Have you ever questioned the nature of your reality? No? Well, this isn’t Westworld, and I wouldn’t have expected you to. I certainly hadn’t until I decided to read Invisible Women.
Invisible Women: Exposing Data Bias in a World Designed for Men by Caroline Criado-Perez explores the implicit biases that plague women in all facets of their lives. As data scientists, our goal is to find the most efficient way to do things. Oftentimes this means that if there is an accepted convention that has been shown to work, why fix it? It turns out, however, that many current structures and methods of doing things are engineered—often without anyone realizing it—to cause systemic problems.
In an effort to encourage you all to read this amazing book, I will primarily focus on how this information changed the way that I approach my life and future career as a data scientist. I condensed my thoughts into three major takeaways: intention, input, and audience.
The implications of Criado’s message are easiest to understand through one of her first examples: public transportation.
Public transportation was designed to allow all people residing in densely populated areas to travel freely. When these systems were created, they were intended to transport workers into the city center where their jobs were located. At that time, the majority of travelers were men. Today, however, we find that these systems are predominantly used by women.
“In France, two-thirds of public transport passengers are women; in Philadelphia and Chicago in the US, the figure is 64% and 62%, respectively. Meanwhile, men around the world are more likely to drive” (Criado, p. 45).
Around the world, men are more likely to own cars and/or take the family vehicle to work. As a result, women and/or lower-income individuals rely more heavily on public transportation like buses, trains, or subways. As a consequence of that disconnect, it is difficult to truly optimize the public transport system for its current passengers.
By not recognizing the demographic most utilizing the service, there is a divergence between intention and the service provided. Like many industries, transportation departments are historically male-dominated. The absence of diversity of thought has consequences.
As it turns out, female travelers have very different needs and take alternate travel paths than male travelers. Rather than go strictly from point A to B, female travelers tend to make multiple stops: grocery store, errands, childcare. The current systems that employ pay-per-trip nature and the radial design of routes make traveling more expensive and cumbersome for female travelers.
Looking beyond the inconveniences that the current system creates for women, there are serious causes for concern. Safety is the biggest concern for female travelers. The numerous transfers and wait times associated create additional vulnerabilities for female travelers. Thus far, there has been little effort to optimize the transportation system based on safety, the most important consideration for the passengers using public transportation.
As it turns out, something as basic as the public transportation system can be loaded with biases that deeply hinder the lives of women around the world. The easiest solution that Criado points out is to look closer at data on women and consider gendered differences throughout the process. Regarding public transportation, these problems could have been identified, remedied, and optimized had the officials and data scientists considered the female transportation experience.
As data professionals, this book provides a new lens to sharpen our analytical and problem-solving skills.
Intention, input, and audience are now a part of my list of considerations when approaching any project. I would urge anyone interested in equity, women’s rights, or simply data science to make the time to read Invisible Women: Exposing Data Bias in a World Designed for Men.
Columnist: Camille Carter