The idea of discourse alerts us to a number of issues about the social construction of social problems. It suggests that we need to look beyond competing theories or perspectives to look at how knowledge is organised around central themes that allow the different theories to compete. Discourses define what the problem is, and it is because theories share the definition of the problem that they can compete and argue. Perspectives that start somewhere else – or do not share the definition of the problem – have great difficulty in making themselves heard or understood. They do not fit the terms of reference of the discourse. Thus the discourse represented by the COS rules is not one that would allow for the discussion of poverty either as a consequence of the behaviour of the rich or as a consequence of structural social inequalities.
Second, discourses shape what sorts of knowledge are meaningful, worth having and ‘truthful’. As we have seen, the discourse of poverty means that we look for, collect and use knowledge about ‘poor people’. To suggest that this category of people does not exist, or that we should not try to know more about them sounds meaningless. Of course, ‘everybody knows’ that there are poor people. They can be identified, measured and investigated (and we can only do something about the problem of poverty if we know more about it). But the category of ‘poor people’ only exists because of the discourse of poverty, and we should not assume that ‘they’ are ‘different’ because of that. Placing ‘them’ in a separate category can be seen as a means of marginalising or excluding ‘them’ from normal life.
Third, the idea of discourse points us to the importance of looking at the ways in which the poor are institutionalised in social arrangements and relations of power. The workhouse, charity organisations, benefits offices, means tests, cohabitation rules and so on are ways in which the discourse of poverty has been institutionalised (and they also reflect different sorts of theories, policies and perspectives within the discourse). They involve relations of power between groups of social actors – claimants, assessors, case workers, fraud investigators, and so on.