Elite Perceptions of the Poor: Reflections for a Comparative Research Proiect AE T his ~rticlc p~esents the ~hcorctica.l foundations for an ()ngoin~ resc:lrl~h proJect on elites and theIr perceptIons of poverty and poor people within their own societies. We arc addressing this topic for t,\\'o reasons. First, virtually nothing is known about such perceptions in nlOS( so,ictics, cspl-ci,llly in less developed countries where poverty is nlost serious and where our work is mainly focused. Second, the \vay in which clites pen:eivc and define poverty can powerfully affect social policy and the quality of life of the poor.

In the following I shall use comparative research on poverty as a frame of reference for the discussion. The context of research on poverty highlights the difficulties, but also the advantages of adopting a comparative focus.

No abstract

Poverty understanding and poverty research can be said to have gone through roughly three phases: Tale-telling, studies with a client focus as well as development research, and knowledge building about poverty reduction. The stage is now set for a new phase, that of understanding the processes that produce poverty and continue to produce poverty at a rate no present poverty reducing measures can possibly win over or even compete with. The challenge ahead is to make poverty production visible and place it firmly on the research agenda.

The arena of social policy could be a constant battlefield were it not confined within political institutions. Conflicts are extended into the organisation of social programmes and exert pressure on policy objectives and the contents of the programmes. Social programmes can accommodate these pressures in different ways. During this process programmes become transformed, some to such an extent that their social policy value diminishes and they serve other needs than those which they were designed to meet. This is the muffling effect of social policy. Ideally a social programme can be constructed with a low conflict potential. In this paper some basic characteristics of a low conflict system of social security are identified. Three different systems of social security are compared with regard to these characteristics and it is asked how conflict-ridden the different systems have been since they were first introduced. The model seems to have some predictive power. The methodological problems of such an approach are discussed.