Fixing Diversity Diversity

Some job seekers are already developing ways to subvert the system by trading answers to employers’ tests and by creating fake applications as informal audits of their own. In fact one HR employee for a major company recommends “slipping the words ‘Oxford’ or ‘Cambridge’ into a CV in invisible white text, to pass the automated screening.”16 In terms of a more collective response, a federation of trade unions called UNI Global Union has developed “a charter of digital rights for workers touching on automated and AI-based decisions, to be included in bargaining agreements.”

The danger of New Jim Code impartiality is the neglect of ongoing inequity perpetuated by colorblind designs. In this context, algorithms may not be just a veneer that covers historical fault lines. They also seem to be streamlining discrimination – making it easier to sift, sort, and justify why tomorrow’s workforce continues to be racially stratified. Algorithmic neutrality reproduces algorithmically sustained discrimination.

Fixing Diversity Diversity Inc., a company I introduced to you earlier, does not employ fancy machine learning algorithms but statistics that use people’s first and last names to predict their ethnicity. Ethno-racially coded names are a resource for computer coding and thus a valuable commodity in many other industries that are looking to tailor their products to different groups. One of the things I find fascinating in this context is how ethnicity operates as a proxy for individuality. So, when people who work in direct marketing talk about ethnic targeting, they see this as acknowledging individuals, not groups. To them, it is the opposite of making vague generalizations, even when ethnic stereotypes are employed. For example, one industry insider who uses names to predict ethnicity explains: “There is no such thing as an average Joe or Jane, Asian Americans have different values than Jewish Americans, if you know someone’s ethnicity or religion, you can predict quite a few things.”