Social Indicators and Welfare Monitoring

In this paper prepared for the 1999 Copenhagen Seminar for Social Progress, Gøsta Esping-Andersen reviews approaches to welfare monitoring since the 1960s. The attempt to record and measure changes in welfare over the past four decades has been influenced by a number of factors, including the changing ideological climate, practical difficulties with concepts and data, and objective changes in social conditions. Countries around the world also have different overall perspectives on welfare provision: they define welfare goals differently and have developed distinct regimes, or sets of policies and institutions, for meeting these goals. Even advanced industrial nations, for example, have differing conceptions of welfare provision, ranging from the minimalist approach of Anglo-Saxon countries to the social-democratic focus of Nordic societies. The first devotes primary attention to identifying the poor and needy, while the second is more concerned with monitoring equal access to resources among the population at large.

Each of these is based on a specific theoretical approach to welfare. The first, and most prevalent, way of thinking about social policy rests on the calculation of risk. Each risk is typically identified as a separate event or state, with its own actuarial probability and defined response. This approach tends to individualize welfare issues—to concentrate on discrete cases—and to encourage the use of indicators such as poverty head counts, years of schooling, mortality and longevity, or unemployment rates. Although this may be adequate when societies develop along predictable lines, and risks are clearly identifiable with distinct stages of life or age categories, a concentration on risk becomes increasingly less useful as societies change in complex ways.