Widening inequality in the distribution of resources, higher rates of unemployment and deteriorating conditions in the inner cities have concentrated attention in Britain, and elsewhere in Europe, upon the idea and exact meaning of ‘deprivation’. This scientific concept is used extensively not only in the analysis of social conditions but also, in an applied form, as an instrument of policy in allocating resources to particular regions, areas and services. This paper argues that the indicators which are chosen to represent the phenomenon are often unduly restricted and even involve double counting. As a result the distribution and severity of deprivation seems to be seriously misperceived and resources misallocated. A review of the available studies shows how the concept might be treated more coherently in relation to that of poverty.

Three approaches to defining poverty levels are discussed—social consensus approaches, budget standard methods, and behavioural approaches. Each addresses different questions and none, of itself, has provided—nor, it is argued, could ever provide—an objective definition of poverty. The paper then raises problems that have been largely neglected in defining poverty. First, the treatment of time and home production: the time and ability of individuals to prepare food or to wash and feed without assistance, for example, vary greatly depending on circumstances and in turn affect income needs. Choices and constraints affecting the household formations in which people live and their budgeting behaviour are also important in assessing poverty. Individual variations in behaviour need to be explicitly recognised if practical definitions of poverty levels are to be found. Finally, the paper condemns discussions of poverty that are part of the problem rather than part of the solution.

Since Beveridge, budget standards have been neglected in British social policy research. Empirical effort has concentrated on developing social indicator methods of investigating relative poverty. This paper explores the potential of budget standards for assessing whether the scale rates of supplementary benefit are adequate. Three applications of budget standard methodology are presented.

The principal problematic issue in all poverty measures is the source and status of the standards of needs and deprivation. Rejection of minimum subsistence or quasi-absolute approaches to defining poverty, and acceptance of the social relativism of poverty, logically demand that the indicators of deprivation equally be derived from the society in question and not be prescribed for it by ‘experts’. The paper reviews the policy implications of the distinction between deprivation and poverty, and discusses two principal approaches to empirical methods of establishing ‘consensual’ measures of poverty: the income proxy method and the deprivation indicator method. The paper also distinguishes sociologically-based poverty lines from politically-based social security scales, outlining some important aspects of the theoretical and methodological relations between them.

‘Consensual’ methods, which seek to establish poverty lines by reference to the views of society as a whole, are an important recent development. Three variants are recognised: those which require the public to estimate an adequate minimum income; those which ask people to specify a list of necessary items and those which ask what level of benefits the public is prepared to fund. This paper suggests that attempts to operationalise the consensual approach have been frustrated by their reliance on survey methodology. Some thoughts are offered on a possible methodology, based on the use of qualitative techniques, which would first explore consensus on the definition of poverty and then, if appropriate, seek directly to determine a socially approved budget standard.

This research note describes the distribution of family income in seven industrial nations, using data from the Luxembourg Income Study. Significant differences between the countries are identified. A ‘poverty cycle’ of the type identified by Rowntree still describes the distribution of income over age groups, but it is neither strong nor universal. Elderly families are found to be a key group for understanding the degree of income inequality in contemporary societies.