A Count Data Model with Social Interactions

Vous êtes ici

Accueil » A Count Data Model with Social Interactions
14 Octobre 2020
Types de publication: 
Cahier de recherche
Auteur(s): 
Elysee Aristide Houndetoungan
Axe de recherche: 
Enjeux économiques et financiers
Mots-clés: 
Discret model
Social Networks
Bayesian game
Rational expectations
Network formation
Classification JEL: 
C25
C31
C73
D84
D85

I present a model of peer effects in which the dependent variable takes integer values. I present an incomplete information game rationalizing the model, and I provide sufficient conditions under which the equilibrium of the game is unique. I estimate the model’s parameters using the Nested Partial Likelihood method. I show that the counting nature of the dependent variable is important and that assuming incorrectly that it is continuous significantly underestimates the peer effects. I estimate peer effects on the the number of extracurricular activities in which students are enrolled. Increasing the number of activities in which friends are enrolled by one implies an increase of 0.295 in the number of activities in which students are enrolled, when controlling for network endogeneity. Ignoring the endogeneity of the network overestimates the peer effects. 
 

Contact: 

Elysée Aristide Houndetoungan : Université Laval, Department of Economics.


I would like to thank Vincent Boucher for his helpful comments and insights. I would also like to thank Bernard Fortin, Arnaud Dufays, Luc Bissonnette, and Maripier Isabelle for helpful comments and discussions. Thank you also to the participants of the Applied Young Economists webinar at Monash University. I provide an easy-to-use R package—named CDatanet—for implementing the model and methods used in this paper.

The package is located at github.com/ahoundetoungan/CDatanet. 

This research uses data from Add Health, a program directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill, and funded by Grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. Special acknowledgment is given to Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. Information on how to obtain Add Health data files is available on the Add Health website (www.cpc.unc.edu/addhealth). No direct support was received for this research.