15 - Data Journalism: What’s Feminism Got to Do With I.T.?

Summary

Abstract

Taking a feminist approach to data journalism means tuning in to the ways in which inequality enters databases and algorithms, as well as developing strategies to mitigate those biases.

Keywords: data journalism, feminism, gender, ethics, inequality, databases

Because of advances in technology over the last 70 years, people can store and process more information than ever before. The most successful technology companies in the world—Google, Facebook, Amazon, Microsoft, Apple—make their money by aggregating data. In business and government, it is increasingly valued to make “data-driven” decisions. Data are powerful— because they are financially lucrative and valued by the powerful—but they are not distributed equally, nor are the skills to work with them, nor are the technological resources required to store and process them. The people that work with data are not representative of the general population—they are disproportionately male, white, from the Global North and highly educated.

Precisely because of these basic inequalities in the data ecosystem, taking a feminist approach to data journalism can be helpful to uncover hidden biases in the information pipeline. Feminism can be simply defined as the belief in the social, political and economic equality of the sexes and organized activity on behalf of that belief. Feminist concepts and tools can be helpful for interrogating social power using gender as a central (but not the only) dimension of analysis. One of the defining features of contemporary feminism is its insistence on intersectionality—the idea that we must consider not only sexism, but also racism, classism, ableism and other structural forces in thinking about how power imbalances can obscure the truth. For journalists who identify with the profession's convention of “speaking truth to power,” a feminist approach may feel quite familiar.

This essay looks across several stages in the data-processing pipeline— data collection, data context and data communication—and points out pitfalls for bias as well as opportunities for employing a feminist lens. Note that a feminist approach is not only useful for data pertaining to women or gender issues, but suitable for any project about human beings or human institutions (read: pretty much every project), because where you have humans you have social inequality.