Today’s NAO Report on Universal Credit implementation is one of the hardest hitting critiques in living memory from a usually restrained institution. I would say “I hate to say I told you so”, but I don’t ‘hate to say it” and I did, three years ago. But first the NAO’s verdict:
“The National Audit Office has concluded that the Department for Work and Pensions has not achieved value for money in its early implementation of Universal Credit. …
Today’s report concludes that the Department was overly ambitious in both the timetable and scope of the programme. The Department took risks to try to meet the short timescale and used a new project management approach which it had never before used on a programme of this size and complexity. It was unable to explain how it originally decided on its ambitious plans or evaluated their feasibility.”
Devastating stuff, but not unexpected, as there’s been a steady trickle of stories about UC’s problems.
In a post on my own Whitehall Watch and on the ‘Public Finance’ blogsite, in November 2010, I spelt out why the implementation of Universal Credit was likely to be a disaster. I think the broad thrust of what I said then still holds true today.
I have heard some blame being attached to Iain Duncan Smith – the funniest quip I’ve heard is “what do you expect when you send a Lieutenant to do a Generals job?” (a reference to IDS’s undistinguished military service, which he’s always made a lot of). He has certainly suffered from a large dose of hubris about what it is possible to do and on what timescales.
But, as I suggested back in 2010, the problems are largely systemic not personal – to do with the nature of Whitehall and its ‘policy’ dominant culture. Anyway – read what I said then and make your own minds up:
“This Government could do with a notice hung in every Ministerial office saying “It’s the implementation, stupid.” This is going to especially apply to the welfare reforms unveiled by Ian Duncan Smith today [25 Nov 2010]……
The bit of the [Welfare Reform] White Paper  that ought to send a chill down the spines of all involved in implementing Universal Benefit is as follows:
“The Department for Work and Pensions will be responsible for the delivery of Universal Credit and will make extensive use of online technology to allow people to better manage their claim and understand the benefits of entering paid work. We expect to start taking claims for Universal Credit from October 2013.”
Oh dear. Thirty odd benefits consolidated into a single system, with all the IT requirements, retraining, process design and a host of other implementation issues are going to be solved in just under three years. Really? Are they serious? Fortunately, Whitehall has such a brilliant track-record of implementing big IT projects….
Moreover DWP has very little experience of providing this stuff on-line. At present the best systems they have are on-line form filling, where the completed forms are then printed off and someone deals with them as good old paper files. They are going to jump from where they are now to integrating 30 odd systems into a web-based system in just three years? Seriously?
And of course implementation will require co-operation across several government departments, including DWP, Treasury, Communities and Local Government, etc – and again, we all know how good Whitehall is at joined-up government…
Here’s just one example from the White Paper:
“Recipients who have earnings from employment will have those earnings automatically taken into account. We intend to use HM Revenue & Customs proposed real-time information system to identify earnings and to calculate the net Universal Credit payment due by applying the appropriate taper to the gross payment. This means that those recipients who receive earnings through Pay As You Earn will not need to inform us for payment purposes if the amount of their earnings change.”
Good luck with that.
Just think about the implementation muddles the Government has managed to generate around the relatively simple proposed change to Child Benefit and multiply that by thirty and you get some idea of the train wreck on the horizon. Just read Chapter 4 of the White Paper, and weep at the naïve optimism.
And of course, all of this requires primary legislation before they can even start to specify and sign-off development plans for the systems and technology needed to make this all work.
Please don’t misunderstand, I am broadly sympathetic to the idea of simplifying benefits. I agree with the aim of people being better off in work than on benefits. I agree there are some people abusing the current systems and this should be stopped as much as possible.
But the potential for disaster and what Ian Duncan Smith charmingly calls “perverse disincentives” and other disasters in this rushed reform are legion. I hope I’m wrong, but that really would be the triumph of hope over experience.”
11 thoughts on “‘Universal Credit’ Crunch Arrives (as I predicted it would)”
“Fortunately, Whitehall has such a brilliant track-record of implementing big IT projects”
I always find it a massive warning sign when totally changing business processes is described as in IT project. It’s a business reorganisation facilitated by IT. An IT project is something like “upgrade network”, “upgrade email”, etc. I doesn’t matter if it’s government or corporate, any time a business change is called an IT project it’ll struggle, because it needs to be a business project first.
Completely agree : this is huge process re-engineering project as well as IT, which is why it would probably have been better to do it in stages.
I am reminded of the disaster that was and is the Child Support Agency.
Disasters also predicted and presumably continuing to this day.
It is as if the Government Social Security system has no collective memory of past failures that it might learn from rather than repeat similar types of mistakes in future>
Yet still they forge ahead with the Probation outsourcing to a ridiculous timetable with a scheme that will destroy much of what exists and leave no fall back situation after it fails.
This was indeed predictable and of course DWP civil servants would have known that at the outset. But what would have been the fate of any civil servant who told IDS in 2010 that his grand plan was not achievable? DWP is actually relatively “can do” for a government department and given a reasonable and achievable task it will generally get it done – merger of DSS and ES, roll out of the new-style JCP offices, (too) many examples of introducing new benefits. If anything the department is culturally inclined to be over-optimistic and disclined to say no to ministers.
Agreed Susan. And indeed we (Carole Talbot and I) have used JCP as an example of a successful merger in a recent book chapter we did – see Chapter 6 The Structure Solution? Public sector mergers in the UK in http://www.amazon.co.uk/Crossing-Boundaries-Public-Management-Policy/dp/0415678242/ref=sr_1_1?ie=UTF8&qid=1378457671&sr=8-1&keywords=janine+o%27flynn
You point out: “ as I suggested back in 2010, the problems are largely systemic.” Yet such systemic nature of complex systems is the very essence as to where the problems reside; yet it evidently does not appear to be understood. The NAO points out its role is to “draw out what good looks like;” yet in reality it has no idea what good should look like when the system aspired to is complex and hence subject to systemic sensitivities. The one organisation, common factor, that has been involved in all these failed initiatives has been the NAO – perhaps it is time the NAO was audited for its own skills and competences as to what good looks like in complex systems?
Dear Chris, whilst I sympathise with some of your points, I’m afraid you completely misunderstand the role of the NAO. To say it is a “common factor” in all these failed projects is a bit like saying the Police are a “common factor” in all burglaries. Yes, perhaps the NAO could do more to intervene earlier when things are obviously going wrong, but its powers and resources are limited and it can’t second guess every management and policy decision in Government.
As for your point about complex systems – the whole point for me is that the Government should have tried to uncomplicate things as much as possible rather than going for what was rolling up several different projects into one: merging different benefits, including passing the necessary legislation; putting the whole application and calculation process on-line; integrating it with a (at the time non-existent) real-time income tax system; etc. These could easily have been broken down into phased projects over a longer time period with much lower risks attached. Trying to do them all in 3 years was always bonkers.
Nice return of service; unfortunately the current role of the NAO is too well understood; it draws out what it says “looks good” in terms that it understand represents the validation of a verifiable set of criteria – yet the criteria, as evidence identifies, has not been policy tested for “viability.” The 3 Vs rather than 3Es! We end up trying to “build the wrong system right” because that’s the way the public sector captures requirements and audit compounds this with what might be described as “errors of omission.”
The previous outgoing Chair of the PAC identified complexity at the top of his ten most significant hurdles to policy delivery; but this creates a common paradox as complexity is the easiest way society solves problems – it makes systems more complex; and we know that complexity is necessary to deliver the richness in our lives (just as the ACPO identifies complexity with policing yet has difficulty identifying what this means in terms of policy and strategy).
To emphasise this difficulty consider the following observations (by a very respected source) regarding a Gov. Depts. requirements analysis:
On the topic of suitability for complicated vs. complex systems (i.e. those that are linear and easily modelled vs. those that are non-linear & counter intuitive) the key point here is whether the non-linearity is a property of the solution domain or the problem domain If they are only a property of the solution domain then the higher level requirements do not need to be aware of them, in this case the identification of the active elements/functions in the system comes from the analysis of the problem and any modelling that is done to support that analysis. If, however, they are a property of the problem domain, then this raises the issues to a higher level and I am tempted to ask whether any such system (requirements capture) really has such a purpose.
Notice that the problem domain has to also appear in the description of the solution domain!
The result is that when events don’t deliver yet another so called wicked problem is identified and no accountability is exercised; yet wicked problems have become an excuse for systems aspired to in a postmodern (everything is relative) period being delivered as if we were still in a comforting modernist time of reductionism and linear practice.
Could it be that today’s PASC Truth to Power: how Civil Service reform can succeed is more than over due and that the role of audit has to retuned to the vagaries, the reality, of our complex society?