Can we trust the poor? Survey on the conditions of a preference-based poverty indicator

Several economists have recently proposed an original way of measuring poverty: an individual is no longer considered poor when he or she cannot afford a minimum basket of goods, but when he or she prefers this basket to his or her situation. This thesis presents a critical analysis of this proposal in two parts. In the first part, three major fields of the literature in philosophy of economics (materialist, sociological and behavioral) are mobilised in order to identify certain limits to this approach, that is the possible dangers of placing blind trust in the preferences of the poor. It appears, for example, that the poor tend to overestimate the well-being of their current situation in order to make it psychologically bearable or because it seems "natural" to them. A preference-based poverty indicator would then ignore these people, which may seem problematic from a normative point of view. In the second part of the thesis, two alternative poverty indicators are therefore proposed: while they are still based on the preferences of the poor, they limit the biases identified in the philosophical literature.