This is our truth, tell us yours
Jem wrote brilliantly a couple of days ago about the half witted example a member of the armed forces used to illustrate data matching as an intelligence practice.
The first rule of data mining is you need to know what you’re looking for, or have good examples to use to enable machine learning to derive patterns.
Instead of looking for needles in haystacks, tracking bank transactions or retail purchases, if you want to stop trafficking you’d look for the haystacks, the properties where the victims of people trafficking are accommodated.
It’s not complicated, and would involve much smaller sets of data than all bank transactions, or all women from poor countries. Local authorities have data sets of all properties on which rates are paid, and all properties on which council tax is paid. Identifying properties where neither is paid, or where there’s a mismatch between public data sets about property would be a simple way of identifying addresses associated with trafficking.
Note that this is about trafficking, not sex work. Trafficking covers a range of industries, including agriculture and catering, not just sex work.
At heart in the article Jem cited though is an intersection of political tensions. The first is that the state doesn’t want to show its hand in terms of revealing how mass surveillance could work. The same sophisticated data matching that could find trafficking associated addresses would be ideal for clamping down on tax evasion. There is a long tradition of hypocrisy in Britain that allows surveillance against those who are othered, not those who are like us. Hence the rows about surveillance being used to counter the serial English middle class vice of lying to get your child into the right school. So data matching can only be used if it’s targeted at othered women, preferably foreign, not, for instance, the criminal gangmasters who use trafficked workers to pick cockles or make sure we have cheap eggs at Tescos.
The second point is that Britain has no tradition of having one body responsible for internal surveillance. You might think this is a good thing, but, in effect, the competing agencies making bids to be at the forefront of the surveillance of the future are engaged in a propaganda war full of exaggeration and dishonesty. At the same time surveillance is deployed to an extent against the poor and excluded that the well heeled and prosperous will not accept. If you don’t believe me, take the the example of schools admissions above, and contrast the way they’re handled compared with, say, a housing benefit application. Both involve the use and potential abuse of public resources, but surveillance is only acceptable for one of them. Before you think about a bank, a private institution,using data mining to track sex workers, think about that dichotomy.