We are indebted to Wedge Greene, Gary Gaines, and James Horney for helpful research assistance. Preparation of this paper was supported in part by grants from the Institute for World Order. We have benefitted from the comments of Lawrence Finkelstein, Ernst Haas, Robert Holt, Henry Nau, Joe Oppenheimer and anonymous referees. An earlier version of this paper was presented at the meetings of the International Political Science Association, Edinburgh, Scotland, 16–21 August 1976.

¹ Halbouty, Michael T., “Future Trends in Exploration to Discover New Hydrocarbon Reserves Throughout the World and the Role to be Played by OPEC Member Countries and National Oil Companies,” paper presented at the OPEC-Seminar on “The Present and Future Role of the National Oil Companies,” Vienna, 10 10 1977Google Scholar.

² The members of the OECD are: Australia, Austria, Belgium, Canada, Denmark, Finland, France, the Federal Republic of Germany, Greece, Iceland, Ireland, Italy, Japan, Luxembourg, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, the United Kingdom, and the United States.

³ See Keohane, Robert O., “The International Energy Agency: State Influence and Transgovernmental Politics,” International Organization 32, 4 (Autumn 1978)CrossRefGoogle Scholar. Keohane provides a rare inside look into the OECD's energy arm.

⁴ The collective goods literature is now well known to most international relations scholars. The first complete exposition in English is Samuelson, Paul A., “The Pure Theory of Public Expenditure,” Review of Economics and Statistics (11 1954): 387–89Google Scholar. The classic discussion for social scientists with varied interests remains Olson, Mancur, The Logic of Collective Action (Cambridge: Harvard University Press, 1965)Google Scholar.

⁵ This characteristic is regarded as more fundamental to collective goods than the other key trait, excludability. As Singer puts it, “The publicness of a public good depends on economics not on technology.… Even though it may be technologically possible to restrict consumption of the public good, economic efficiency requires that it be made available to all consumers at a zero price since the opportunity cost of extending consumption is zero.” Singer, Neil, Public Microeconomics (Boston: Little, Brown Company, 1972), p. 95Google Scholar. The best discussion of the significance of both characteristics is Head, John G., “Public Goods and Public Policy,” Public Finance 3 (1962): 197–219Google Scholar.

⁶ Tilton, John E., U.S. Energy R&D Policy: The Role of Economics (Baltimore: The Johns Hopkins University Press, 1974), pp. 29–30Google Scholar, stresses the multiple industry impact. Tilton also mentions human capital spillovers resulting from learning curve effects as a developmental externality justifying public support. For a thorough recent discussion of R&D in terms of the “appropriability” problem, see Magee, S. P., “Information and the Multinational Corporation: An Appropriability Theory of Direct Foreign Investment” in The New International Economic Order, Bhagwati, Jagdish, ed. (Cambridge: The MIT Press, 1977), pp. 317–340Google Scholar.

⁷ Statement by Nelson F. Sievering, Jr., Assistant Administrator for International Affairs, Energy Research and Development Administration before the Environment, Energy and Natural Resources Subcommittee of the House Committee on Government Operations, 12 May 1977, mimeo; Olson, Mancur Jr, and Zeckhauser, Richard, “The Efficient Production of External Economies,” American Economic Review (06 1970): 512–516Google Scholar. For a discussion which stresses the advantages of independence, ceteris paribus, see Ruggie, John Gerard, “Collective Goods and the Future of International Collaboration,” American Political Science Review (09 1972): 874–893Google Scholar. Our discussion of R&D can be reconciled with Ruggie's if it is understood that R&D capacity is a private, national good, while its results are a public good.

⁸ Holloman, J. Herbert and Grenon, Michel, Energy Research and Development (Cambridge: Ballinger, 1975), p. 129Google Scholar. Also see: Prodi, Romano and Clo, Alberto, “Europe,” Daedalus 104, 4 (Fall 1975): 91–5Google Scholar; Energy Policies in the European Community, U.S. Energy Research and Development Administration (02 1975), ERDA–51Google Scholar.

⁹ Ibid.; Walsh, John, “European Community Energy Policy,” Science 184, 4142 (14 06 1974): 1158–1161CrossRefGoogle ScholarPubMed.

¹⁰ Nau, Henry R., “Collective Responses to R&D Problems in Western Europe,” International Organization 29 (Summer 1975): 617–53CrossRefGoogle Scholar, and the estimate by one high General Electric R&D manager in Hearings before the Subcommittee on Energy Research, Development, and Demonstration and the Subcommittee on Domestic and International Scientific Planning and Analysis of the Committee on Science and Technology, U.S. House of Representatives, 94th Congress, 2nd Session, (Washington: U.S.G.P.O., 1976), p. 127.

¹¹ Quoted in Lester, James P., “Energy R&D: U.S. Technology Transfer to Advanced Western Countries,” Graduate Program in Science, Technology, and Public Policy, The George Washington University, Washington, 11, 1975, p. 14Google Scholar.

¹² Pollock, Herman and Congdon, Michael B., “International Cooperation in Energy Research and Development,” Law and Policy in International Business (06/08 1974): 686–688Google Scholar.

¹³ Nau, , “Collective Responses,” p. 638Google Scholar.

¹⁴ Comparison of these data from a U.S. government source with those provided by OECD change specific numbers and add cooperative activities involving other members of OECD, but the same general implications are there; Organization for Economic Cooperation and Development, Energy R&D (Paris: OECD, 1975), pp. 115–9Google Scholar.

¹⁵ Ibid., pp. 155–56.

¹⁶ Olson, , Logic of Collective Action, pp. 30–31Google Scholar.

¹⁷ OECD, Energy R&D, p. 157Google Scholar, and Holloman, and Grenon, , Energy Research and Development, pp. 113–133Google Scholar.

¹⁸ Prodi, and Clo, , “Europe,” pp. 91–95Google Scholar.

¹⁹ Pollock, and Congdon, , “International Cooperation,” p. 687Google Scholar.

²⁰ OECD, Energy R&D, p. 158Google Scholar; see also Nau, Henry R., National Politics and International Technology (Baltimore: Johns Hopkins University Press, 1974)Google Scholar; Donnelly, Warren H., Commercial Nuclear Power in Europe, prepared for the Subcommittee on National Security Policy and Scientific Developments, Committee on Foreign Affairs, U.S. House of Representatives, 92nd Congress, 2nd Session, 12, 1972Google Scholar; Gilpin, Robert, France in the Age of the Scientific State (Princeton University Press, 1968)CrossRefGoogle Scholar.

²¹ Nau, , “Collective Responses,” and Baker, Steven J., “Monopoly or Cartel?” Foreign Policy 23 (Summer 1976): 207–208Google Scholar.

²² Holloman, and Grenon, , Energy Research and Development, p. 136Google Scholar; on the severe need of U.S. reactor vendors for foreign orders, see Jaskow, Paul L. and Baughman, Martin L., “The Future of the U.S. Nuclear Energy Industry,” Bell Journal of Economics 7, 1 (Spring 1976): 3–32CrossRefGoogle Scholar.

²³ OECD, Energy R&D, p. 158Google Scholar; Energy Research and Development Administration, A National Plan for Energy Research, Development and Demonstration, Vol. I, The Plan, ERDA–48 (Washington: U.S.G.P.O., 1975), p. VI–2Google Scholar; Nau, National Politics.

²⁴ Gall, Norman, “Atoms for Brazil, Dangers for All,” Foreign Policy 23 (Summer 1976): 157, 165Google Scholar.

²⁵ Simonet, Henri, “Energy and the Future of Europe,” Foreign Affairs 53, 3 (04 1975): 450–63CrossRefGoogle Scholar.

²⁶ This section is based on public documentary sources. Inferences have been checked and enriched through off-the-record interviews of a number of former and current U.S. officials with responsibilities for energy policy and analysis of international energy matters.

²⁷ Hammond, Allen L., “Nuclear Proliferation,” Science 193 (9 07 1976): 125–130CrossRefGoogle ScholarPubMed; for a thorough treatment of the other major joint effect of the two events see Bloomfield, Lincoln P., “Nuclear Spread and World Order,” Foreign Affairs 53, 4 (07 1975): 743–55CrossRefGoogle Scholar.

²⁸ International Herald Tribune, 15–16 July 1978, p. 4.

²⁹ On endowment differences, see: Energy Policies in the European Community. One European energy official remarked of his colleagues from countries with substantial oil and natural gas that “lue-eyed sheikhs are as bad as brown-eyed ones,” The Economist, 7 september 1974, p. 64; Choucri, Nazli, The International Politics of Energy Interdependence (Lexington: D. C. Heath, Lexington Books, 1976)Google Scholar .

³⁰ These requisites are developed from those in Energy Policies in the European Community, pp. 92–3.

³¹ Sievering, , “Statement,” p. 7Google Scholar.

³² For the voting rules, see: Selected Readings on International Cooperation in Energy Research and Development, prepared by the Subcommittee on Domestic and International Scientific Planning and Analysis and the Subcommittee on Energy Research, Development, and Demonstration of the Committee on Science and Technology, U.S. House of Representatives, 94th Congress, 2nd Session (Washington: U.S.G.P.O., 1976), pp. 186–8Google Scholar.

³³ Keohane, “International Energy Agency,” passim.

³⁴ Quoted in Department of State Bulletin, 4 March 1974.

³⁵ Hearings, pp. 13–14; A Time to Choose, Final Report by the Energy Policy Project of the Ford Foundation (Cambridge: Ballinger, 1974), pp. 157–67Google Scholar; Arab, Uzi B., “Is the Atlantic Alliance Soluble in Oil?” in The Energy Crisis and U.S. Foreign Policy, Szyliowicz, Joseph S. and O'Neill, Bard E., eds. (New York: Praeger, 1975), pp. 150–1Google Scholar.

³⁶ On the rationale for the charge, see: Oppenheim, V. H., “The Past, We Pushed Them,” Foreign Policy (Winter 1976): 24–57Google Scholar; Washington Post, 10 July 1977, pp. Al ff.

³⁷ Kissinger, is quoted in the New York Times, 12 02 1974Google Scholar. For a number of his major statements including comments on energy R&D, see: Readings, pp. 115–18, 136–43. Also see: Federal Energy Administration, Project Independence (Washington: U.S. G.P.O., 1974)Google Scholar.

³⁸ U.S. Department of State, International Cooperation in Energy R&D, A Report to Gov. John A. Love, Director, Energy Policy Office, from William J. Casey, Under-Secretary of State, Economic Affairs, summarized in Pollock and Congdon, “International Cooperation.”

³⁹ Poole, James, “The Next World Oil Crisis: American Energy Policies Are Seen as a Global Menace,” Sunday Times (London)Google Scholar, reprinted in Atlas, July 1976, pp. 10–14; for a useful review of the European disarray, see Robert J. Lieber, Oil and the Middle East War (Center for International Affairs, Harvard University, 1976). See also Katzenstein, Peter J., “International Relations and Domestic Structures,” International Organization 30, 1 (Winter 1976): 23–44CrossRefGoogle Scholar.

⁴⁰ Energy Policies in the European Community, p. 38.

⁴¹ Hammarlund, Jeffrey R. and Lindberg, Leon L., eds., The Political Economy of Energy Policy: A Projection for Capitalist Society (Institute for Environmental Studies, University of Wisconsin-Madison, 12 1976)Google Scholar; IES Report 70; and Tsurumi, Yoshi, “Japan,” Daedalus 104, 4 (Fall 1975): 113–28Google Scholar.

⁴² Pollock and Congdon, “International Cooperation”; Lester, “Energy R&D”; OECD, Energy R&D.

⁴³ Holloman, and Grenon, , Energy Research and Development, p. 130Google Scholar; Energy Research and Development Administration, A National Plan for Energy Research, Development and Demonstration, Vol. I, The Plan, 1975, ERDA–48, p. C–3Google Scholar.

⁴⁴ Enders, Thomas O. quoted in New York Times, 31 01 1976, p. 37Google Scholar.

⁴⁵ Readings, p. 63; Energy Research and Development Agency, A National Plan for Energy Research, Development and Demonstration: Creating Energy Choices for the Future, 1976, Vols. 1, 2, 1976Google Scholar; Sievering, , “Statement,” p. 7Google Scholar.

⁴⁶ Readings, pp. 73, 83–5; Hearings, pp. 16, 77–8, 141.

⁴⁷ Suboptimality can be avoided only if one country's demand dominates all others. See Olson, , Logic of Collective Action, pp. 28 ffGoogle Scholar. and the final section of this paper.

In a stimulating piece, Oppenheimer has argued that the political process is so fraught with difficulties as an aggregating agent that it would be surprising to find the predictions of the theory of collective goods actually borne out. Joe A. Oppenheimer, “Reflections on the Applicability of the Theory of Collective Goods to International Politics,” presented to the 1978 meetings of the International Studies Association. Our view is less pessimistic, and we think the results of this paper provide some justification for it. For us, using collective goods theory to predict disproportionate cost-sharing is rather like using the theory of cost-benefit to predict which of a set of possible bridge-sitings will actually be acted upon: it is an imprecise device but it will not err randomly across a broad range of situations.

⁴ This expectation is familiar to students of industrial competition where a failure of tacit collusion often leads to excessive advertising in an industry by comparison with the joint profit-maximizing level. See Scherer, F. M., Industrial Market Structure and Economic Performance (Chicago: Rand McNally & Company, 1970), p. 336Google Scholar.

Also in this connection, some recent literature implies that insofar as it is conducted by (or perhaps only sponsored by) the government, a more than optimal amount of expenditure will take place. This allegedly results from motivations and opportunities towards overproduction suggested in works such as William Niskanen's Bureaucracy and Representative Government. Although some appear to believe that an overproduction bias is firmly established in such work, most critical readers remain unconvinced. As Earl Thompson has pointed out in a review of Niskanen's book, “ … bureaus want tremendously high outputs to increase their factor rents, but … government representatives, who are aware of the desires of the bureaucrats and are disciplined by political forces involving majority rule voting, keep these desires in check, while existing voting models do not give as much of a hint as to whether the outputs determined by the governmental representatives are too large or too small.” Journal of Economic Literature (September 1973): 952. Indeed, some years earlier Anthony Downs put forward a view of sub optimal spending by government, with goods like R&D, which produce “Benefits … remote from those who receive them, either in time, space, or comprehensibility,” subject to especially great short-falls. Downs, Anthony, “Why the Government Budget is Too Small in a Democracy,” World Politics (07 1960)Google Scholar. Given the strong biases on both sides of the bargaining table on an issue like R&D, the direction of distortion simply cannot be known, but in our view it is unlikely to be systematically large one way or the other across countries and R&D sectors (taking into account of course that future benefits will be discounted by voters in the ordinary way); we believe that the burden of proof is solidly with those who would contend that the distortion is systematic or substantial. In terms of the relations between industry and government, pioneering attempts to model the impact of the structure of an industry on its pressure to provide either “sub-optimal” or “super-optimal” amounts of R&D have been made by Takashi Negishi. See his “The Excess of Public Expenditures on Industries,” Journal of Public Economics 2 (1973): 231–240CrossRefGoogle Scholar, and “Excess Burden and The Voluntary Theory of Public Finance: The Case of Intermediate Public Goods,” Economka (May 1975): 177–81. The assumptions in this work are much too strict to be directly useful in the present paper.

⁴⁹ TheOLS coefficient for GNP is 1.16; t = 26.49.

One of the points stressed most heavily by Oppenheimer in the paper cited in footnote 47 is the extent to which the valuation of a public good is dependent upon income distribution both within and among countries. As is implicit in our discussion, we think an optimal level of expenditure would probably be quite robust with respect to changes in distribution both internally and internationally. The reason is simply that energy is heavily involved in the delivery of goods and services in exchange for dollar votes—virtually regardless of who casts them within the richer countries.

⁵⁰ The issue was apparently first raised in the international relations literature by Russett, Bruce and Sullivan, John, “Collective Goods and International Organization,” International Organization (Autumn 1971): 845–865Google Scholar, where they contend in a footnote that their use of ratio correlation serves “in no way to inflate the apparent relationship between the variables …”, p. 862. No mention is made of the fact that their procedure is biased away from confirming their hypothesis. Although the understanding of the direction of bias goes back as far as Karl Pearson in 1897, it has been recently brought to the attention of political scientists in a dispute between Uslaner, and Lyons, in the American Journal of Political Science in 1976 and 1977 (numbers 20, 21)Google Scholar. A clear statement is Vanderbok, William, “On Improving the Analysis of Ratio Data,” Political Methodology 4, 2 (1977)Google Scholar. Vanderbok gives the expected measured Pearsonian correlation for different relative coefficients of variation for the numerator and denominator of the ratio when the ratio and one of its elements are actually independent.

We have checked the literature carefully and have consulted with several statisticians who have worked on the ratio data problem. It appears that the exact bias involved in rank correlation data has not yet been established. Furthermore, the test of significance for Kendall's partial tau is undetermined. Because of the more precise results concerning bias and significance for partial correlation coefficients which appear to be offered by the use of the ordinary Pearsonian product moment correlation coefficients, some scholars might be tempted to turn directly to them in an attempt to test public goods hypotheses. In our experience, this is most unwise. The following table shows the relation between NATO expenditure as a percentage of GNP and GNP for 1964 as examined by Olson and Zeckhauser (Olson, Mancur Jr, and Zeckhauser, Richard, “An Economic Theory of Alliances,” Review of Economics and Statistics (11 1966): 266–279)Google Scholar; the relationship between total R&D expenditures by 17 OECD countries in the mid-sixties as a percent of GNP and GNP examined by the writers in preparation of the present paper (based on data in OECD, The Level and Structure of R&D Efforts in OECD Member Countries, [Paris, 1967])Google Scholar and the data presented in Table 7. The reader can confirm that the relative strength of the relation (ignoring bias) varies among the three public goods depending on which measure is used, and, in particular, that the simple correlation coefficient for the energy R&D percentage to GNP relation is dramatically lower than the rank order relation. What this suggests is that this relation is monotonic but otherwise very “badly behaved.”

⁵¹ The reader is reminded that simply finding a negative correlation between per capita income and percent of GNP spent (in our case –.067) does not disprove some actual positive relation because the spurious correlation is strongly negative. Thus Olson and Zeckhauser's claim that the relation between per capita income and percent of GNP spent on NATO is negative is incorrect (Olson, and Zeckhauser, , “An Economic Theory,” p. 275)Google Scholar, although their central conclusion about the relationship between GNP and percentage spent on defense is even stronger than they claimagain because of spurious correlation.

⁵² Some countries will naturally concentrate on domestic resources because only they will significantly benefit from successful exploitation; examples are certain geothermal projects in Iceland and parts of Scandinavia. Still others may concentrate on domestic sources for increased price-cost margins of exports or for security. Nevertheless, for many, perhaps most, countries, the desire for cheap energy might be expected to prevail.

⁵³ The private monopoly system is an extremely inefficient instrument for encouraging industrial innovation—except by comparison with the alternatives. For a lucid discussion see Johnson, Harry G., “The Efficiency and Welfare Implications of the International Corporation,” in The International Corporation: A Symposium, Kindleberger, Charles P., ed. (Cambridge: The MIT Press, 1970), pp. p36–39Google Scholar.

⁵⁴ For the fifteen countries for which a breakdown between nuclear and non-nuclear R&D expenditure is available, Kendall's tau on the relation between GNP and percentage spent on R&D is.524 for nuclear spending and.522 for all otherspending. (Data from OECD, Energy R&D.)

⁵⁵ For a discussion of the “national champion strategy,” see Moran, T. H., “MNC's and the Political Economy of Atlantic Relations,” 1977Google Scholar, mimeographed.

⁵⁶ The energy R&D expenditure figure for the U.S. is from Russell, Milton, “Energy” in Setting National Priorities: The 1978 Budget, Pechman, Joseph A., ed. (Washington: The Brookings Institution, 1977)Google Scholar. Furthermore, spending more than doubled within the next two years. After this study had already been edited, more recent figures became available from the OECD for a group of countries quite similar to that presented in Table 5. (Organization for Economic Cooperation and Development, Energy Policies and Programmes oflEA countries, 1977 Review (Paris, 1978)Google Scholar. These data suggest that while most countries increased their energy R&D efforts as a percent of GNP the disproportionality was probably greater than in 1974 (Kendall's tau equals.55 rather than.45 for a similar set of countries).

⁵⁷ Russett and Sullivan, “Collective Goods”; and Olson, “Increasing the Incentives.”

⁵⁸ For an informative exploration, see Hammarlund and Lindberg, Political Economy of Energy Policy.

⁵⁹ This discussion benefits from those of Holloman, and Grenon, , “Energy Research and Development,” pp. 130, 137Google Scholar, and Nau, , “Collective Responses,” p. 652Google Scholar.

⁶⁰ For an early contribution which recognizes the possible role of “means-testing” in the provision of international public goods, see de Strihou, Jacques van Ypersele, “Sharing the Defense Burden Among Western Allies,” Review of Economics and Statistics (11 1967): 527–536Google Scholar.

⁶¹ Kissinger, Henry A., “Cohesion of Industrial Democracies,” Office of Media Services, Bureau of Public Affairs, Department of State (21 06 1976), p. 5Google Scholar.