Government ITprojects need not fail so dramatically
It is time to clarify the debate on government IT "disasters". High-profile projects, such as digitising the passport system, have generally involved a mismatch between IT and human organisations - and the public sector does not have a monopoly on that. What is unique, however, is government's inability to learn from failure.
When central government wants a large-scale, bespoke solution to a problem of national importance - be it air-traffic control or national insurance - it looks to a supplier with a track-record for such projects.
When things go wrong there is a merry-go-round of committees and reports, but this does not prevent the government running back to the same suppliers that failed before.
In the 1990s stinging penalties on civil engineering contractors taught them to deliver on time and to budget with government contracts. The down side is that many walked away from government work when asked to deliver it to private sector standards. We now hear the same whinge from IT suppliers: "there's not enough money in government IT to make it worth our while".
The mass production of stable, high-performance platforms is pushing the suppliers to add value ie make money through services rather than the kit itself.
Government IT solutions using "legacy" communications protocols or lumbering "add-ons"should have rung alarm bells years ago - but the public sector continues to roll out such systems today.
In the end it comes down to managing the relationship with suppliers. The Public Accounts Committee slammed government performance on this but if the current round of boards, enquiries and committees is to bring real improvements it must structure the supplier relationship differently.
It could start by considering these ground-rules: pay for the system only when it works, penalise poor performance, empower the user to resist supplier blackmail - and bring the best of private sector project management into the commissioning organisations.