Mardi 15 octobre 2019 — Séminaire DYNAMICS

INVITES :
  • Catia Batista (Nova School of Busi­ness and Economics) ;
  • Riccardo Turati (UCLou­vain).
HORAIRES ET LIEU :
  • 16:30–19:00
  • PSE, 48 Bd Jourdan 75014 Paris, Salle R1-09
RESUMES :

Catia Batista (Nova School of Busi­ness and Econo­mics): Testing Classic Theo­ries of Migra­tion in the Lab (travail conjoint avec David McKenzie)

We use incen­ti­vized labo­ra­tory expe­ri­ments to inves­ti­gate how poten­tial migrants make deci­sions between working in different desti­na­tions in order to test the predic­tions of different classic theo­ries of migra­tion. We test theo­ries of income maxi­mi­za­tion, migrant skill-selec­tion, and multi-desti­na­tion choice and how the predic­tions and beha­vior under these theo­ries vary as we vary migra­tion costs, liqui­dity constraints, risk, social bene­fits, and incom­plete infor­ma­tion. We show how the basic income maxi­mi­za­tion model of migra­tion with selec­tion on observed and unob­served skills leads to a much higher migra­tion rate and more nega­tive skill-selec­tion than is obtained when migra­tion deci­sions take place under more realistic assump­tions. Second, we find evidence of a home bias, where simply label­ling a desti­na­tion as “home” causes more people to choose that loca­tion. Thirdly, we inves­ti­gate whether the inde­pen­dence of irre­le­vant alter­na­tives (IIA) assump­tion holds. We find it holds for most people when deci­sions just involve wages, costs, and liqui­dity constraints. However, once we add a risk of unem­ploy­ment and incom­plete infor­ma­tion, IIA no longer holds for about 20 percent of our sample.

Riccardo Turati (UCLou­vain): Network-based Connec­ted­ness and the Diffu­sion of Cultural Traits 

This paper empi­ri­cally inves­ti­gates the impact of network-based connec­ted­ness on the diffu­sion of cultural traits. Using Gallup World Poll data over 148 coun­tries on indi­vi­dual connec­ted­ness, opinions and beliefs, we find that indi­vi­duals who have a connec­tion abroad are asso­ciated with higher levels of social beha­vior, reli­gio­sity and gender-egali­ta­rian atti­tudes. The effect is stronger among indi­vi­duals living in regions charac­te­rized by low levels of reli­gio­sity and gender-egali­ta­rian views, sugges­ting that migra­tion favors cultural conver­gence across regions along those traits. The effects are robust to connec­ted­ness diffu­sion, and country and indi­vi­dual open­ness towards foreign coun­tries. The cultural effects of connec­ted­ness on each trait are stronger among less educated indi­vi­duals rather then highly educated ones. The effects are also robust to a set of propen­sity score matching and cova­riates matching tech­niques, under­mi­ning the poten­tial threat driven by selec­tion into connec­ted­ness by obser­vables. Statis­tical tests are care­fully imple­mented to quan­tify the selec­tion threat driven by unob­served factors, which appears negli­gible. The effects are sizeable on social beha­vior and gender-egali­ta­rian views, parti­cu­larly on low educated indi­vi­duals, once simu­la­tions based on esti­mated coef­fi­cients are performed. Although robust, the pro-reli­gio­sity effect of connec­ted­ness is limited and negligible.