CMStatistics 2018: Start Registration
View Submission - CMStatistics
Title: Looking for gender bias Authors:  Javier de Vicente Maldonado - Carlos III University (Spain) [presenting]
Abstract: Measuring the possible bias in the allocation of intrahousehold resources is a difficult exercise, mainly due to the particular nature of the household expenditure data. However, an accurate assessment could lead to the identification of gender discrimination which, at the same time, would be a key factor for women empowerment and economic development. We represent the latent Engel curves, i.e. the substantial drivers of consumption patterns, as an approximate factor model in which the factors are extracted using an algorithm based on the maximum likelihood method. In addition, we propose a new measure of gender bias called latent outlay equivalence ratio based on the original procedure proposed previously. Finally, we illustrate this new approach using the commonly used data from the 1889/90 US Bureau of Labor report, which consists of 1024 budgets of British families in the textile, coal-mining and metal manufacturing industries. We clearly identify the existence of two latent curves that can be defined as basic necessities (e.g. food) and luxuries (e.g. alcohol) respectively. In contrast to the results obtained in previous analysis, we find strong evidence of gender discrimination.