45 percent of contract renegotiations) are not associated with either socialization of the economy or nationalization of an industry as opposed to only 22 percent of nationalizations and 33 percent of forced sales.

⁵⁴ Our findings diverge from those of Hawkins, Mintz, and Provissiero (op. cit.). They found a relatively higher proportion of nationalizations (60 versus 50 percent) and interventions (16 versus 9 percent) and a significantly lower share of forced sales (11 versus 32 percent). The discrepancy is not accounted for by differences in the sectoral distribution of takings.

⁵⁵ The findings of the second phase of the study are consistent with this hypothesis. Again, abstracting from a relatively small number of mass expropriations, the propensity to force divestment is a function of pressures for local control over the economy resulting from the process of modernization, the relative importance of FDI, and the level of host country human resource capabilities. See Kobrin, , op. cit.Google Scholar

⁵⁶ See Kobrin, Stephen J., “When Does Political Instability Result in Increased Investment Risk?,” Columbia Journal of World Business 12 (Fall 1978): 113–122Google Scholar and Kobrin, Stephen J., “Political Risk: A Review and Reconsideration,” Journal of International Business Studies (Spring 1979).CrossRefGoogle Scholar