Read on Twitter
Fun little nugget I just discovered in the National Data Strategy is the action in s.4.2.1 to “implement the recommendations of the Joined up data in government” paper #brief-summary">https://www.gov.uk/government/publications/joined-up-data-in-government-the-future-of-data-linking-methods/joined-up-data-in-government-the-future-of-data-linkage-methods #brief-summary">https://www.gov.uk/governmen...
The paper describes itself as setting out to, "engage with the data linkage community across government, academia, the third sector and internationally to understand challenges faced and identify state-of-the-art data linking methods to help realise those benefits."
But I’d go as far as saying it’s a wolf in sheep’s clothing. It comes across as a fairly technocratic guide to improving methodologies. Privacy experts are better placed than I to assess its approach to Privacy-Preserving Record Linkage. But - dear reader - it buries the lede.
What exactly is data linkage across government? Well, it sets it out quite clearly in this example of data sharing between the MoJ and DfE to increase "understanding of the links between childhood characteristics, education outcomes and (re)-offending…"
To continue, data linkage, "will assist in identifying the population that requires support through early intervention and evaluating these projects to understand whether they are effective."
There is no mention that I can see of data ethics, oversight, limitations on how this data will be used. More efficient researchers are presumably already aware of this work, but the reference is buried in an auxiliary data methods paper; you have to work hard to find it.
I& #39;d expect this to be accompanied by extensive protections and safeguards, with redress methods built in. I can’t find any mention of these in either paper, despite the commitment in the strategy to ensure data helps create “A fairer society for all”.
There is, however, a commitment in s.4.2.1 to implement an Integrated Data Platform for govt to make this kind of analysis easier - wch sounds *a lot* like Palantir’s Foundry.
Given yesterday’s revelation that positive Covid test data will be passed to the police, I can’t see any reason to think this data will be used benevolently. Kristian Lum and @wsisaac’s 2016 paper “To Predict and Serve” is a useful cautionary tale here: https://rss.onlinelibrary.wiley.com/doi/full/10.1111/j.1740-9713.2016.00960.x">https://rss.onlinelibrary.wiley.com/doi/full/...
As is the sorry tale of the London Gangs Matrix https://www.wired.co.uk/article/gangs-matrix-violence-london-predictive-policing">https://www.wired.co.uk/article/g... The National Data Strategy says, "Data can be used to hold a mirror up to society - to understand how different groups are faring”, yet there is no mention of mitigating structural racism in any of the data.
