Child-Gender Preference Generalized Maximum Entropy Approach

Supanika Leurcharusmee, Jirakom Sirisrisakulchai, Natthaphat Kingnetr, Songsak Sriboonchitta

Abstract

This study measures child-gender preference in Thailand by examining the probability of having an additional child given genders of the previous children using the generalized maximum entropy (GME) approach. The GME method is robust to multicollinearity problem, which allows us to examine different child-gender preference pattern among families with different characteristics. The results show that girls are preferred in the families where the mothers have no high school degree. However, boys are preferred in the families where the mothers have a high school degree.  Moreover, regardless of mothers' levels of education, the results show the evidence of mixed gender preference. Thai parents are less likely to have an additional child once they already have at least one boy and one girl.