International Choice Modelling Conference, International Choice Modelling Conference 2017

Font Size: 
The influence of individual husband’s and wife’s preferences on joint parental school choice for their children
Petr Mariel, Riccardo Scarpa, Ainhoa Vega-Bayo

Last modified: 28 March 2017


The objective of this paper is to study school choice as a deliberate joint parental decision. This decision is affected by the underlying husband and wife’s preferences and it is analysed by using a bargaining discrete choice model, applied on a data set that has been obtained by carrying out a survey in the metropolitan area of Bilbao, Spain. To the best of our knowledge, this is the first application of bargaining discrete choice model to the school choice topic.

Choosing a school for their children is definitely one of the most important and hardest decisions parents face. It is widely known that school type is related to the variation in students’ grades and that the school decision directly affects the social environment of the children. The vast quantitative empirical literature on the subject of parental school choice usually focuses its attention on the relationship between the socio-economic characteristics of the families and type of school (normally, private or public).

The objective of this study is slightly different, because we aim to analyse the school choice as a deliberate joint parental decision, which is affected by the underlying husband and wife’s preferences, by means of both stated and revealed preference information collected via a specific survey in the metropolitan area of Bilbao, Spain. Analyses of the differences between individual and household preferences in the context of stated preference methods have been conducted in many other different fields such as transport, marketing or environmental economics but not school choice. The specific contribution of this paper is thus the first application of a bargaining discrete choice model to school choice.

Within this framework we apply an innovative modelling approach of a modified bargaining discrete choice model with random bargaining coefficients in order to handle the different nature of stated preference (SP) individual data (husband and wife’s preferences) and revealed preference (RP) household data (the school actually chosen in real life). Our approach deviates from the standard bargaining model used in the literature because, generally, the joint choices are expressed in a hypothetical context and do not represent the real behaviour of the couple.

We assume that members of a couple bargain over their joint evaluations of the alternatives on the basis of their respective individual utilities. That is why the proposed joint decision model includes a bargaining parameter specific to each attribute. The joint parameter of each attribute is therefore defined as a linear combination of the SP coefficients. As the parents’ decision on school is a RP choice, the corresponding parameters in the joint part of the model can deviate from the proposed linear combination of the SP coefficients due to the different nature of the data. This is why we add, to each coefficient, a new error term. These error terms convert the coefficients of the joint decision part of the model into random parameters and should represent the difference between SP and RP data.

The first result we obtain is that the results representing stated preferences of the couple do not deviate drastically from the revealed preferences. This is probably due to the fact that the school choice is one of the most important and hardest decisions parents face in their life and they took the stated part of the questionnaire seriously. Another reason related to the previous one is probably the fact that we received only voluntarily filled questionnaires without supervision and sent altruistically by post without any financial profit. The second interesting result is the prevalence of husband’s preferences in the school choice. This appears to go somewhat against other previous findings, which suggest that mothers are on average more involved in school choice.


Conference registration is required in order to view papers.