Measuring representations of gender and race in children's content and understanding their impact on inequality in educational outcomes

As a child goes through life, her beliefs are shaped by what she comes in contact with. This affects what she sees as her strengths and weaknesses, and what paths she thinks hold more or less promise for her future. A large body of research shows that bias – the notion that one group defined, for example, by gender, ethnicity, or sexuality, is inferior to another on some dimension – harms children through affecting their beliefs about themselves as well as their effort in school and hobbies. Because learning builds on itself, small differences in enthusiasm or effort induced by bias can compound over time and have large negative consequences later in life. In this project, I will further our understanding of the transmission of gender bias via the early sources of information a child comes in contact with. First, I will use a novel method to to quantify the extent of gender bias the child encounters early in life. Then, using modern causal inference techniques and a large administrative data set, I will estimate how increased exposure to such bias affects her performance in school and beyond.