Platformland and thoughts about trust and legibility

On Saturday my pre-ordered copy of Richard Pope’s Platformland arrived. I read it in an afternoon.

I had a bit of trepidation at the start of the book as it discussed the early days of Universal Credit. I’ve found as of late that books using early GOV.UK products as a reference can describe a world that feels different to what the people are experiencing now or even in the more recent past. I don’t know what Universal Credit is like these days, but the one thing that I noticed from my time at DWP Digital between 2019–2021, was that UC (as we called it) felt very cut off from everyone else. This was because the tooling choices, likely made in those early days, meant that they had different email, confluence and even Slack platforms from the of the Digital group. This is a fairly minor thing, but I think is perhaps a reminder that product ways of working can seem different both at different points in time but even between those on the team and those adjacent to it. Any different perspectives can be hugely helpful to head off this potential disconnect.

I did also feel my spidey-senses tingle at discussions of proactive life event updates. Whenever I hear these discussions, I think of a couple of stories I know of:

someone who was advised at a jobcentre to apply for parental support only to be told a few years later they had been overpaid and needed to repay several thousand pounds. They now have very little trust in government and are reticent to get anything else from government unless they completely understand the process.
someone who, whenever they need a DBS check for a new job, rings a special phone number. This is because they are trans: their records are only processed by a specialist group of people who are authorised to see their protected characteristics.

I often fear that the stories of proactive services assume the ‘happy path’ and ignore those where proactive services either have gone wrong or could go very wrong. That said, to Richards’s credit, he covers some of these concerns in the latter half of the book, where he suggests that in fact government should not be seamless but instead should help people understand law and policies. He even admits that GOV.UK’s early days probably did not understand this point and missed it. (That said, England is a particularly challenging country to do such work in, given that it is a country with an unwritten constitution with many of its rulings being based on precedent, but that’s for another day).

I also liked Richard’s notes about services patterns of journals, accounts, and tasks, as inspired by Universal Credit.

The one thing I’d have maybe liked a bit more about is trust. I already mentioned knowing people who do not trust government, but I think that there are other ones:

does the public’s trust in government affect how quickly a country can lean into initiatives like digital credentials? Someone, I can’t remember whom, had the intriguing comment that both Taiwan and Estonia—leading people in digital government activities—both also became independent from a neighbouring country relatively recently, meaning that inhabitants may trust its new government with their data. In comparison, England has famously been anti-register since World War 2 (and Richard notes that there was an early colonialism privilege of Britons travelling without papers) .
how much do government agencies trust each others’ data capture? Richard talks about how income can mean different things to different departments, but there’s also something about whether departments can take the risk of trusting another department’s data compared to doing it again themselves so that they can really trust it. Perhaps in the UK at least, GOV.UK One Login will manage some of this by holding data centrally: watch this space I guess?

All in all, I do think it’s a good book and worth reading. A call out to other people in government though, I’d love to see books that include more experiences of the not-so-happy paths—even based on author’s lived experience?–as for example is talked about in Design Justice, Data Action and The Tech That Comes Next. Admittedly this is a tricky one to navigate with professional codes and such, but a girl can dream.