IDS should learn from Polaris technology

 - it's not quite rocket science

Andreas Whittam-Smith, in the pages of the newspaper he co-founded, has twice in recent months attacked the election-timetable-obsessed Cameron government for failure to plan. In September, the peg on which he hung his argument was HS2; yesterday it was Universal Credit.

There are various ways of planning and controlling large projects. PERT/CPM and Gantt are the most well-known. There is an overview here. The irony is that, as Mr Whittam-Smith points out, the huge US healthcare insurance programme has suffered setbacks, in the nation which established a means of bringing big government projects in on time.

In addition to answering the questions posed by Mr Whittam-Smith, DWP needs seriously to apply a project-planning tool. By "seriously", I mean that a permanent control team headed by a senior executive has to be set up, realistic estimates have to be drawn up, outcomes reviewed regularly and anyone with a stake in the outcome has to be involved. That does not mean that ministers should micro-manage the project; the planning charts would be hierarchical and the higher up the chain of command, the broader and less detailed the charts become. It should still be possible to see in what areas the pinch-points are occurring.

Of course, DWP civil servants may already have done all this but then been hit by a political diktat - the end-date shown by the planning software displeases the minister, who imposes his own go-live date. As a consequence, corners are cut and time-scales artificially shortened - testing is usually the function which suffers. I've seen it happen at close quarters; happily, the first major check caused a pause for thought and reassessment, putting the project back on track to a successful implementation. Sadly, any lessons learned from this experience would have been lost along with experienced staff as government computing was privatised from 1979 onwards.

It is clear that both Iain Duncan Smith and probably the permanent secretary are floundering. It needs someone at ministerial level - and therefore with some clout - with IT management experience to be put in charge for the duration of the Universal Credit implementation. As a Liberal Democrat, I immediately think of Richard Allan, but no doubt there are Conservative Lords or MPs with similar experience. However, there are advantages to appointing a LibDem. There is an outside chance that the Conservatives will not be the largest party in the Commons after the next election. The UC programme is virtually certain to continue into the next administration, and for the sake of continuity a Liberal Democrat would be more acceptable than a Conservative at the helm in 2015. From the Conservative point of view, a LibDem would be a convenient scapegoat if there were a further failure in the UC system.