The datafication of race and ethnicity

The New York Times published a fascinating overview of the American census forms since the late 18th century. It shows how the form keeps trying to ‘capture’ the country’s demographics, “creating and reshaping the ever-changing views of racial and ethnic identity.”

These ever-changing ways of categorizing people find their way into policies and governmental institutions. In the US, the ‘ethnicities’ are a fixed list, but you can at least self-identify and pick the (literal) boxes that most fit you. In the Netherlands, the Central Bureau of Statistics (CBS) up until very recently divided people into non-immigrants and immigrants (depending on where your parents were born), with the latter group being subdivided into ‘non-Western’ and ‘Western’ immigration backgrounds according to some weird post-colonial logic. Gerwin van Schie has shown (see his doctoral thesis (PDF)) how these categorizations lead to discrimination. As he says (machine translated): If a data system uses information about migration background or ethnicity for evaluation or prediction, discriminatory outcomes are not an accident but an inherent part of the design. In 2022, the CBS started using a new way of categorizing Dutch citizens. You can only be datafied as an immigrant if you weren’t born here (but you can be a child of an immigrant). They have also dropped the Western/Non-Western binary, now using Europe and non-Europe instead, while ensuring they can track the ‘classic’ migration countries Turkey, Morocco, Surinam, Indonesia, and the Dutch Caribbean. This is what the Netherlands looks like from the viewpoint of the CBS:

These categorizations still leave the door wide open for using them in all sorts of governmental tooling and modeling. Even with the very recent attention to the possible stigmatizing effects of using migration status, we can’t be surprised if we find discrimination in these systems in the near future. See: An American Puzzle: Fitting Race in a Box at the New York Times, Indeling van burgers op basis van migratieachtergrond kan leiden tot discriminatie at Universiteit Utrecht, and CBS introduceert nieuwe indeling bevolking naar herkomst at CBS.

https://racismandtechnology.center/2024/06/22/the-datafication-of-race-and-ethnicity/