¹ Consider L. E. Fredman's remark in Fredman, L. E., The Australian Ballot: The Story of an American Reform (East Lansing: Michigan State University Press, 1968), p. 119.Google Scholar ‘The conduct of elections now attracts little attention from political scientists. It is assumed that they are fair and orderly, and an accurate expression of the popular will’. Notice too the equally complacent remark on the secret vote, from fifty years earlier, in Evans, E. C., A History of the Australian Ballot System in the United States (Chicago: University of Chicago Press, 1917), p. 72.Google Scholar ‘It has cleared away the obstacles which formerly prevented a free expression of the public will. It has made good government possible, if the electors really want it.’

² Harsanyi, John, ‘Rational Choice Models of Behavior Versus Functionist and Conformist Theories’, World Politics, 22 (1969), 513–38CrossRefGoogle Scholar; Taylor, Michael, ‘Rationality and Revolutionary Collective Action’, in Taylor, , ed., Rationality and Revolution (Cambridge: Cambridge University Press, 1987).Google Scholar

³ Rawls, John, A Theory of Justice (Oxford: Oxford University Press, 1971).Google Scholar

⁴ Nossiter, T. J., Influence, Opinion and Political Idioms in Reformed England (Brighton: Harvester, 1975).Google Scholar

⁵ When this article already existed in draft, we found that a similar distinction had been drawn recently by others, and in similar terms. See Coleman, Jules and Ferejohn, John, ‘Democracy and Social Choice’, Ethics, 97 (1986), 6–25CrossRefGoogle Scholar; and Cohen, Joshua, ‘An Epistemic Conception of Democracy’, Ethics, 97 (1986), 26–38.CrossRefGoogle Scholar

⁶ For convenience of expression we shall refer to the voter throughout as if the voter were male.

⁷ Although we shall discuss these two ideals under the characterizations offered, it is worth noting that there is a version of the judgement ideal that softens the contrast between the two. According to that version it is best for all, not if everyone votes out of consideration of what is best for all, but rather if everyone votes out of consideration of what is best for himself or for those in some group to which he belongs. On this version of the judgement ideal, each is required to vote, not for the option that he judges to be best for all, but in conformity to a strategy which he judges to have that merit. This version of the judgement ideal softens the contrast with the preference ideal, particularly if it is assumed, as it often is, that a person's preferences are generally self-interested. But nevertheless everything that we say about the judgement ideal in general applies also, although sometimes with obvious modifications, to this version. And so, for convenience of presentation, we shall write as if the judgement ideal only comes in the standard version given above.

⁸ On Bentham and James Mill, and on the protective theory of democracy, see Pateman, Carole, Participation and Democratic Theory (Cambridge: Cambridge University Press, 1970), chap. 1.CrossRefGoogle Scholar On the economic theory of the state see Hamlin, Alan, Ethics, Economics and the State (Brighton: Harvester, 1986).Google Scholar

⁹ See Hospers, John, Libertarianism (Los Angeles, Calif.: Nash, 1971).Google Scholar

¹⁰ To describe the preference model as ‘non-interactive’ does not, of course, deny that voters will have reason to seek information about the consequences of certain policies or candidates, or that they may consult others in that connection.

¹¹ Pettit, Philip, ‘Towards a Social Democratic Theory of the State’, Political Studies, 35 (1987), 42–55.CrossRefGoogle Scholar

¹² Such ‘internal aggregation’ may not entirely obliterate familiar problems of majoritarian cycling and the like, but it seems likely to moderate those problems significantly.

¹³ For a recent treatment of the role of majority rule in ironing out errors, which connects Rousseau's notion of the ‘general will’ to Condorcet jury theorems, see Grofman, Bernard and Feld, Scott, ‘Rousseau's General Will: A Condorcetian Perspective’, American Political Science Review, 82 (1988), 567–76.CrossRefGoogle Scholar It is crucial for the Condorcet logic that the individual judgements be independent. We differ from Rousseau if he assumes, as Grofman and Feld allege, that discussion compromises independence in the sense relevant to Condorcet's theorems. We think that the decentralized discussions we wish to foster would certainly not undermine this independence, at least in a pluralistic society.

¹⁴ Elster, Jon, Sour Grapes (Cambridge: Cambridge University Press, 1983), chap. 2CrossRefGoogle Scholar, and ‘The Market and the Forum’, in Elster, Jon and Hylland, A., eds, Foundations of Social Choice Theory (Cambridge: Cambridge University Press, 1986)Google Scholar, argues that the educative effect can only be attained as a byproduct of pursuing some other end; we mean to take his point. See Pettit, Philip and Brennan, Geoffrey, ‘Restrictive Consequentialism’, Australasian Journal of Philosophy, 64 (1986), 438–55, for the relevant ethical perspective.CrossRefGoogle Scholar

¹⁵ Pateman, , Participation and Democratic Theory, chap. 2.Google Scholar

¹⁶ Braithwaite, John and Pettit, Philip, Not Just Deserts: A Republican Theory of Criminal Justice (Oxford: Oxford University Press, forthcoming).CrossRefGoogle Scholar

¹⁷ Mill, J. S., Considerations on Representative Government (London: Everyman Books, 1964; first published in 1861), p. 305.Google Scholar

¹⁸ Mill, , Considerations on Representative Government, p. 299.Google Scholar

¹⁹ Notice that in this sense of ‘objective’, utilitarianism would supply such an answer.

²⁰ Barry, Brian, Economists, Sociologists and Democracy (Chicago: University of Chicago Press, 1970).Google Scholar

²¹ We argue later that the preference ideal of voting is unrealizable but that the judgement ideal may not be. To that extent we align ourselves with the republican tradition in democratic thinking; we hold that if democracy is to thrive, it must begin to conform to the republican image. But, for the record, we do not think that this means that democracy must be direct rather than representative; on the contrary, we hold by the superiority of a representative regime. The position that we wish to occupy then depends on reopening distinctions which the habits of political theory would have us close.

²² Downs, Anthony, An Economic Theory of Democracy (New York: Harper, 1957), p. 37.Google Scholar

²³ Broome, John, ‘Review of Kenneth Arrow's Collected Papers’, Economic Journal, 95 (1985), 210–11CrossRefGoogle Scholar, highlights some general flaws. Another flaw becomes apparent in the light of the argument in Section II.

²⁴ On related matters see the congenial claims of Sen, Amartya, Choice, Welfare and Measurement (Oxford: Blackwells, 1982), Part 1.Google Scholar

²⁵ The distinction between preference and judgement ideals is similar to that drawn between the ‘market’ and ‘forum’ conceptions of democratic process that Elster discusses in his chapter entitled ‘The Market and the Forum’ in Elster, Jon and Hylland, A., eds, Foundations of Social Choice Theory.Google Scholar Moreover, the arguments that underlie the forum notion, and in particular Habermas's belief that the forum context encourages the expression of public judgements, are somewhat similar to our own concerning discursive defensibility. Elster offers several criticisms of the Habermasian account of the forum ideal, some of which have to do with doubts about whether unanimity would emerge even under ‘ideal speech’ conditions. Our own position does not involve any expectation of unanimity: differences of judgement will in general remain, which some procedure (simple plurality, perhaps) will have to reconcile. Nor do we have in mind the discursive confrontation of each with all as characteristic of the forum. Our open vote proposals more modestly expose each only to the risk of being observed by a variety of others. Therefore, we do not invite the problem of demogoguery in the way that Habermas's vision does. But the spirit of our argument is Habermasian. We reckon, with Habermas, that the open arena encourages one to direct one's speech and one's voting in a public interest direction. It is interesting that Elster does not question the Habermas ideal on the grounds that the forum invites bribery and corruption simply because it is open (see Section IV). Neither, to our knowledge, is this question raised by others. Elster's self-confessed preference (p. 128) is for a view of democratic politics lying somewhere between the ‘market’ conception on the one hand and the ‘forum-as-social-ethics-seminar’ on the other. Such a conception is to hand in the ‘judgement ideal’, and may well be tolerably realizable under some modest reforms of current practice. Unveiling the vote is, in our view, one such reform.

²⁶ Brennan, Geoffrey and Buchanan, James, ‘The Normative Purpose of Economic “Science”: Rediscovery of an Eighteenth Century Method’, International Review of Law and Economics, 1 (1981), 155–66.CrossRefGoogle Scholar

²⁷ Mueller, Dennis, Public Choice (New York: Cambridge University Press, 1979).Google Scholar

²⁸ Brennan, Geoffrey and Buchanan, James, ‘Voter Choice’, American Behavioral Scientist, 28 (1984), 185–201CrossRefGoogle Scholar; Brennan, Geoffrey and Lomasky, Loren, ‘Inefficient Unanimity’, Journal of Applied Philosophy, 1 (1984), 157–63.CrossRefGoogle Scholar

²⁹ We ignore the payoff associated, not with actually getting what you want, but with contributing to an electoral swing, if there is one, in that direction. This payoff is also problematic because of the vanishing contribution of a single vote to such a swing.

³⁰ For a decision theory which explicitly allows for the representation of such preferences see Jeffrey, Richard, Logic of Decision, 2nd edn (Chicago: University of Chicago Press, 1983).Google Scholar Notice however that on Jeffrey's evidential theory a person of type X may find it rational to vote for A rather than B on the basis of the associated outcomes, because of reasoning that if he does so then there is a better chance of other type-X people doing so. This ‘Newcomb’ result is a problem for that theory, however, rather than for us. On related matters see Price, Huw, ‘Against Causal Decision Theory’, Synthese, 67 (1986), 195–212CrossRefGoogle Scholar; and Pettit, Philip, ‘The Prisoner's Dilemma is an Unexploitable Newcomb Problem’, Synthese, 76 (1988), 123–34.CrossRefGoogle Scholar

³¹ In this discussion we move freely, since there is little danger of misunderstanding, between preferences over states of affairs and preferences over properties of states of affairs. See Pettit, Philip, ‘Decision Theory and Folk Psychology’, in Bacharach, Michael and Hurley, Susan, Essays in the Foundations of Decision Theory (Oxford: Blackwell, forthcoming).Google Scholar

³² Brennan, and Buchanan, , ‘Voter Choice’.Google Scholar

³³ See Pettit, , ‘Towards a Social Democratic Theory of the State’.Google Scholar Remember here the qualification mentioned in fn. 7; it explains the ‘normally’. In the spirit of the position described in fn. 7, someone may object that the consequentialist will be able to justify any old vote on the grounds that his vote makes no difference. Perhaps, but we take courage from the fact that consequentialists rarely resort to this move on being asked what they say – by parallel with what they would vote – that society should do.

³⁴ This theme will be familiar from the work of Jürgen Habermas. But for critiques of some aspects of Habermas's development of the idea see Pettit, Philip, ‘Habermas on Truth and Justice’, in Parkinson, G. H. R., ed., Marx and Marxisms (Cambridge: Cambridge University Press, 1982)Google Scholar and Elster, , Sour Grapes, pp. 35–42Google Scholar, as well as Elster, Jon, ‘The Market and the Forum’.Google Scholar

³⁵ Pettit, Philip, ‘The Freedom of the City: A Republican Ideal’, in Hamlin, Alan and Pettit, Philip, eds, The Good Polity (Oxford: Black well, 1989), chap. 1.Google Scholar

³⁶ Kinder, D. R. and Kiewiet, D. R., ‘Sociotropic Politics: The American Case’, British Journal of Political Science, 11 (1981), 129–61.CrossRefGoogle Scholar

³⁷ Kinder, and Kiewiet, , ‘Economic Discontent and Political Behavior: The Role of Personal Grievances and Collective Economic Judgements in Congressional Voting’, American Journal of Political Science, 23 (1979), 495–527.CrossRefGoogle Scholar

³⁸ For a more general statement of a republican political philosophy see Pettit, , ‘The Freedom of the City: A Republican Ideal’Google Scholar; and Braithwaite, and Pettit, , Not Just Deserts.Google Scholar It must be remarked, however, that some republicans were explicitly in favour of secrecy in voting. See Fink, Zera, The Classical Republicans (Evanston, III: Northwestern University Press, 1945).Google Scholar

³⁹ Quoted in Bourke, Paul and DeBats, Donald, ‘Charles Sumner, the London Ballot Society, and the Senate Debate of March 1867’, Perspectives in American History, New Series 1 (1984), 343–57.Google Scholar

⁴⁰ Bourke and DeBats draw attention to the fact that if Mill conducted an argument on broadly the American lines – they do not comment directly on his views – still there was an important contrast between the British and the American debate. ‘The argument for viva voce voting in Britain was often expressed in terms of maintaining legitimate influence over the voter. In the United States, where it was less common to concede either the existence or the legitimacy of social influence, the argument could take the form of urging that only by public recognition could republican independence be encouraged and psychologically rewarded’. Bourke, Paul and DeBats, Donald, ‘Identifiable Voting in Nineteenth-Century America: Towards a Comparison of Britain and the United States before the Secret Ballot’, Perspectives in American History, 11 (1977–1978), 259–98, at p. 263.Google Scholar

⁴¹ Mill, , Considerations on Representative Government, pp. 302–6.Google Scholar Nor it seems was Mill excessively optimistic, at least about politics in the large cities; see Fraser, Derek, Urban Politics in Victorial England (Leicester: Leicester University Press, 1976).Google Scholar Mill's line of thought is not of course without precedent in his intellectual culture. Consider for example the following remark by William Godwin, to which John Broome has drawn our attention: ‘Virtue will always be an unusual spectacle among men, till they shall have learned to be at all times ready to avow their actions, and assign the reasons upon which they are founded’; Godwin, William, Political Justice (Harmondsworth, Middx: Penguin Books, 1981), pp. 628–9.Google Scholar

⁴² See Finn, P. D., ‘Electoral Corruption and Malpractice’, Australian Federal Law Review, 8 (1977), 184–230, p. 197Google Scholar; ‘With the advent of full adult suffrage, bribery of electors ceased to be a potent force of influencing electoral results. Large and anonymous electorates destroyed its breeding ground’.

⁴³ It may be undesirable, exceptionally, in the closed, homogeneous communities that pluralistic societies sometimes contain.

⁴⁴ We are most grateful to Paul Bourke for sharing some of his knowledge of nineteenth-century voting patterns with us and for providing us with the stimulus of his criticism and incredulity. We have been aided by discussions with colleagues too numerous to mention but we would like to record our gratitude for written comments received from John Broome, John Burnheim, Bob Goodin, Frank Jackson, Chandran Kukathas, Loren Lomasky, Amartya Sen and Hugh Stretton, for comments received from anonymous referees, and for verbal comments offered when it was presented as a paper to the Department of Political Science, Research School of Social Sciences, Australian National University.