Learning to Spot the Revealing Gaps in Our Public Data Sets
The most vulnerable people, and acute social issues, can be near-invisible to policymakers
By Delphine Strauss
As art installations go, it is low key: a filing cabinet filled with meticulously labelled hanging folders. Visitors are welcome to browse under any heading that sparks their interest: publicly available gun trace data; the Nanjing massacre death toll; English language rules internalised by native speakers; how much Spotify pays each artist per play of song. The folders are all empty.
The work, titled “The Library of Missing Datasets”, is by Mimi Onuoha, an artist and adjunct professor at New York University. The aim, she says, is to expose the “blank spots in spaces that are otherwise saturated with data”. The blanks can reveal hidden biases in a society.
In my work as an economics reporter, I am used to sifting through statistical releases that do not quite answer the questions one might like them to: gross domestic product is a poor proxy for wellbeing; one measure of poverty tells a different story from another; a high employment rate does not necessarily ensure job security or an acceptable standard of living.
