I have to thank a nice, serendipitous Friday night in the pub with Sym, Rose, Tom and Anna for this train of thought. They prompted me to think about all this stuff, but my constitution of opinions here isn’t necessarily theirs.
What’s going on?
Soon, quite soon, or at least by 24th January 2025, we’re going to have another election. It’s never a done deal, but it seems quite possible that there might be a quite comprehensive change of leadership. Before the 2010 general election, a fair few people involved in mySociety, Channel 4, the BBC and other groups with a stake in the government being a bit more modern and a bit less shit got involved with trying to advise the incoming Conservative administration. This wasn’t entirely well received by the ministers in the Labour government they had also been advising.
And now, a few people from the current civic tech/open gov movement are having meetings with SpAds. So, for all of us not invited into the tent, what’s left to ponder is “What should Labour’s government technology policy actually be?”.
You’ve never had it so good
Half the problem is that a lot of the suggestion is basically to keep going. It’s boring, deeply difficult to get money for something that isn’t on fire. GOV.UK is… fine? Boring, fine, needs more money, content could be better, but broadly it’s not awful.
Compared with the position in 2010 where some online services kept office hours and turned off overnight or the infamous stories of £10k per word change on Home Office sites, it’s not that bad now. We have for the most part, succeeded in getting modern-ish approaches to digital services integrated. And that makes it hard to hang a manifesto on digital services as objects in their own right.
Except, surely we’ve all had a service that falls apart? Anything where you’re doing something for your kid and the service can’t decide on grammar, policies that are designed with adminstrative burdens to penalise the service user for political or financial reasons (30 hours childcare allowance/Universal Credit), anything that had Verify on it (I never successfully logged in to Verify, even having had a Barclays account and working on digital ID policy). There is still loads to be done, loads to be improved and maintained. There is no end to the work, because the work of digital government is the work of government.
Design constraints
Labour have committed to following the current government’s spending plans on taking office. This basically means no planning for spending money, and instead having to do it as an emergency when the thing that was meant to be replaced breaks. So, the government may end up being quite unpopular with its base as it is unable to change fiscal course, and unpopular with the right for not being hang-em-and-flog-em enough, although the Party seems to be keen on emphasising that it has no empathy so maybe that’ll win over Tory voters. It’s also very likely that there will be significant tax rises during the next Parliament if Labour keeps to pre-existing plans.
So, Labour might get a decent majority, a fractious and poor public, and might end up having to enact unpopular things left for them by the previous government. It is entirely plausible that they might not get a second term.
So what’s this got to do with digital services, beyond the obvious question of who pays the bill?
Missions: impossible?
We’re deep in the AI hype bubble. We’d only just finished the crypto one, but now, here we are again, everything needs an AI on it. And I’m sure that’s tempting in govtech, after all, what sounds better than replacing boring work with a Computer That Can _Really_ Think! (except when it can’t)? Unlike “putting benefits on the chain” or whatever, AI decision making is already and will continue to be awful for users. There is no user need: it is all business need. Users need sensible, auditable decisions that are flexible to circumstances and are kind, fair and considerate. That is not AI’s vibe. I also worry with AI that it priveleges the technology rather than the outcome. When the government put so much time into a Cloud First strategy, that was to unlock the ability to do better services. I don’t see that with AI, just a rush to the new thing before you get the so that… part of the user story nailed.
If everything had followed the hype in the 2010 government, then “Facebook, but for Government” would have been built, and it wasn’t. Not from lack of people asking for it in the mistaken belief it would be A Good Thing (I remember a lot of client requests at the time being very twitter/facebook themed “please make this social media about planning rule changes go viral” etc etc). But, people who built twitter helped to come in and start conversations about architecting massive services with loads of concurrent users and scalable infrastructure. That’s what you need, the expertise with the underlying structures, not always the shiny hype thing.
So, assuming that we can avoid the hype cycle, perhaps we could get digital services in a manifesto to where they need to be: part of the delivery of government services. Imagine bringing people in to deliver a Green New Deal style raft of policies. More appealing for many people in tech than the current raft of core government policies.
So,what if the technology strategy of the next government was actually dependent on the policy goals that they want to achieve? What if we phrased a technology policy around what vision the new government is trying to achieve? That might be a way for technology to stop being the last thing on the list, a delivery problem rather than a policy constraint and it could even allow for some level of joined up imagination around data and service design. We can always hope.
Post script: if you’re from Labour and read this, there are tonnes of good things you could do with what’s already there. Make service assessment reports part of the list of things that select committees scrutine. Involve service user representatives in them. Make the practice of delivering policy more open, more willing to iterate. Send me an email.
