Gender Disparities in Child Custody Sentencing in Spain: a Data Driven Analysis

In our work, we investigate biases in judicial decisions using data analytics. Specifically, we are interested in analyzing the impact of the gender of both the judge and the plaintiff on the probability of winning a case. With this aim, we analyze a dataset comprising information from over one thousand second-instance appeals for child custody in Spain. Our results indicate significant differences in how legal arguments and facts are utilized in the final sentences, depending on the gender of the plaintiff. We also examine the impact of the requested type of custody (sole or joint) on the probability of winning a case, with a focus on its relationship to the plaintiff's gender. Moreover, our analysis reveals statistically significant differences in the winning probability of a case depending on the judge's gender. To further understand these findings, we conduct additional analysis to establish the causal relationship between the judge's gender and the probability of winning, revealing weak but consistent patterns. Our research provides a consistent methodology for evaluating biases in judicial systems and offers intriguing insights into the context of child custody in Spain.