A test of a unitary model on labour supply using the information of household decision-making systems

This article tests whether a unitary model is consistent with household behaviour using the data of two-earner couples. It focuses on the unitary model assuming that all family members have the same utility function. The analysis investigates the difference in a husband's and wife's labour supply between the household that determines the wife to be the main decision-maker and the household that selects a different decision-making system under the control of individual and household characteristics. The estimation employs a treatment effects model to consider the selectivity bias caused by unmeasured characteristics. Results show that the household with the wife as the main decision-maker increases the husband's working hours by 15% and decreases the wife's working hours by 59%, compared to the household that selects a different decision-making system. This implies that the unitary model is rejected. Additionally, the husband's wage rate, the husband's and wife's health status, and their gambling addiction determine the household decision-making system such as the variables that determine the reservation utility of not being married. The effect of the decision-making system on the labour supply and that of the determinant factors on the decision-making system are consistent with the implications obtained from Nash bargaining models and collective models.