¹ Truman, , The Governmental Process (New York: Knopf, 1951 and 1971).Google Scholar Reference will be made to the second edition.

² Olson, , The Logic of Collective Action (Cambridge, Mass.: Harvard University Press, 1965 and 1971).Google Scholar Reference will be made to the second edition. Other critiques of group theory include: Macridis, Roy C., “Interest Groups in Comparative Analysis,” Journal of Politics, XXIII (February 1961), 25–45CrossRefGoogle Scholar; Odegard, Peter H., “A Group Basis of Politics: A New Name for an Ancient Myth,” Western Political Quarterly, XI (September 1958), 689–702Google Scholar; Rothman, Stanley, “Systematic Political Theory: Observations on the Group Approach,” American Political Science Review, Vol. 54 (March 1960), 15–33.CrossRefGoogle Scholar

³ Barry, , The Liberal Theory of justice (Oxford: Clarendon Press, 1973), 118.Google Scholar

⁴ Olson (fn. 2), 2; emphasis in original.

⁵ Truman (fn. 1), xxix.

⁶ Hardin, , “Collective Action as an Agreeable n-Prisoners' Dilemma,” Behavioral Science, XVI (September 1971), 472–81.CrossRefGoogle Scholar

⁷ For an explanation of the Prisoners' Dilemma, see section III.

⁸ For a lucid and fuller summary of the characteristics of a public good, see Taylor, Michael, Anarchy and Cooperation (London: Wiley, 1976), 14–15.Google Scholar

⁹ Riker, William H. and Ordeshook, Peter C., An Introduction to Positive Political Theory (Englewood Cliffs, N.J.: Prentice-Hall, 1973), 74.Google Scholar

¹⁰ Olson (fn. 2), 15.

¹¹ Frohlich, Norman and Oppenheimer, Joe A., “I Get By With a Little Help From My Friends,” World Politics, XXIII (October 1970), 109.Google Scholar

¹² In principle, the wider bridge produces a different good, but in practice it would be unlikely to be perceived as being significantly different.

¹³ In this example it is, in principle, possible for the union to negotiate only on behalf of its members and not of the entire category of workers involved. Thus the good is not strictly non-excludable; computerized payrolls that make a distinction between union members and non-members are feasible. However, the distinction seems rarely to be made in practice, presumably because of the extra cost to firms of having to bargain with individual non-members. Thus the possibility of free-riding occurs in practice.

¹⁴ Indeed, some groups may collapse and then reorganize when their interest is threatened again, as with the Wing Airport Resistance Association. Having successfully fought the proposal to build London's third airport at Wing, WARA was wound-up, only to be reactivated when the issue was later reopened. No doubt when the matter is finally settled, WARA will collapse again. For a description of WARA's organization and first campaign, see Kimber, Richard and Richardson, J. J., “The Roskillers,” in Kimber, and Richardson, , eds., Campaigning for the Environment (London: Routledge, 1974), 165–211.Google Scholar

¹⁵ See, for example, Shubik, Martin, ed., Game Theory and Related Approaches to Social Behavior (New York: Wiley, 1964), 38.Google Scholar For a critical evaluation of the conventional view, see Rapoport, Anatol, Fights, Games, and Debates (Ann Arbor: University of Michigan Press, 1960), 173–77.Google Scholar

¹⁶ Hardin (fn. 6), 473–74.

¹⁷ Taylor (fn. 8), 7.

¹⁸ Olson (fn. 2), 9–16.

¹⁹ Ibid., 126; emphasis in original.

²⁰ Ibid., 35.

²¹ Frohlich, Norman, Oppenheimer, Joe A., and Young, Oran R., Political Leadership and Collective Goods (Princeton: Princeton University Press, 1971).Google Scholar

²² Richelson, Jeffrey, “A Note on Collective Goods and the Theory of Political Entrepreneurship,” Public Choice, XVI (Fall 1973), 75.Google Scholar

²³ Olson (fn. 2), 12.

²⁴ With this approach there is a difficult extreme case in which, if the net value is sufficiently small, the increased burden of one individual leaving transforms a small net benefit for the others into a net loss, thus causing them to leave the group. Consider a large group of, say, 10,000 individuals providing a good at a total cost of £999,000. The cost per head is thus £99.90. If the value to each individual is £99.90½, then each individual has a net benefit of ½p; small but perceptible. If one person now leaves, the new cost per head becomes £999,000 divided by 9,999 which is £99.91. Thus, the original net benefit is transformed into an equally small but perceptible net loss, and the group should collapse. This may not be a major objection to the increased-burden approach, but any theory should be able to cope with the extreme cases.

²⁵ Olson (fn. 2), 12.

²⁶ Ibid., 106.

²⁷ See, for example, Salisbury, Robert H., “An Exchange Theory of Interest Groups,” Midwest Journal of Political Science, XIII (February 1969), 1–32.CrossRefGoogle Scholar

²⁸ “Strategy” here refers to the decision at which an individual arrives when he has completed his calculation based on the relevant factors. The societal process envisaged here has some similarity with the ethologists' view in which evolution takes place against a background of an evolutionarily stable set of genes. See Dawkins, Richard, The Selfish Gene (Oxford: Clarendon Press, 1976), 93.Google Scholar

²⁹ Truman (fn. 1), 159.

³⁰ Olson (fn. 2), 133. The process that gives rise to the by-product is unclear. The problem is not so much to explain how large groups come to be organized, but rather to explain how they come to provide collective goods.

³¹ Marsh, , “More on Joining Interest Groups,” British Journal of Political Science, VIII (July 1978), 384.Google Scholar