Kenza Elass*, Bertille Picard**

Internal seminars
phd seminar

Kenza Elass*, Bertille Picard**

AMSE
Sex and the density: Urban wage premia and the gender wage gap*
Fairness of welfare-maximizing algorithms in experimental designs**
Joint with
Cecilia García-Peñalosa, Christian Schluter*
Cancelled
Venue

MEGA Salle Carine Nourry

MEGA - Salle Carine Nourry

Maison de l'économie et de la gestion d'Aix
424 chemin du viaduc
13080 Aix-en-Provence

Date(s)
Tuesday, January 31 2023| 11:00am to 12:30pm
Contact(s)

Camille Hainnaux: camille.hainnaux[at]univ-amu.fr
Daniela Horta Saenz: daniela.horta-saenz[at]univ-amu.fr
Jade Ponsard: jade.ponsard[at]univ-amu.fr
Nathan Vieira: nathan.vieira[at]univ-amu.fr

Abstract

*In France, the gender wage gap for the 20% of the workforce living in the denser locations is 22% lower than that for those living in the bottom 20% of the density distribution, indicating that women benefit more from density than men. This paper explores the importance of geographical location for understanding the gender wage gap. Following the recent literature on economic geography that takes into account the endogeneity of location, we estimate the difference in the returns to urban density across genders. Results show that the elasticity of the wage premium increases with density for both men and women, with a significantly higher female urban wage premium. We consider different mechanisms to explain this pattern.

**Welfare-maximizing algorithms offer new insights in experimental economics. During an experiment, they identify the most beneficial treatment for the subjects and thus maximize the experiment's overall welfare impact. However, for experimentalists or policy implementers, this implies transferring decision-making power to an algorithm. Allocating individuals to treatment arms using an algorithm exposes us to contemporary criticisms of artificial intelligence, such as discrimination or exacerbation of inequalities. Can we meet the requirements of fairness in these automated designs? I will present preliminary work implementing strategies to control inequalities generated by basic algorithms using recent results on inference in this experimental setup.