
Soutenance
Soutenance de thèse de Mohamed Soual
le 13 décembre 2021
14h
Pauvreté, expérimentation et développement dans les travaux d’Esther Duflo
Membres du jury :
Summary : the J-PAL approach is widely credited with meeting the criteria of internal validity. This thesis seeks to challenge this judgment by drawing on a comprehensive critique informed by Quine and Duhem's analyses of the status of theories. Thus, it emphasizes that evidence of causality is not provided as a result of the discrepancy between the conditions of randomization and the assumptions underlying the statistical tools, the evacuation of heterogeneity following the determination of the average effect, and the contingent nature of the results. These limitations, which compromise generalization and hence the establishment of a firm basis for ensuring the commutativity of results, stem from the undue primacy given to empirical observation at the expense of theory. On this basis, the central proposition developed is that the admissibility of the poverty approach in terms of experiments is due, not to the validity of its method and its statements, but to the qualification acquired in the scientific forums, as an experimental microeconomics of development with a Gold Standard, and in the forums of expertise and public policy making. In these respects, it contributes to the narrowing of the object of development economics by reducing its scope to the sole microeconomics of decisions and reactions to interventions. In so doing, it evacuates the impact of structures on the choices of agents, thus consecrating the recusal of macroeconomics. At the same time, it confines the study to local objects, giving a predominant status to the individual behavior of agents. In short, based on observation reports without any theoretical framework, the J-PAL approach appears to be more normative and performative than positive.
- M. Rédouane TAOUIL, Université Grenoble-Alpes, Directeur de thèse
- M. Vincent SPENLEHAUER, École Nationale des Ponts et Chaussées, Rapporteur
- M. Pierre-Noel GIRAUD, Mines Paris Tech, Examinateur
- M. Ahmed TRITAH, Université de Potiers, Examinateur
- Mme Sophie SAGLIO ROSSINI, Université Paris 8 - Vincennes St Denis, Examinatrice
- M. Juan José JARDON URRIETA, Universidad Michoacana de San Nicolas de Hidalgo, Rapporteur
Summary : the J-PAL approach is widely credited with meeting the criteria of internal validity. This thesis seeks to challenge this judgment by drawing on a comprehensive critique informed by Quine and Duhem's analyses of the status of theories. Thus, it emphasizes that evidence of causality is not provided as a result of the discrepancy between the conditions of randomization and the assumptions underlying the statistical tools, the evacuation of heterogeneity following the determination of the average effect, and the contingent nature of the results. These limitations, which compromise generalization and hence the establishment of a firm basis for ensuring the commutativity of results, stem from the undue primacy given to empirical observation at the expense of theory. On this basis, the central proposition developed is that the admissibility of the poverty approach in terms of experiments is due, not to the validity of its method and its statements, but to the qualification acquired in the scientific forums, as an experimental microeconomics of development with a Gold Standard, and in the forums of expertise and public policy making. In these respects, it contributes to the narrowing of the object of development economics by reducing its scope to the sole microeconomics of decisions and reactions to interventions. In so doing, it evacuates the impact of structures on the choices of agents, thus consecrating the recusal of macroeconomics. At the same time, it confines the study to local objects, giving a predominant status to the individual behavior of agents. In short, based on observation reports without any theoretical framework, the J-PAL approach appears to be more normative and performative than positive.
Localisation
Salle visioconférence du Creg, Bateg
Mis à jour le 26 novembre 2021