Ulrich Aiounou*, Laura Contreras**

phd seminar

Ulrich Aiounou*, Laura Contreras**

Treatment effect estimation : Lasso vs Autometrics*
The dual faces of development: Assessing the effects of industrial place-based policies on informality in Colombia**

IBD Amphi

Îlot Bernard du Bois - Amphithéâtre

5-9 boulevard Maurice Bourdet
13001 Marseille

Tuesday, May 30 2023| 11:00am to 12:30pm

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


*Selecting relevant variables in a model is a common problem faced by economists in general and in particular in causal effect estimation models. We show, with simulation evidence that autometrics-based methods perform well in general, and especially perform well when lasso-based methods fail. We apply these methods to two application in real database. We retain that autometrics-based methods should be preferred to lasso-based methods for their stability and their property for allowing to control the number of irrelevant variable selected.

**The disparities in economic activity across and within regions have been one of the most studied empirical regularities due to their implications for social and economic inequalities. Governments have implemented place-based policies to promote the economic development of their most remote and lagged areas and so help to mitigate their regional disparities. Despite the high prevalence of developing economies implementing this policy tool, the existing literature on place-based policies has predominantly focused on examining their effects in developed economies. Nevertheless, the experience with place-based policies in developed countries can hardly be transferred to less-developed economies where a substantial share of their labour force is employed informally. To shed light on these, this paper examines the effects of the Free Economic Zones (FEZs) regime established in Colombia in 2005. Based on a novel data set that combines georeferenced data on FEZs, municipalities and household characteristics, I measure the regional effects of the construction of these zones on labour market outcomes, particularly focusing on informality, using a novel IV strategy. Preliminary findings suggest that the Free Economic Zones regime increases the population and employment in the regions hosting a zone. To assess the aggregate effects of these policies, I aim to construct a quantitative spatial economic model that incorporates labour market frictions and imperfectly enforced regulations, heterogeneous firms sorting into the formal or informal sector, and explains how the input-output linkages between firms within and outside the zones, influence the generation or destruction of informal employment opportunities.