How to Navigate Data’s Pink and Blue Problem
Gender associations are not universal, as data visualisation experts know
By Alan Smith
Once a year, graphic designers at the Office for National Statistics become particularly excited: the release of the latest baby name rankings heralds all kinds of amusing graphics that bring government statistics to social media. Many serious government statisticians may scoff (after all, it is not GDP), but it is a clever tactic.
Seven years ago, while working at the ONS, I got caught up in all this silliness and created a chart of the latest baby name rankings. It was pink and blue.
I remember agonising over the decision. But ultimately, I justified it because the strong cultural association of the colours meant I did not need to put a separate legend on the chart. This, in turn, meant my graphic would be “cleaner”, something to which all designers aspire.
I reasoned that if it had been a chart of unemployment data, I would not have made the same decision because it would have been an inappropriate reinforcement of adult gender stereotypes. But perhaps it was OK with babies. After all, a trip to the local gift-card shop would reveal a swath of pink and blue cards to celebrate the arrival of newborns.
