View article

Amartya Sen's work on economic poverty has brought about a serious reconsideration of how poverty is measured. The time-tested method of counting the number of poor is now dismissed for failing to indicate the intensity of poverty; the more recent vintage of statistics based on the mean or aggregate income of the poor are criticized for failing to reflect the distribution of income among the poor. These observations have led to the development of several new summary statistics or measures of poverty, almost all of which are based firmly in the framework suggested by Sen. But what exactly is Sen's approach to poverty measurement, and how is it justified? How does the Sen measure compare to subsequent measures presented Where does the traditional problem of setting a poverty within his framework? line figure into this literature? In this paper we shall examine critically the recent work on poverty measurement with the goal of providing clear answers to these basic questions.

We begin, of course, with Sen's approach to poverty measurement, examining first his three motivating axioms: the focus axiom, the monotonicity axiom, and the weak transfer axiom. Next the Sen measure is defined and we trace the steps leading from the general principles to the specific form Sen adopts. We shall pay particularly close attention to those aspects of the argument that reappear in the subsequent literature: Sen's conception of relative deprivation, the relation between poverty and inequality measurement, and the use for which the poverty measure is sought.