Universal Credit Cock-ups: the Computer Industry Speaks.

“Basic flaw rests in the idea that we can “personalise” benefits for millions of people. There are just too many moving parts; and in a system with millions of iterations, anything that can go wrong will go wrong. “

Following the advice of many people on this Blog, I have been looking at what the chaps and chapettes in the Computer industry have to say about the piss-poor Universal Credit scheme.

It’s with little doubt that the Macedonian News Factory, the ‘alterative facts’ of the DWP will say otherwise but these are some of their views;

Department for Work and Pensions to spend £90m enhancing heavily criticised Universal Credit IT system – much of which will be thrown away

The Department for Work and Pensions (DWP) is to spend a further £90m on enhancing the heavily criticised IT system developed for Universal Credit – much of which will eventually be thrown away.

But the new “twin-track” approach to the project – which involves developing a new digital system largely in-house with little or no contribution from the existing IT contractors – is budgeted to cut the overall IT costs by £135m.

The figures are revealed in the latest draft Universal Credit business case, excerpts of which have been seen by Computer Weekly.

The document shows that the total IT investment cost for the government’s flagship welfare reform programme is expected to be £535m – down from the £670m forecast in a previous business case in late 2012.

The largest element of that cost is £397m for the existing system– the previous total of £670m was entirely attributed to this element of the project – with £307m of that figure already spent.

“Costs reflect actuals of £307m and future expenditure of £90m to cover the stabilisation of the existing live service in Pathfinder, development of the service to accept claims from couples in Spring 2014, families in later 2014 and tax credit claimants in Jan [stet] 2015,” said the draft business case document.

The department admitted in December last year that it was writing off £40.1m of the IT work already completed on Universal Credit, with a further £91m to be written off over a five-year period instead of 15 years as previously planned. This means that by the time Universal Credit is fully rolled out in 2017/2018, at least £131.1m of the planned £397m IT spend on the current system will have been thrown away.

Most of the overall decrease in budget appears to come from the change in approach that sees a new “end-state” digital solution being developed in parallel with enhancements to the existing system, which is currently being used to support the Pathfinder trials for Universal Credit.

As revealed by Computer Weekly earlier this week, the four big IT suppliers working on the existing system – HP, Accenture, IBM and BT – will have “significantly less” involvement in the digital system, with in-house DWP resources preferred.

Building the “end-state solution” will cost £106m based on the latest assessments, according to the business case document. That figure includes external IT costs of £69m and the in-house “Design and Build” team at £37m.

Computer Weekly. 8th of January.

In the same Computer Weekly Paul Spicker comments,

It is easy to blame the IT when things go wrong, but when people are asked to do impossible things, it is not surprising if they do not deliver

The government’s Universal Credit scheme has been bedevilled by a series of disasters. The problems are nowhere more visible than in the Department for Work and Pensions’ failure to set up the IT.

After three and a half years, the systems the Department for Work and Pensions’ (DWP) needs to make the benefit work do not yet exist. Much of what has been produced has been written off as useless; pilots have been repeatedly delayed; the implementation depends heavily on temporary stop-gaps and manual processing; the continuity of management has been blighted by illness and staff leaving.

It is easy to blame the IT when things go wrong, but when people are asked to do impossible things, it should not be surprising if they do not deliver.

Universal Credit (UC) is supposed to combine a range of existing benefits into one, simpler system. Those benefits include:

• Jobseekers Allowance, for the unemployed;
• Employment and Support Allowance (ESA), for those too ill to work;
• Housing Benefit, for tenants on low incomes;
• Child Tax Credit, for families in and out of work;
• Working Tax Credit, for people on low wages; and
• Income Support, for others (such as lone parents and people with disabilities) on very low incomes.

Benefits are complicated for good reasons. They are trying to do lots of different things, in lots of different circumstances. UC won’t make the system markedly less complex. All the component elements will still be there, fortified by a clutch of extra rules.

The big idea was that UC would deliver “personalised” benefits in “real time”. There were obviously difficult IT issues that needed to be resolved – such as matching up records from the DWP, local authorities and HM Revenue & Customs for people in different jobs – and they haven’t been fixed. However, the problems are more fundamental.

Any computer system could only ever hope to go as fast as the information that goes into it. People whose lives are constantly changing often can’t give clear answers to straight questions. Many current claimants are working in “flexible”, insecure employment, but people who are working casually often don’t know if they’re really employed or if they’re going to be paid.

The Institute of Chartered Accountants has warned that many of the self-employed people they work with find it hard to give a sensible account of their income once a year, let alone once a month. People whose relationships are breaking up, or being newly formed, don’t know what their status is. A DWP minister has tried to explain how complicated UC becomes when two divorced people join their families together – the claim has to be dealt with manually.

A recent survey of people with disabilities – all eligible for benefits – found that 62% of respondents did not think of themselves as disabled. Of the rest, 11% said they were disabled “sometimes”.

The problems of benefits have been extensively researched.

People don’t understand the rules. They don’t know when they’re entitled to, or when they stop being entitled. Many people don’t claim benefits they are supposed to receive. There are particular difficulties associated with means-testing – setting up rules that can be enforced, trying to balance equity and incentives and all while trying to keep tabs on fluctuating incomes and constantly changing circumstances.

The most complicated benefits, such as Pension Credit and ESA, are dogged by “fraud and error”, a tally mainly made up of mistakes by claimants and staff.

Past governments have presided over a long series of administrative fiascos. There was Unified Housing Benefit, which relied on a complex calculation and an arcane set of rules. There was the Child Support Agency, which asked for the current income, liabilities and household circumstances of two families. There was the Tax Credit scheme, condemned by the Ombudsman as being unsuited to the needs of people on low incomes.

UC incorporates just about every design feature that hasn’t worked in the course of the last 30 years. And it adds others, such as:

• “Digital by default”, which can’t work because of basic security requirements – it has already been given up as hopeless:
• Applying conditions to couples jointly, which means that declarations by one person are never enough;
• Setting impossible conditions for job search, which can’t be complied with or policed effectively; and imposing sanctions that cut people off benefits for arbitrary fixed periods.

The most basic flaw rests in the idea that we can “personalise” benefits for millions of people. There are just too many moving parts; and in a system with millions of iterations, anything that can go wrong will go wrong. 

To be able to manage, benefit claimants need secure, predictable incomes. A benefit system needs to be simpler, less individual, slower to change and less intrusive. This one is collapsing under the weight of its overweening.

Bear this in mind!