Over the last year of my research a key theme has emerged - the split between how the state creates and cares about the functional requirements of digital systems that under pin services, and the non-functional requirements. To coin a phrase I have staked a lot on that analytical difference.
My penultimate working paper before I start writing up takes a proper go-back-stake-the-undead approach to this.
The good news for you is that you don’t need to read it - nor the appendix neither - as the results are so clear.
First lets recap - functional requirements are things that are specific to a particular system: what information is required, what decisions will be taken, how things are to be calculated, what processes are to be followed.
Non-functional requirements on the other hand are much more generic - and it can be hard to work out at first blush which particular system or service they pertain to: there will be a database, the staff will sign on. And critically joined-up government and data sharing.
There are non-digital non-functional requirements too: there will be building, it will have a roof and a canteen, there will be bus stop nearby serving a population that can be hired.
The premise is that the government is organised for the pre-digital world when the functional and non-functional requirements are almost totally uncoupled, and not the digital age when they are tightly enmeshed.
If that is the case then looking at the legislation of a major digital system - in this case Scottish Social Security - should not turn up a trace of non-functional requirements.
Simple task - read the legislation and see. There are 3 bits of primary legislation and 76 pieces of secondary legislation.
So I read ‘em. And for each section I marked it in one of four ways:
non-specification 725
functional specification 419
endo non-functional specification 3
exo non-functional specification 2
It’s all functional.
(Exo non-functionals specify the relationship of a social security system to other government systems, and endo’s specify the relationship of different social security sub-systems to each other.)
I wouldn’t bother reading the attached working paper unless you want to see the qualifications on the methodology. I am so sure you that you don’t want to read the 85 page appendix that has all the gory details that I’m not attaching it. If you are that desperate you can work out how to contact me.
Only 1 more paper to finish - on experimental digital legislative processes. Upward on onward.
Digital state systems like social security need to be joined up to other systems - but they’re not. - is that what you are saying @Gordon Guthrie. Asking for the remedial class.