¹ Hollis, Martin and Nell, Edward, Rational Economic Man: A Philosophical Critique of Neoclassical Economics (Cambridge: Cambridge University Press, 1975), p. 34.CrossRefGoogle Scholar Chapter 1, ‘Relevance and Falsification’, contains a more extended discussion of the issue. Further discussion of the precise nature of Friedman's claims can be found in the last section of the article.

² Clearly within the scope of an article there is not room for an extensive treatment of the different positions of individual libertarians. I simply seek to characterize a claim commonly made by them and to consider the theoretical grounds which underpin it. The choice of Friedman, Hayek and von Mises was inspired not only by the desire to avoid the charge of ducking the issue by picking straw men. Their relevance is heightened by my claim that underlying methodological assumptions play a crucial part in the issue. The figures discussed were chosen because their treatment of such metatheoretical issues is commonly regarded as also grounding the work of others who deploy, but do not analyse, the same arguments.

³ See particularly the first and last chapters of Book V of The Principles of Political Economy.

⁴ These words are carefully chosen. It is true that Hayek and von Mises acknowledge the empirical limitations in applying the model of the perfect market, although it is notable that others such as Nozick and Friedman, especially in his policy-oriented writing, do so hardly at all. However qualified by some of the more sophisticated ‘new right’ theorists, the perfect market model is an immensely powerful and simplistic paradigm, and it is this version which ‘has lodged in the crania of politicians’ as Nick Bosanquet puts it (After the New Right (London: Heinemann, 1983), p. 44).Google Scholar To suggest that anyone taking it seriously should be dismissed as jejeune (as have several commentators on earlier versions of this article) would seem to be a rather chilling way of marginalizing criticism of a position which has gained immense political and rhetorical power.

⁵ Hayek, F. A., ‘The Moral Element in Free Enterprise’, in Studies in Philosophy, Politics and Economics (London: Routledge and Kegan Paul, 1967), pp. 233–4.Google Scholar Hayek almost universally appears as a libertarian, as he does here and, for example, in the Constitution of Liberty, where he criticizes cultural and national conservatism for providing a ‘bridge… to collectivism’ (The Constitution of Liberty (London: Routledge and Kegan Paul, 1960), p. 405).Google Scholar But it is interesting to note that his policy interventions can reveal a preoccupation with the very maintenance of cultural hegemony to which his libertarian defence of the market would seem opposed. See his letter on immigration to The Times, I March 1968, cited in part in Levitas, Ruth, ed., The Ideology of the New Right (Cambridge: Polity Press, 1986), p. 71Google Scholar: ‘the ordinary man only slowly reconciles himself to a large increase in foreigners amongst his neighbours, even if they differ only in language and manners,… the wise statesman… ought to aim at keeping the rate of influx low’.Google Scholar

⁶ A similar case is made by von Mises in Human Action: A Treatise on Economics (London: William Hodge, 1949), p. 280Google Scholar: ‘it is true that the employer has the right to fire the employee. But if he makes use of this right to indulge in his whims, he hurts his own interests. It is to his own disadvantage if he discharges a better man in order to hire a less efficient one. The market does not directly prevent anybody from arbitrarily inflicting harm on his fellow citizens; it only puts a penalty upon such conduct.’ However, von Mises is rightly more cautious about the ultimate power of the market to overcome the individual's ideal (or prejudice) regarding economic behaviour: ‘Dependence upon the supremacy of the consumers is not unlimited. If a man has a weighty reason for defying the sovereignty of the consumers he can try it… he may have to pay a price for his conviction… but it is left to a man's own decision to choose between a material advantage and the call of what he believes to be his duty’ (pp. 283–4).

⁷ See also Sowell, Thomas, Race and Economics (New York: McKay and Co., 1975), chap. 6.Google Scholar Sowell argues that the higher discrimination found in state and regulated industries can be explained by their not having to bear the full costs of job discrimination. His explanation of this cost is substantially the same as Friedman, 's (Race and Economics, pp. 168–9).Google Scholar A further resumé of the libertarian case is to be found in Gilder, George, Wealth and Poverty (New York: Basic Books, 1981), chap. 12 (‘The Myths of Discrimination’).Google Scholar

⁸ For example, Sowell's previously cited work and his Markets and Minorities (Oxford: Blackwell, 1980)Google Scholar; Freeman, Richard, ‘Black Economic Progress since 1964’, The Public Interest, 52 (1978), 52–68Google Scholar; and Banfield, Edward, The Unheavenly City Revisited (Boston. Mass.: Little, Brown, 1974).Google Scholar The opposing case is discussed below but see also Garfinkel, I. and Haveman, R., Earnings Capacity, Poverty, and Inequality (New York: Institute for Research on Poverty, 1977).Google Scholar Since the historical context of black emancipation can hardly be claimed to approximate to free-market conditions, the relevance of such studies must raise problems and always begs the ceteris paribus condition outlined above. Indeed, in some cases the failure of the data to fit the thesis is brazenly blamed on counterfactuals whose relevance is directly contingent on the truth of the very thesis in question. Thus Gilder, trying to explain a 20 per cent discrepancy between white and black earnings, when other identifiable indices are held constant, avers quite speculatively that ‘the remaining gap. the evidence increasingly suggests, relates not chiefly to discrimination against blacks but to earlier discrimination against their parents and to government-induced dependency and female-headed families’. No evidence is offered for the first statement, while the second hinges on the blithe assumption that ‘welfare can be shown to intensify and perpetuate the causes of poverty’. See Gilder, , Wealth and Poverty, pp. 132, 133.Google Scholar

⁹ Belshaw, C. S., Traditional Exchange and Modern Markets (Englewood Cliffs, NJ: Prentice Hall, 1965), p. 77.Google Scholar

¹⁰ Geertz, C., ‘Suq: The Bazaar Economy in Sefrou’, in Geertz, C., Geertz, H. and Rosen, L., eds, Meaning ami Order in Moroccan Society (Cambridge: Cambridge University Press, 1979), p. 198.Google Scholar

¹¹ Sahlins, Marshall, Stone Aye Economics (London: Tavistock Press, 1974), pp. 191 ff.Google Scholar

¹² Mintz, S., ‘Pratik: Haitian Personal Economic Relationships’, Proceedings.of the American Ethnographic Association (Seattle: Washington University Press, 1961), pp. 54–63.Google Scholar

¹³ Davies, W. G., Social Relations in a Philippine Market, Self-interest and Subjectivity (Berkeley: University of California Press, 1973), p. 216.Google Scholar

¹⁴ Wilson, J. A., ‘Adaptation to Uncertainty and Small Numbers Exchange: The New England Fresh Fish Market’. Bell Journal of Economics, 11 (1980), 491–504, p. 491.CrossRefGoogle Scholar

¹⁵ Plattner, Stuart, ‘Economic Custom in a Competitive Marketplace’, American Anthropologist, 85 (1983), 848–58, pp. 852, 856.CrossRefGoogle Scholar

¹⁶ An economic historian noting the phenomenon in the emergent commodity markets of England's commercial revolution, comments: ‘buyers and sellers often attempted to reduce the risks of short-run fluctuations in supply and demand by attempting to nurture special trading relations with an opposite number’ (Anderson, B. L., ‘Entrepreneurship, Market Process, and the Industrial Revolution in England’, in Anderson, B. L. and Latham, A. J. H., eds, The Market in History (London: Croom Helm, 1986), p. 16).Google Scholar

¹⁷ One might (charitably) suggest that the social exclusivity of the pre-Big Bang London Stock Market was another way of minimizing risk in a market where trust was al a premium and surveillance difficult and costly.

¹⁸ Trager, L., ‘Customer and Creditors: Variations in Economic Personalism in Nigeria’, Ethnology, 20 (1981), 133–46.CrossRefGoogle Scholar

¹⁹ Davis, , ‘Social relations in a Philippine market’, p. 216 (for suki)Google Scholar; Trager, , ‘Customer and Creditors’, p. 137 (for onibara)Google Scholar; Plattner, , ‘Economic Custom in a Competitive Marketplace’, p. 852Google Scholar; all these emphasize the risk-minimizing function of the practice. Risk aversion in peasant economies is well documented: Johnson, Allen, ‘Security and Risk-taking among Poor Peasants’, in Dalton, G., ed., Studies in Economic Anthropology, American Anthropological Association Monograph, 7 (Washington: American Anthropological Association, 1977)Google Scholar; Lipton, Michael, ‘The Theory of the Optimising Peasant’, Journal of Development Studies, 4 (1968), 327–51, esp. pp. 334–5Google Scholar; Bliss, C. J. and Stern, N. H., Palanpur: Economics in an Indian Market (Oxford: Oxford University Press, 1982), pp. 300ff.Google Scholar

²⁰ Davis, , ‘Social Relations in a Philippine market’, pp. 229, 256.Google Scholar

²¹ Plattner, , ‘Economic Custom in a Competitive Marketplace’, p. 851.Google Scholar

²² Arensberg, C., The Irish Countryman (London: Macmillan, 1937), pp. 154, 161.Google Scholar

²³ Compadrazgo involves the ritual ‘adoption’ of a trading partner. The practice is cited in Davis, , ‘Social Relations in a Philippine Market’, p. 229.Google Scholar See further: Mintz, S. and Woolf, E., ‘An Analysis of Ritual Co-parenthood (Compadrazgo)’ South West Journal of Anthropology, 6 (1950), 341–68.CrossRefGoogle Scholar

²⁴ Market neutrality is a complex concept. A good recent discussion of the various possible connotations is to be found in O'Donnell, A. T., ‘The Neutrality of the Market’, in Goodin, Robert E. and Reeve, Andrew, eds. Liberal Neutrality (London: Routledge, 1989).Google Scholar Most importantly, for the discussion here, O'Donnell distinguishes between the issue of neutrality between transactors and neutrality between preferences (p. 42).Google Scholar

²⁵ Another rather different problem relates not to public goods, as normally conceived, but to jointly-supplied, ideal-regarding goods – what David Miller has called ‘group-oriented desires’, such as the desire to produce in a co-operative rather than a hierarchical environment. Here, too, as Miller shows, the market fails to deliver without political intervention. See Miller, David, ‘Market Neutrality and the Failure of Co-operatives’, British Journal of Political Science, 11 (1981), 309–29CrossRefGoogle Scholar; and see also his more recent Market, State and Community (Oxford: Oxford University Press, 1989), pp. 83–93.Google Scholar

²⁶ Radio-direction beacons already exist at sea of course, but any suitably tuned radio-receiver can pick them up without charge. In the present political climate one hesitates to put forward what might appear as a reductio ad absurdum.

²⁷ In this sense discrimination may be ‘inefficient’, but not sub-optimal, given the inclusion of discrimination in the preference-schedules of employers. Even so, affirmative action legislation may be a necessary device to bring home to an employer the financial cost of his prejudice, by coercing at least some in-group market actors to break ranks and thereby realize the market advantage available to them. See O'Donnell, , ‘Neutrality of the Market’, p. 55.Google Scholar

²⁸ Decision theory runs into difficulties in deciding whether observed ‘inconsistencies’ in preferences ‘disprove’ the axioms of decision theory or reveal changes in ceteris paribus conditions, for example by displaying ‘learning’, or require the attribution to subjects of differential risk aversion, or of perverse characterizations of the choices facing them. It appears to be impossible to devise satisfactory tests for discriminating amongst these various possibilities. A classic article is Tversky, A., ‘A Critique of Expected Utility Theory: Descriptive and Normative Considerations’, Erkenntnis, 9 (1975), 163–74.Google Scholar For a brief discussion, see Davidson, Donald, ‘The Philosophy of Psychology’ in Davidson, Donald, Essays on Actions and Events (Oxford: Oxford University Press, 1980), pp. 269ff.Google Scholar A related issue is the difficulty traditional economic theory has in accommodating the notion of commitment. On this point, see the seminal article by Sen, Amartya K., ‘Rational Fools: A Critique of the Behavioural Foundations of Economic Theory’ in (inter alia) Philosophy and Puhlic Affairs, 6 (1976–1977), pp. 317–44Google Scholar: and in Hahn, F. and Hollis, M., eds., Philosophy and Economic Theory (Oxford: Oxford University Press, 1979)Google Scholar: ‘The characteristic of commitment with which I am most concerned here is the fact that it drives a wedge between personal choice and personal welfare, and much of traditional economic theory relies on the identity between the two’ (Sen, , in Hollis, and Hahn, , eds, Philosophy and Economic Theory, p. 97Google Scholar). Intolerance, or the prejudicial behaviour resulting from it, may be regarded analytically as a form of commitment.

²⁹ Indeed, if in-group cohesion extends to other workers and customers, as well as to the employing group, boycott responses could actually impose costs on defecting employers.

³⁰ The latter example was suggested to me by Jeremy Shearmur, who was present on the original occasion when the paper on which this article is based was delivered.

³¹ Sowell, one of the more judicious apologists for free-market policies in these areas, does concede that ‘under special conditions, discrimination can be made profitable’ (Race and Economics, p. 169Google Scholar). But these ‘special conditions’ turn out to be not at all unusual: significant imbalance of numbers between the groups, easy identification, a high desire to exclude and a low cost of exclusion. ‘Cost’ to Sowell is clearly a money-price cost and therefore begs the whole question (pp. 169–72).

³² For example: ‘Praxeology is indifferent to the ultimate goals of action… It applies the term happiness in a purely formal sense. In the praxeological terminology the proposition: man's uniqueaim is to attain happiness, is tautological. It does not imply any statement about the state of affairs from which man expects happiness’ (von Mises, , Human Action, p. 15).Google Scholar

³³ Some discussion of the difficulties involved in defining neutrality in political procedures can be found in Raz, J., The Morality of Freedom (Oxford: Oxford University Press, 1986), chap. 5Google Scholar, and in Gauthier, David, Morals by Agreement (Oxford: Oxford University Press. 1986), chap, ivGoogle Scholar: ‘The Market: Freedom from Morality’. However, whilst both celebrate the formally liberating effect of the market from moral and other-regarding considerations, neither author seriously considers the issue of the scope thereby allowed for those who wish to sustain such elements in their choices. Miller, David, in Market, State and Community (chap. 3)Google Scholar, acknowledges the importance of the issue but elects to ‘forgo any criticism of the market on the grounds that it induces people to adopt (e.g.) commodity-based conceptions of the good’ (note 7, p. 75).

³⁴ The interesting exception to this position is Samuelson, who argues that as a matter of logic, hypotheses whose consequences are fully specified, realistic and confirmed must rest on assumptions which are realistic. This is strictly entailed only if all the consequences are specified and confirmed and no other assumptions will generate equivalent hypotheses. Of course, not all – in fact very few – hypotheses are such that their consequences can all be specified, and even fewer are all confirmed, so the force of the logical point is considerably blunted. He falls back on a chauvinistic gut feeling that realism in assumptions is indeed important: ‘All economic regularities that have no commonsense core that you can explain to your wife will soon fail’ (see Samuelson, P., ‘Comment on E. Nagel's.“Assumptions in Economic Theory”’, in The Collected Scientific Papers of Pani A. Samuelson, 3 vols (Cambridge, Mass.: MIT Press, 1966), vol. II, p. 1175).Google Scholar

³⁵ von Mises, L., Epistemological Problems of Modern Economics (Princeton, NJ: von Nostrand, 1960)Google Scholar, excerpted in Hahn, and Hollis, , eds. Philosophy and Economic Theory, p. 64.Google Scholar Von Misessees what he calls the basic principles of human action as akin, in the practical world, to Kant's synthetic a prioris. ‘They are the necessary mental tool to arrange sense data in a systematic way, to transform them into facts of experience, then these facts into bricks to build theories, and finally the theories into technics to attend ends aimed at’. See von Mises, Ludwig, The Ultimate Foundation of Economic Science (Princeton, NJ: von Nostrand, 1962), p. 16Google Scholar; see also pp. 19–21. Unlike Friedman, who is a fairly unreconstructed logical positivist, and Hayek, who endorses positivistic science but sees it as a basis for deductive forays into untestable territory, von Mises rejects the application of positivism to the human sciences. This is because he sees positivism as an historical reaction of the natural sciences to the subversion of their paradigm, Euclidean geometry, by non-Euclidean geometries. But destruction of synthetic a prioris in natural science has, he claims, no implications for the study of human action, where they can therefore continue to be applied (see Hayek, , ‘Degrees of Explanation’ in Studies (which is dedicated to Popper), pp. 5–7Google Scholar; and von Mises, , The Ultimate Foundation, p. 5).Google Scholar Thus, although each of the three thinkers occupies a different position on the episiemological status of hypotheses or axioms, they are united in denying that they are descriptions of reality.

³⁶ Friedman, Milton, Essays in Positive Economics (Chicago; University of Chicago Press, 1953), pp. 14–15.Google Scholar