The development of Universal Credit has been dogged by missteps. So can it still deliver a new system of “dynamic benefits”? Colin Marrs asks the experts
Iain Duncan Smith’s zeal for welfare reform was sparked after a deeply affecting visit to Glasgow’s Easterhouse estate in 2002. Two years later, following his short stint as Conservative Party leader, he founded the Centre for Social Justice, a think tank focused on tackling poverty. Duncan Smith became convinced that the rate of benefits withdrawal provided a massive disincentive for the unemployed to return to work.
In a 2009 report, the CSJ outlined a new system of “dynamic benefits” designed to address the problem. Although the policy was omitted from the 2010 Conservative manifesto, a hint of what was to come appeared in the coalition agreement. “We will investigate how to simplify the benefit system in order to improve incentives to work,” the blueprint said.
According to reports, Duncan Smith was only willing to accept the role of work and pensions secretary in May 2010 on condition he be allowed to implement his benefits reform concept. He got his wish, despite scepticism from an austerity-focused Treasury machine nervous about the initial set-up costs. Following this May’s general election victory, he was reappointed, giving him a chance to prove critics of the programme wrong, and to overcome its significant teething problems.
PCS Universal Credit staff start two-day strike
DWP failed to consider welfare reform risks, says spending watchdog
Daniel Thornton: If the government wants to get big projects right, it needs to become agile by default
Spending review: An opportunity to redesign services for the better?
In November 2010, the Department for Work and Pensions published its proposals for the new benefit, based on the dynamic benefits concept, but now dubbed Universal Credit. It would replace six means-tested benefits and tax credits – Jobseeker’s Allowance, Housing Benefit, Working Tax Credit, Child Tax Credit, Employment and Support Allowance and Income Support. Chancellor George Osborne set aside £2bn to implement the policy, affecting an estimated 8 million households, by 2015.
The concept was given a warm welcome by poverty charities and private sector organisations alike. Even the TUC, which raised a number of worries about implementation, stated “the government’s ambition to ensure that all benefit claimants who move into jobs are better off financially is laudable”. And, crucially, civil servants in the DWP were also enthusiastic, according to Chris Goulden, head of the poverty team at the Joseph Rowntree Foundation. “There were many in DWP who wanted to take it on board because they were aware of the problems with the existing system,” he says.
It is no exaggeration, however, to say that the initial handling of the programme by officials was disastrous. The problems were laid bare in an excoriating 2013 report by the National Audit Office: NAO head Amyas Morse said the programme’s ambitious timescale “was not matched by an appropriate management approach” and that it suffered from “weak management, ineffective control and poor governance”.
The watchdog said the department had ignored warnings that it needed a detailed plan for how Universal Credit was supposed to work. “By mid-2012, this meant that the department could not agree what security it needed to protect claimant transactions and was unclear about how Universal Credit would integrate with other programmes,” it said. These concerns led to the Cabinet Office rejecting the department’s proposed IT hardware and networks solution as representing poor value for the proposed £55m outlay.
Other criticisms by the NAO identified a clumsy implementation of “agile” project management techniques, which had never before been used on a project of Universal Credit’s scale, a “fortress mentality” and “good news” reporting culture within the programme, along with poor control over supplier spending and ineffective departmental oversight. “The lack of a detailed plan or management information meant that the department has never been able to measure its progress effectively against what it is trying to achieve,” it said.
Perhaps worst of all, the DWP was failing to learn from its mistakes. The NAO said that recommendations from the department’s own internal assurance reviews were not being implemented. “Although the nature and emphasis of its recommendations changed over time, the key areas of concern raised by the Major Projects Authority in February 2013 had appeared in previous reports,” it said.
The concerns raised by the MPA were so serious that the project had been “reset” earlier in the year. The team in charge of the process was charged with developing a proper plan for the programme, implementing the delayed piloting of the project and searching for a new senior responsible owner of the project.
Following the reset, DWP adopted a “twin-track” approach to delivering Universal Credit. This meant continuing to roll out the ICT developed by external suppliers, dubbed the “live service”, while simultaneously delivering a new “digital” service built in-house, aimed at improving functionality, enabling all six benefits to be incorporated. In 2013, the Public Accounts Committee found that the DWP expects to reuse just £34m of technology of the £344m it spent on the live service when the digital service is operational.
Such figures were jumped on by critics, but Neil Couling, the current senior responsible owner on the project, tells Civil Service World that the approach was the right one. “If you spool the world back to 2010, we [in government] didn’t even have things like smartphones and iPads then, so it was just a very bad time, I think, to start developing an IT project,” he says.
Jos Creese, president of BCS, the Chartered Institute for IT, is sceptical about the explanation, blaming the difficulties partly on “questionable methods for development”. He says there are huge challenges inherent in implementing any IT project as ambitious as Universal Credit. “It is easy to simply say ‘keep it simple and start small’, but it is very often just not possible to develop a basic level of national foundation without a broad system base that adequately reflects many areas, geographies and interests,” he says. “That inevitably creates a level of complexity and risk that is rarely paralleled in the private sector.”
In future, according to Creese, government departments may have to completely rethink their approach to large-scale ICT projects. “It is my contention that the best national systems of the future are going to come from a simple high-level national policy, framework and infrastructure, from which local public services such as health authorities and local councils can be franchised to implement, following the blueprint, but informed by local geography, democracy, existing infrastructure, technology and priorities,” he says.
Nonetheless, Couling says the twin-track has allowed some lessons to be learnt as simpler new cases move onto the live service. “As an SRO I’m very pleased to have twin track because, effectively, I’ve got two horses in the race. Everything so far is suggesting that the digital service will be ready and capable of scaling and delivering on time, but if it isn’t then I’ve got the twin track of the live service to flip to if I need to, and I’m deliberately holding off decisions on that until I’ve got a full level of confidence around the digital solution,” he says.
And although the administrative costs under twin track are £244m higher than scrapping the live service and waiting for the digital solution to be developed, “the live service pays for itself because it allows me to bring forward about £240m of economic benefits a couple of years earlier [than] rolling out the digital service,” Couling says. “Which is why the Treasury signed it off. They would never have signed it off if it wasn’t a good economic case.”
The reset also resulted in a range of governance reforms to the programme. “We’ve broadened the programme board, so it’s now got senior representatives from the Cabinet Office, the Treasury, HMRC and local government and it’s chaired by Sir Robert Walmsley, a former civil servant who is now independent of government,” says Couling. “That has really strengthened the scrutiny of what’s going on, so now I take my plans to them and they give them a good going over, and there’s really good challenge in the system now.”
The changes appear to be bearing some fruit, if the department’s figures are to be believed. In February, it published a study of the first 6,000 Universal Credit cases showing that, compared to Jobseeker’s Allowance (JSA), claimants are more active in seeking work, finding it more quickly, staying in work longer and earning more. Over a four-month period, claimants were 13% more likely to have been in work than on JSA. Couling says: “I think that gets a bit lost in the whys and wherefores about systems, which are important, but it’s the outcome that really matters.”
The roll-out of the live system is currently on track to meet its (admittedly twice revised) timetable, Couling says. At the time of our interview, 314 of 714 job centres in the UK are now operating it for simple new claims, with the rest of the country due to come on board by March next year. Couling says that the timetable “allows me to test and learn going forward, so instead of assuming we’ve got all the answers here in Whitehall, we’re actually working with people on the ground to drive out answers to learn.”
The Treasury signed off a revised business plan for Universal Credit in September, delaying full delivery of the digital service – when legacy benefit claimants and those getting tax credits are added to the system – for two years until 2019. The new digital system is currently being trialled in south London, and will be integrated with existing systems over the next year. Couling says the process is “going well”. “Our claimants like it, our staff like it, and we’re learning how to put changes in quickly.” But, says Finch, the success or failure of the entire project rests on getting this part of the programme right.
In June, the Major Projects Authority’s annual report gave the Universal Credit project an amber/red rating. Such ratings are given, according to a Cabinet Office spokeswoman “because successful delivery of the project is in doubt, with major risks or issues apparent in a number of key areas”. However, documents released alongside the report listed no specific areas for concern, saying that “delivery remains on track against plans announced in September 2014”.
Neither the DWP nor Cabinet Office are willing to reveal the major risks or issues identified by the MPA in its assessment of Universal Credit. However, a spokesman for DWP denies any contradiction between the traffic light rating and the positive progress outlined. “The amber/red rating reflects the size of the project,” he says. “All the projects the MPA looks at are large but Universal Credit is probably one of the larger reforms in Whitehall. It also has to do with the timescale of the projects – those that have amber or amber/green ratings tend to be slightly smaller or closer to delivery.”
David Finch, senior economic analyst at think tank the Resolution Foundation, says the hurdles remaining to final delivery should not be underestimated. He lists a number of in-built complications which are likely to cause problems, such as the requirement that the rent element of Universal Credit is paid directly to claimants, and imposing monthly reporting requirements for the self-employed. “There are some unnecessary complications that have been decided on as policy despite experts – not vested interests – warning they were not a good idea,” says Finch.
Bill Irvine, benefits consultant and former local government adviser to the now defunct DWP housing benefit standing committee, agrees, pointing to the belated and “grudging” introduction of a pilot with one housing association allowing some payments to go directly to the landlord, along with some small improvements to the complaints process. “There were a lot of experts they could have consulted. If they had done that they would have resolved a lot of problems years ago,” he says. “But even though the DWP has no prior experience in administering housing benefit, it chose to ignore them.”
The JRF’s Goulden raises a bigger question about the development of the Universal Credit in isolation from other departments. He says: “There is a danger that the policy will make it easier to take work but the work is of low quality. There needs to be some thinking within Whitehall about how Universal Credit links to industrial policy and other labour market policies so the result does not just become people moving off benefits to be stuck in low wage jobs.”
Despite such ongoing challenges, Couling says that civil servants from other countries are keen to learn from the UK’s experience. “Yes, it’s tricky. Yes, it’s a big programme,” he admits. “But that’s why I’ve got 200 people working on the digital solution and I’ve got another 300 working out how to then implement that across the country. There isn’t anybody’s experience to draw on – that’s why we’re taking it carefully, so we can deliver in a steady and secure way.”