The contribution of social theory to the field of social policy is too often one which begins in ideology and ends in rhetoric. The subject matter of social policy contains many of the most urgent problems of our time. Consequently theories and models of social welfare tend to be highly normative, explaining how men ought to behave if they wish to accomplish certain results, as discussed in Horton (1966).

Currently the two most influential theoretical formulations in social policy are based respectively on the ‘institutional’ and the ‘residual’ models of social welfare. Both models draw on historical and sociological evidence to predict future trends in social policy. The residual model rests on moral assumptions about the selfevident virtues of competition and self-help. The institutional model rests its moral case on the ethics of co-operation and mutual aid.

In both cases psychological assumptions are made about the attitudes of individuals towards welfare provision. Their common weakness is a tendency to confuse academic perceptions of social reality with those of the ordinary users of social services. There is, however, no firm evidence as yet that sizeable sections of the community are strongly committed either to the ethic of mutual aid or to the liberties of the free market. The end result for the discipline of social policy and administration is that too much is prescribed, too much indicted and too little explained.

The central and unresolved issue in this debate has been defined by Richard Titmuss as the problem of developing socially acceptable selective services within an ‘infrastructure of universalist services’ in such a way that stigma is reduced to a minimum. This approach correctly focuses attention upon the subjective realities of everyday life for those in need.

The aim of this paper is to set out a number of empirically testable hypotheses in the form of a model of social welfare. If substantiated, the model could be used to classify welfare systems in terms of their stigmatising propensities. It should then be possible to formulate a theory explaining why people are elevated or debased in exchange situations and which conditions of provision and usage are most likely to engender stigma in industrial societies.