Saturday, 21 March 2015

The system is fine. It's the users that don't work



It has fallen to Bryan Glick, the estimable editor of Computer Weekly, to perform the first post mortem on the Rural Payments Agency's (RPA) computerised Basic Payment Scheme (BPS) which was discontinued yesterday and replaced with paper – "successive software releases failed to resolve the problems with the mapping tool".

We have known for two years that Mr Mike Bracken, the executive director of the Government Digital Service (GDS), was heavily involved in the development of BPS. "I'm on the Board", as he told us, "I'm trying to help them every week ... GDS will be working very closely with them ... it's going to help us deal with Europe in a different way, and quite rightly we're building it as a platform. It's going to be another example of government as a platform".

Mr Glick reveals that in addition, Mr Liam Maxwell, the Government's Chief Technology Officer is also the senior responsible owner of BPS, "overseeing progress", and that this is a mark of "the importance of RPA to the GDS strategy".

These are highly respected people:
Highly respected, but it doesn't seem to have helped with BPS.

Perhaps the cheerleaders who voted for Messrs Bracken and Maxwell should have looked harder at that GDS strategy, which was heavily criticised at the time by at least four professors:
  • The professors deemed the strategy to be too flimsy, the detail was missing, it was going to be of no practical assistance to a large local authority or, as it turns out, to RPA.
  • The strategy underestimated the complexity of ultra-large-scale government systems, it ignored the relevant academic studies that might have helped GDS to understand the "complex cultural, political and regulatory environment" in which the "technologically diverse, long-lived set of transactional services" of government have to operate.
  • Against that, the appeal to open source, agile, the cloud and SMEs is "over-simplistic", the professors said. "There are risks that rapidly changing [agile] services will deter the takeup of digital services, not encourage it" and "the [Government Digital Strategy] is remarkably (perhaps alarmingly) silent on the issue of how to coordinate SMEs in project delivery":
  • Both of those problems have been experienced by BPS according to Mr Glick. "The iterative development process was also causing problems for farmers. “The system is frequently going down at short notice for upgrades, making it difficult for farmers and agents ...".  One of the suppliers is a "Belfast company that uses offshore developers in Gdansk, Poland", there are "hundreds of IT experts" also involved and more than 100 products that need to be integrated.
  • "We see little discussion of a concrete and practical change management process to support the 'digital by default' strategy", the professors said, back in January 2013. Two years later, that is clearly still a problem.
And what does Mr Glick say?
... the complex guidelines for the new Basic Payment Scheme (BPS) runs to 84 pages ... the situation differed considerably from the days of SPS [the predecessor system].
Clearly we're not dealing with anything "ultra-large-scale" here. That can't be the problem. Nor can the considerably different situation – there are still only 84 pages and the RPA have had at least two years and £154 million to work on BPS with the assistance of GDS's agile expertise.

"Scalability of the system had already been identified as one of the biggest challenges", according to Mr Glick:
GDS chief Mike Bracken acknowledged the complexity involved in a blog post in December 2014. “It’s not just the policy that’s complex. For this exemplar alone, we’re talking about roughly 110,000 farmers and 1,200 land agents,” he wrote.
By no stretch of the imagination is 112,200 111,200 people a large user base, 84 pages do not make for a complex policy – wait till GDS see the 15,000 pages of our tax code – and if the problem of scalability had been identified why did anyone inflict the system on the poor unfortunate users, particularly GDS, who claim to put the user uniquely first?

Mr Bracken is further quoted as saying:
Farmers themselves are a diverse group of people, whose properties can range from a smallholding to an industrial-scale business. The average age of farmers in the UK is also quite high, with many in their 60s and 70s.
Is Mr Glick complaining or being asked to complain that farmers aren't standard enough? And that they're too old, damn them? And too stupid: "There must have been question marks around how less digitally literate farmers would cope"? And even worse, that they live in the countryside?
Broadband access in remote rural areas was another issue, said ... a report in Farmers Guardian.
The system is fine? It's the users who don't work?


GDS will be working very closely with them ...
It's going to be another example of government as a platform ...

No comments:

Post a comment