GOV.UK Accounts — thinking about the future

I’ve been thinking a bit recently about GDS’s recent statements about creating “GOV.UK accounts” — this one introducing the concept and this one on related user research.

It’s very interesting that, on face value at least, this feels like a complete reversal of GDS’s longstanding policy of transactional services not requiring a user to log in. I’m positive that back in 2012 there was a convincing amount of evidence that accounts were neither user friendly or required.

To be clear up-front — I’ve always thought that user accounts were a good idea. It’s important to note though that I’m a million miles away from being a user researcher or service designer. It’s entirely possible that we’re talking about a new level of complexity for transactions, thanks to things like the ever-increasing number of government APIs, hopefully leading to the kind of thing Tom Loosemore (in conjunction with Richard Pope) was showing off back in 2015 (it’s all great but the really good stuff kicks in about 25 minutes in).

As Tom says in that talk, the Estonians have been doing this kind of thing for years — specifically by creating X-Road, the missing data layer we don’t have (yet) in the UK. It’ll be important to see what the new Government Chief Data Officer, as and when they are appointed, has to say about separating data from services within departments and what kind of data infrastructure we want to put in place. I’m very much hoping one where not only can every citizen start to see what information government holds on them via their shiny new government accounts, but also that records every time a piece of data is shared between organisations (I’m not saying *cough*blockchain*cough* but…).

What I’ve actually been thinking about most though is what all this means for the “tell us once” principal and the NO2ID fight against ID cards and centralised databases all the way back in 2010. Many people (some but not all of policy wonks, service designers, developers, data folks, etc) would love there to be single canonical sources of citizen data. In Estonia that’s the law — if you have to provide the same piece of information more than once the government has to update the system.

However, the arguments from 2010 are as powerful today as they were then. Centralised data is easier to steal, leak, corrupt and generally mishandle. In particular there’s been an unspoken understanding in government that even if specific organisations hold data related to their area of authority (driving licence numbers at the DVLA, benefits info at DWP, health data — well, that’s a very complex story) there’s never been a push to have one central authority storing basic citizen information like name, address, date-of-birth, etc. Even GDS Registers, the failed attempt to produce some simple reusable sets of data that I was part of, only looked at information relating to non-personal information such as lists of countries or school locations.

I’m not as sure as I was ten years ago that centralising common data is a bad idea, especially if data entered for one reason highlights that citizens are entitled to other benefits or services they may not know about. However, the people who say it is are very smart and should definitely by listened to.

Creating GOV.UK Accounts is a very small first step on a long journey. We could be anything up to a decade from something like X-Road, even if we decided tomorrow that was going to be the policy going forward.

All I’m saying at this point is that there needs to be a lot more opening up of this conversation once the new Government Chief Digital Officer and Government Chief Data Officer are in place.