1 It goes without saying that politics is envisaged, here, at the macro‐level and that by ‘decision’ I intend an enactment.

2 Lest I should be accused of oversimplification, for the complexities on which I cannot dwell see Sartori, G., ‘What is Politics?’, Political Theory, 02 1973, pp. 5–26.CrossRefGoogle Scholar

3 For instance, in Buchanan, J. and Tullock, G., The Calculus of Consent, University of Michigan Press, 1965 Google Scholar, ‘costs of decision‐making’ and ‘external costs’ play a major role. While I am greatly indebted to their fundamental work, it will be seen that my argument largely departs from their premises. Compare, e.g. their ‘external‐costs’ function (pp. 63–8) with my ‘external risks’. Also, Buchanan and Tullock do not distinguish between ‘collective’ and ‘collectivized’ decisions.

4 For the simplicity of the argument the variable is so defined as to imply that the deciding body is a group. But the point can be extended to the dictatorial hypothesis, to the ‘rule of the one’. Depending on whether the single ruler acquires office by heredity, election, force, conquest, lot (anyone), lot‐and‐rotation (anyone in turn), indefinitely or temporarily, it can be easily shown that the external risks vary.

5 Note that ‘representative group’ draws from, and points to, the sociological theory of representation (‘representativeness’), whereas I am exclusively concerned with the theory and practice of political representation. Throughout this article, therefore, ‘representation’ is conceived as ‘responsiveness and ‘responsibility to’. See my article ‘Representational Systems’ in The International Encyclopedia of the Social Sciences, Free Press & MacMillan, 1968, vol. XIII, to which I must also refer for many other facets and complexities of the institution.

6 Remember that, in outcome, external risks turn out to be either costs (damages) or benefits. Analogously, the outcome of decisions can be declared damaging and beneficial—provided that we make clear for whom (for the decision makers or for third parties). The argument is not pursued for lack of space.

7 Reference is made to Schmitt, Carl, whose major writings are now collected in Le Categorie del Politico, Bologna, Il Mulino, 1972 Google Scholar. Class war also enters this perception of politics.

8 I no longer say ‘diverse’ but ‘contrary’ preferences, because majority rule turns a gamut, or a spectrum, into a yes‐no structure of alternatives. We thus have an instrument of conflict resolution which is not conflict‐minimizing.

9 Majority rules are actually referred to four confusingly classified majorities. A qualified majority is first contrasted to a simple majority (to mean that it is not qualified). In turn a simple majority is absolute if referred to the universe, and relative if based on those who are actually present or voting. But then we also use relative majority for whatever majority (whoever first passes the post).

10 Other characteristics of ‘group’, such as shared goals or a sense of belonging, are immaterial to a minimal definition of ‘committee’.

11 It is only obvious that these considerations apply to any large organization. The argument is confined to political systems for the sake of brevity.

12 According to Buchanan and Tullock op. cit., ‘if full side payments are allowed to take place, any decision‐making rule for collective action will lead to positions that may be properly classified as Pareto‐optimal’ (p. 189). It is appropriate to note once again that the the route followed here leads to very different conceptualizations and conclusions.

13 I switch from ‘small group’ to ‘committee’ on the assumption that what is at stake is participation in decision making, i.e. in those groups entrusted with a flow of decisions which afford the highest participatory efficacy.

14 While a literally self‐governing democracy can be said to coincide with ‘true’ direct democracy, it should be recalled that I have given earlier a narrower definition of the latter (in terms of observability).

15 Democratic Theory, Wayne University Press, 1962 (and Praeger 1967), esp. chaps. 2, 5 and 10. Here I treat the problem of self‐government as an inverse relationship between the connotation and the denotation of the term: ‘The intensity of self‐government attainable is in inverse proportion to the extension of the self‐government demanded’ (p. 60 ff.). This formulation is somewhat more flexible, but adds up to the same conclusion.

16 My distinction bears only on the extremes of the size continuum. Micro‐democracy cannot extend beyond the size of the polis, of the small city. Conversely, by macro‐democracy I intend the state level of our territorially dispersed, large scale democracies. In most other respects I would agree with Danl and Tufte that it is very difficult to establish relations between size and democracy (Size and Democracy, Stanford University Press, 1973).

17 Compare the sterility and monotony of the debate on how to maximize the power of the people, to the vitality and development of the ‘politics of allocation’ literature, as testified, in one of its recent variants, by the ‘Rawls principle’ (see Rawls, John, A Theory of Justice, Harvard University Press, 1971 Google Scholar; the overall appraisal of Barry, Brian, The Liberal Theory of Justice, Oxford, Clarendon Press, 1973 Google Scholar; and the follow‐up, ‘Symposium on Rawls’, American Political Science Review, forthcoming).

18 A solution is Pareto‐optimal when at least one person is better off and nobody else is worse off. A solution is Rawls‐preferred, instead, when it gives as much as possible to those who have least.

19 This point is probed in detail, with reference to the distinction between ‘predominant’ and ‘countervailing’ majorities, by R. D’Alimonte, ‘Regola di Maggioranza, Stabilità e Equidistribuzione’, Rivista Italiana di Scienza Politica, I, 1974, sect. 3.

20 This conclusion entails that I do not subscribe to the guideline suggested by Buchanan and Tullock: ‘the logrolling process [full side‐payments] provides the general model of analyzing the various choice‐making rules’ (op. cit., p. 123).

21 I neglect other identifiable units—such as small groups in general, or occasional (non‐institutionalized) assemblies—because the former are not amenable to any precise operational code, and the latter are utterly rule‐less.

22 To this effect Lindblom, C. E., The Intelligence of Democracy: Decision Making Through Mutual Adjustment, Free Press, 1965 Google Scholar, remains the unsurpassed analysis.

23 On the decay of constitutionalism as an intransitive, power‐checking and power‐limiting structure, see Sartori, G., ‘Constitutionalism: A Preliminary Discussion’, American Political Science Review, 12 1962 CrossRefGoogle Scholar; and, more in general, Democratic Theory, chaps. 13 and 15 (esp. pp. 288–91, 306–13, 372–4). On the multiplication of the potentialities of power entailed by technology, my concerns are stated in ‘Technological Forecasting and Polities’, Survey, Winter 1971, esp. pp. 63–8, and, extensively, in ‘The Power of Labour in the Post‐Pacified Society’, published in Italian in Rivista Italiana di Scienza Politico, 1, 1973.

24 See Etzioni, Amitai, Demonstration Democracy, Gordon and Breach, 1970 Google Scholar.