How database schemas from the 1950s through the 1990s, which encoded gender as a binary value, are affecting non-binary persons today
Meredith Broussard / Slate : Tweets: @libshipwreck , @futuretensenow , @bmatb , @schemaly , @roavingxss , and @espiers Tweets: Librarianshipwreck / @libshipwreck : “Over and over again, we have seen technical problems arise because the people designing computer systems were committed to replicating a rigid, retrograde status quo.” Great piece by @merbroussard on how biases get built into technologies: https://slate.com/... Future Tense / @futuretensenow : “Over and over again, we have seen technical problems arise because the people designing computer systems were committed to replicating a rigid, retrograde status quo,” writes @merbroussard https://slate.com/... Dr Ben Britton / @bmatb : “Most of these problems come from the way that 1950s U.S. and U.K. social perspectives informed how computer schemas were created.” - this resonates with some of the discussion circling about our increased reliance on machine learning trained with non-representative data https://twitter.com/... Soraya Chemaly / @schemaly : The next frontier in gender rights is inside databases. Bandukwala is right: Many computer scientists and engineers are personally and professionally committed to the gender binary and cisgender, heterosexual norms. https://slate.com/... G. Howler / @roavingxss : @Techmeme @merbroussard Predictable result. Imagine someone coding a product with 4 gender options in the 50s. He'd have his career destroyed and be chemically castrated like Turing. Elizabeth Spiders / @espiers : I often hear (white male cis) tech people claim technology is, by definition, egalitarian. We already have plenty of evidence that it's not true, starting with machine learning algos that are trained on data sets that reflect racist human biases. Here's another example: https://twitter.com/...
Context & Ripple Effects
This piece closes a loop on the coverage of who built computing. Earlier reporting traced how women helped create the field when it was considered 'women's work' and were then systematically pushed out of the field in the US and UK, with sexist undercurrents from Playboy imagery to biased personality tests cementing a homogeneous Silicon Valley culture. Broussard's argument is the downstream consequence: when a narrow cohort designs the foundational layer, its assumptions get encoded into infrastructure.
The specific mechanism here is schema design — databases built from the 1950s through the 1990s hard-coded gender as a binary field, replicating the social norm of their era. Because schemas are expensive to change and quietly propagate into every system built on top of them, a design decision made decades ago now surfaces whenever a non-binary person interacts with a form, record, or account.
First-order effects
- Non-binary people are the direct casualties: legacy binary fields misgender or exclude them at every touchpoint where an old schema still runs, forcing institutions either to misclassify them or build workarounds.
Second-order effects
- Organizations running these inherited systems face retrofit costs — remapping fields, migrating records, updating every downstream consumer of the data — turning a cultural blind spot into line-item technical debt.
Third-order effects
- If the pattern holds, schema-level bias gets treated like security debt: a design-stage governance problem that auditors and standards bodies flag before deployment, rather than something patched after users are harmed.
The trend: Infrastructure designed by a demographically narrow industry is being forced to absorb identity categories its creators never anticipated, one costly retrofit at a time.