Discussion: A general theory of welfare
All the welfare theories considered so far are based on some moral value (also represented by sets of axioms) that can be intended as subjective values (pleasure or preferences) or objective values (ideals). As Russell (1954, pp. 100-7) argues, a rational choice between competing moral systems cannot be made since 'a genuine difference exists as to ends', and it is not possible 'to advocate any argument in favour of the one against the other', because 'what a man will consider to constitute his happiness depends upon his passions, and these in turn depend upon his education and social circumstances as well as upon his congenital endowment'.13
This argument prompts us to think in terms of a general theory of welfare (GTW). From the point of view of moral values, it must admit all the possible values that a society can espouse in all possible situations. This means that not only subjective values have to be admitted but also objective values. In terms of the models in which GTW is expressed, we believe that it is a branch of logic. 'Logic' is intended here in a wider sense than mathematical logic, because it is concerned not only with deductive arguments but also with inductive reasoning. In other terms, as regards instrumental rationality, GTW must be based on rational dualism, which claims that when it is not possible to obtain the knowledge necessary for applying the MEU procedure, or more generally the maximisation of the expected value, other procedures must be admitted in order to make rational decisions. Therefore, in order to present some actual cases, we refer to policymakers' practical capacity to pursue sustainable social welfare. Today, in order to pursue this aim, environmental policy not only recognizes the intrinsic objective value of nature (Bellamy and Johnson, 2000) but, depending on the particular situation, it makes reference to either substantive rationality or bounded rationality (Steyaert et al., 2007).