Once-upon-a-time Universal Credit (UC), the UK government’s plan for a comprehensive rebuild of the country’s social welfare payments system, was going to be an IT showcase.
It would use the increased power of networks to update people’s benefit payments in “real time” – aiding labour market flexibility as work patterns changed as well as being designed to take advantage of increased computing power and capacity to ensure that the benefit system was operated as an integrated whole and not a disjointed patchwork.
More than even all this, it would be built using “agile” methods to ensure that it was on time and to budget and that the client – namely the UK taxpayer – got the system wanted as opposed to the system that could be bolted together most simply by the software writers.
None of this is now going to happen.
And to the government’s shame they have even debased their own, quite successful, project management system by pretending that UC is something other than a catastrophic failure.
Today, under the cover of a media focused on reporting a major set of domestic election results, the government has slipped out the annual report of the “major projects authority” (MPA) – the body it established to take a grip on the biggest pieces of government procurement.
The MPA has undoubtedly been a big success – as its report shows it has made a positive intervention in a number of projects and forced ministers to take firm action on weaker projects – even to the extent of scrapping them. But on UC the clarity of that process has been sacrificed to political considerations.
For UC has plainly totally failed but rather than admit it, the MPA has created a whole new assessment category – “reset” – and decided that UC is now a new project to be assessed from scratch.
What we should be getting instead is an honest appraisal of why the project failed and what lessons should be learned – especially in development of major software projects. But that would involve admitting that the government has consistently misled people about the nature of the development process, its timetable and its effectiveness. And that is just not going to be allowed to happen.
cartesian product ... stuff about computing mostly 