Jings, an actual discovery
I actually did find out something new...
Continuing my recap of the research work behind my forthcoming report The Foundations of the Digital State, its time to look at the New! Thing! the actual discovery.
(The report launch will be Monday 18th November, in the Bayes Centre and online, keep a slot free in your diary).
All research starts with some hypothesis and my Research Fellowship was no different.
There were 2 axioms:
The Scottish political class are not good enough customers of digital services (yet). Prior to 1983 almost no legislation anywhere in the world led to the creation of a computer system, now almost no legislation doesn’t. Systems do not optimally do things they were not designed to.
We already have the answers, we don’t necessarily know it.
and 4 hypotheses:
Continuous improvement via process re-ordering will identify defects earlier, reduce iteration time and lead to better outcomes.
The legislative process contains hidden barriers to modernisation.
Legislation implicitly defines data and hence processes in the administrative state.
The audit structure for administrative legislation needs to be extended.
The thinking was that the outcomes would be synthetic, taking other peoples work and experience and systematising it - like Northecote-Trevalyn.
But I did say I would go where the evidence led - and so it proved.
John Sheridan, he of legislation.gov.uk fame, said in an aside to me that I should investigate the secondary legislation and commencement dates of Universal Credit.
So we all knows what happens when someone who really knows what they’re talking about tells you to think about something you hadn’t considered and take it seriously, right?
Yes that’s correct, I totally ignored his advice and carried about on my merry way, until…
Long story short, about 6 weeks later, I find myself reading all the 79 bits of social security legislation for Scottish Social Security and muttering “sorry John, you were right John, my bad John”.
And that’s where I found out the thing I hadn’t planned on looking for.
It’s tricky because the thing doesn’t exist, and its the non-existance that the big thing of Foundations of the Digital State.
It hinges on the difference between functional and non-functional (or infrastructural) specifications.
In the old world, say the 1950s, the infrastructure for social security (buildings with canteens, roofs and windows, on a bus route) was totally decoupled from the functionality of the administrative process (forms and calculations).
In the new world there are additional non-functional and infrastructure considerations: sign-ons, databases and backup, cybersecurity, joined-up government and data sharing. This is tightly coupled with the functional administrative software but lacks any institutional deliberative and enforcement apparatus comparable to legislation and judicial review.
The bit that is missing in the modern state is the institutional structures to develop and make coherent non-functional/infrastructural specification work.
Working Paper 9 has the evidence and Working Paper 0 has a sketch of the institutional solution to the problem.
More recaps to follow.