¹ Barry, B. M., ‘The Public Interest’, Aristotelian Society Proceedings, Supplementary Volume XXXVIII(1964), 1–18CrossRefGoogle Scholar, p. I4f., and his Political Argument (London: Routledge and Kegan Paul, 1965)Google Scholar, Chap. II, section 2. I shall not here offer reasons for favouring Barry's view. A useful recent review of alternative accounts of the public interest is found in Virginia Held, The Public Interest and Individual Interests (New York and London: Basic Books, 1970).Google Scholar Held classifies Barry's as a ‘common interest’ account of the public interest. The discussion in the present article is relevant to one of Held's comments on Barry's position: ‘His approach… assigns no role in evaluations of the public interest to anyone's net interests, which would seem to be what any individual asserting his interests would actually and justifiably ground them upon’ (p. 118).

² This ‘balancing-out’ principle belongs to a family of such ‘harmony of interests’ and ‘long run’ considerations well known in the history of economic thought. (See Myrdal, Gunnar, The Political Element in the Development of Economic Theory (London: Routledge and Kegan Paul, 1953)Google Scholar and the appendix by P. E. Streeten.) It is perhaps useful to point out in advance a distinctive feature of the present instance of such a consideration that protects it against a standard criticism of assumptions about the harmony of interests of individuals. The general truth relied on here relates only to a series of policies, each of which is itself in the public interest. This contrasts sharply with those arguments that seek to show that a body of policies as a whole will probably be in the interest of each member of a community even though the individual policies may be in the interests only of sections of that community, by appealing to the tendency of such inequalities to balance out in the long run. Thus, to take a famous recent example, Hicks, J. R. argued in ‘The Rehabilitation of Consumer's Surplus’, Review of Economic Studies, VIII (1940–41), p. 111Google Scholar, that If the economic activities of a community were organised on the principle of making no alterations in the organisation of production which were not improvements in this sense [i.e., the gainers could ‘overcompensate’ the losers], and making all alterations which were improvements that it could possibly find, then, although we could not say that all the inhabitants of that community would be necessarily better off than they would have been if the community had been organised on some different principle, nevertheless there would be a strong probability that almost all of them would be better off after the lapse of a sufficient length of time. One objection to this is that it is not obvious that any inequalities resulting from the pursuit of such a principle will tend to cancel out in the long run rather than lead to continuing inequalities. If however the constituent policies are themselves individually in the public interest, it can be expected (I argue) that balancing out will probably occur.

³ There is one kind of inequality in the benefit conferred by a policy that is in the interests of a group, that will constitute an important qualification of this general claim. Inequalities may be due to the fact that financial benefits are distributed to members of a group in a way broadly proportional to the wealth already possessed by the individual members. The less well-off members will have their interests served less by such a policy than will the better-off, and with a number of such policies the effect will of course be cumulative. I shall not pursue this point in the following pages, but will leave it to be understood that the continuance of serious inequalities of this kind is compatible with its being otherwise true that as the number of policies increases, so does the probability that the interests of each member of the group benefit equally.

⁴ If we wish to allow for the possibility that there is more than one public interest policy that would satisfy the requirements of the higher-order policy, then a more complex condition must be satisfied if the individual policy is to be justifiable in the way discussed here. An individual public interest policy P a1 will be a member of the body of policies required for a higher-order policy H, if: (1) P a1 is one among a number of policies of kind P a, one of which must be followed if H is to be pursued; and (2) if P a1 were replaced by one of the alternative policies of type P a, such as P a2, then someone's interests would suffer as a result of this realization of H more than the first individual's interests suffer as a result of the inclusion of P a1 in H.

⁵ Here I follow Barry with minor differences of wording. See ‘The Public Interest’, p. 4, and Political Argument, Chap. 10.

⁶ An illustration may be helpful here. We may imagine a situation where three major policy programmes are competing for public support and limited funds. One higher-order policy aims at increasing employment throughout the country by various means, and requires such individual policies as regional investment incentives, state subsidies, industrial training programmes and environmental rehabilitation schemes. Each policy is required to be in the public interest compared with the accepted available alternative, and will if necessary include a compensatory element for those groups whose interests would otherwise not benefit, even when the individuals are considered qua members of the public. A second higher-order policy aims at improving health, welfare and educational standards through such policies as new regional hospital organizations, health centres, preventive medical services, and improving standards in existing educational institutions. Again each such policy is required to be in the public interest, and may involve a compensatory element. The third higher-order policy aims at a great expansion of defence arrangements, and requires such individual policies as expansion of certain armament and related industries, and the expansion of certain areas of education. Here again each policy is required to be in the public interest, and may involve a compensatory element. It is clear from this simplified example that while each constituent policy may correctly be described as being in the public interest, compared with the accepted available alternative to it, only one of the higher-order policies could correctly be described as being in the public interest compared with the accepted available alternatives (unless of course two or more were equal in this respect).

⁷ The qualifying phrase ‘considered as members of that majority’ reminds us that there may be some members of such a majority who do not benefit or who benefit much less than others from a certain majority interest policy. And if we suppose that a majority of the majority group will benefit, this may conceivably not be a majority of members of the public. The various possibilities need not be pursued here. Even if we suppose that certain members of such a majority never benefit from a body of such majority interest policies, this amounts only to the supposition that the membership of a minority that never benefits from the body of majority interest policies is slightly increased. It is the likelihood of there existing such a minority that is being emphasized in this paragraph.

⁸ It must of course be conceded that as the number of instances of majority interest policies in a body of policies increases, there is a greater probability that the membership of the minorities will increase in variety, and the distribution of benefits become less unequal, other things being equal. If other things are to be equal, there must not exist factors that would tend to cause a certain minority to benefit much less than others, even when the number of instances of majority interest policies increases. This requirement corresponds of course to the requirement in the case of the public interest policies that exceptional circumstances should be absent. There is however in the case of public interest policies a further reason to expect the balancing-out to tend to occur as the number of instances increases, which is not paralleled for the case of majority interest policies. In the case of public interest policies it is implied that each constituent policy will be in the actual interest of each individual member of the public so long as the policy does not conflict with other interests of his. The ‘weaker’ considerations about balancing-out suggest that with many and varied public interest policies such conflicts will not be expected to occur in many cases for any particular individual. In the case of majority interest policies we are not entitled to assume that each constituent policy will be in the actual interest of each individual member of the public so long as the policy does not conflict with other interests of his. For we have to allow for those majority interest policies that are in the interests of a majority of individuals qua members of that majority, and not in the interests at all of individuals belonging to a minority.