Mar 18, 2016
Shall we just do it? Just build it and get this over with?
We have it anyway, don’t we? Just in a distributed and not-very-accountable way. So why not do it properly?
The stuff I wrote yesterday about registers is just a part of a vastly bigger story about information, people, and government.
[tl;dr of that piece: using ‘registers’ – lists of authoritative data – to make government services better has lots of benefits, and raises interesting questions]
It’s a story that’s so big it doesn’t really have a beginning, or an end. How we meet the needs of people, society, democracy, everything – with technology, data, organisations, everything.
So I’ll home straight in on one part. Probably the most sensitive registry of all would be a register of citizens. Of people. Of the entitled-to-vote. Of permanent residents. Yes, tricky, hey? Let’s just call it people.
The Promised Land of a canonical list of people sat (sits?) behind the for-the-moment-abandoned (I expect this to change/is changing!) concept of a national identity card.
It sits behind lots of other things too – either as the manifestation of the ultimate authoritarian state, or as the lubricant for a trillion safer, more secure, more efficient digital transactions. Depends on who you ask, what they’re trying to sell, and the weight they give to various arguments of logic, experience, ideology and emotion.
It’s hugely political, obviously. The argument that it is “poor civic hygiene” is usually high on the list of “why nots”. A future government may be in a position to do all sorts of terrible things to its people if it can track and target information very precisely at individual level, or even make people appear and disappear at will, through manipulating a central megadatabase.
And that’s to say it’s even possible to procure, build and operate such a beast. The track record at this scale isn’t great.
It’s so sensitive that registers of personal or sensitive data have been explicitly excluded from the current scope. Instead, Verify is doing sterling work to do digital identity checking through the use of third parties – essentially using what outside organisations know about people as a proxy for government’s knowledge, then accepting that trust as being good enough for subsequent interactions with government. A very neat, and widely welcomed, sidestep around the problems and concerns that bedevil a central people register. But it has limitations – you can use it to check facts about people, but you can’t write information back to it, or assemble a master list of people you could then sign up for electronic voting (or any other new thing you dreamed up).
So none of this means that the clamour for a central person register has gone away. It never will. It’s what James Randi once described as an “unsinkable rubber duck.” An idea that no matter how many times you unpack it, debunk it, resolve it…will always bob back to the surface. It’s so tempting. The perfect answer for those who love hierarchy and are convinced that hard-edged systems can save the world. (But Estonia!)
Yes, yes, ok, Estonia etc. – there needs to be a better response available to the “But Estonians”. Your vulnerable minister and officials will be regularly swept over there to marvel at how all this digital identity and database stuff just…works. Nobody dies because of it, the tanks don’t roll in, there isn’t a monitoring screen in every house. I’ve asked a lot of people who should know about this stuff what the solid counter should be to the But Estonians. Curiously, I haven’t found one yet. Have you?
And then, I think – hang on, is any of this resistance actually meaningful?
We may not have a single people register, but we have lots of things that are a lot like it. You may be surprised by some of the questions you get asked when you use Verify. How did they know that? They know lots, really, those identity providers. That’s why they’re identity providers. They’ve spent years buying and integrating things about you. It helps commerce operate. But it’s private, opaque, unaccountable. Sure, it’s not government, but it’s still a thing.
Or what about the Police National Computer? Who knows how they refer to you? But they know things about you. Try getting stopped in the street by the cops and not showing any “ID” (don’t start me…) You’ll find some of their questions to you, and radio checking, pretty interesting too.
So whether it’s done through a single unique identifier (ooh – somebody said “just use the National Insurance Number! DRINK!) or through the patchwork of private and occult registers, we live in a database state anyway. The infrastructure, and the surveillance powers, are already such that pretty much any bad consequence could already happen (is happening?). Data sharing work is developing apace. If one of the main concerns about a centralised people register is its vulnerability to attack, then those concerns apply to the private registers too, no? Ok, but the prize is bigger, but still… The police manage to do it. Experian manage to do it.
Is all the protestation just for show, really – we attack the thing we’ll be able to see because we can’t attack the things we can’t?
My personal view on this (as a non-practising civilian with a lifelong interest in civic data) is that the central register has some benefits. But enormous risks. And that the risks scale faster than the benefits. You aggregate that much in one place and the consequences of error, or breach, or yes, totalitarianism, are unthinkable. So it’s a bad thing.
My friends Achilles and Tortoise teased out some of these issues for me a while ago.
But I’m not convinced I’m right. That would require a level of evidence I don’t have, or a level of ideology I find distasteful.
Help me out here – what would it really take to sink, or float, that rubber duck?
At least for a bit?