There is no escape from value choices in welfare systems. The construction of any models or the elaboration of any theories which have anything to do with policy must inevitably be concerned with ‘what is and what might be’; with what we (as members of a society) want (the ends); and with how we get there (the means).

Not only is policy all about values but those who discuss problems of policy have their own values (some would call them prejudices). But, whatever they are called, it is obvious that the social sciences, and particularly economics and sociology, are not, nor can ever be, ‘value-free’. No doubt there are some practitioners, concerned about the professional status of their subject or discipline (which in itself is an odd word suggestive of a master–servant relationship), who like to think they are value-free, floating in an abstracted social world of an exact Robinson Crusoe science. They take refuge in the description of facts, or in mathematical model-building, or in a mystique of casework or psychoanalysis. But even the labelling of one’s subject or the choice of topics for research or for teaching in the social sciences can reveal the existence of value premises. Consider the use of such terms as ‘race relations’. Does it not imply certain judgements about the unscientific concept of race? Or the use of such valuative and emotive terms as ‘economic growth’, ‘progress’, ‘productivity’, ‘health’, ‘social mobility’ and ‘equality’.

There are, of course, some social phenomena that may be studied with a certain degree of cold rational disinterestedness while never achieving the lack of involvement displayed by a mathematician analysing a quadratic equation. But this is not possible with the study of social policy or, to take another example, social deviation, which is a concept inextricably mixed up in all fields of social policy, medical care, social work and psychiatry. One of the value assumptions, for instance, is that the residual non-market sector (the public social policy sector) should concern itself with the social deviants; the ‘bad actuarial risks’, those who are unwilling or unable to provide for their own needs and the needs of their families through the normal mechanisms of the market.