By Tony Collins
Some people, including those in the know, suspect Universal Credit will be a failed IT-based project, among them Francis Maude. As Cabinet Office minister Maude is ultimately responsible for the Major Projects Authority which has the job, among other things, of averting major project failures.
But Iain Duncan Smith, the DWP secretary of state, has an ace up his sleeve: the initial go-live of Universal Credit is so limited in scope that claims could be managed by hand, at least in part.
The DWP’s FAQs suggest that Universal Credit will handle, in its first phase due to start in October 2013, only new claims - and only those from the unemployed. Under such a light load the system is unlikely to fail, as any particularly complicated claims could managed clerically.
The second phase of Universal Credit, which is due to begin in April 2014, is the important one, in terms of number of claimants. But this phase may be delayed with a general election approaching, according to Government Computing, which quotes the FT.
This is from the DWP’s website:
“Universal Credit will start to take new claims from unemployed people in October 2013.”
“For people in work this process will begin in April 2014. The remainder of current claims will be moved to Universal Credit from 2014, with the process being complete by 2017.”
The projected costs of real-time information, an HMRC project on which the success of Universal Credit depends, have increased by tens of millions from an initial estimate of £108m, according to Ruth Owen, Director General, Personal Tax, HMRC. At least HMRC is being open about RTI – relative to the DWP which continues to deny FOI requests for the risk register or independent assessments of the progress or otherwise of the Universal Credit IT project.
Auditors at the National Audit Office found that the Rural Payment Agency’s Single Payment Scheme for farmers dealt with so few claims that it could have been handled manually for a fraction of the cost of an IT system that went awry. Perhaps Iain Duncan Smith has learnt from that episode.
As Universal Credit phase one will handle only new claims from the unemployed, there may be no need initially for complicated monthly interactions with HMRC’s Real-time information [PAYE] systems.
There may be further restrictions on go-live UC candidates. The DWP may insist that unemployed new claimants are single, childless, between certain ages and not receiving certain benefits or tax credits. They may have to have a valid bank account.
So the numbers of claimants and simplified processing will maximise the chances of a go-live success.
This may explain why the Major Projects Authority has not intervened (yet) to delay the October 2013 go-live date.
It makes sense to minimise complications when going live. But the Passport Agency found that although the go-live of new systems in 1999 went well, extra IT-related security checks slowed down the issuing of passports, such that backlogs built up, people lost their holidays and queues built up at passport offices. It was a project disaster.
The real test of the agile-based Universal Credit project will be when existing benefit claimants move onto the new systems in large numbers. This will not happen before the next general election. The plan is for the roll-out to be completed by the end of 2017.
Meanwhile does Iain Duncan Smith plan to claim a victory for the go-live of Universal Credit when the initial transactions are so simple, and the numbers involved so insignificant, they could be managed clerically if necessary?
As long as Universal Credit does not reduce payments to the genuinely disabled and the most needy, it is generally regarded as a good idea. It should cut fraud and administrative costs.
It’s a pity though that no central department can be open about the progress of its major IT-related projects; and on forcing these progress reports out of dark departmental corners the coalition has made no difference at all.