Government Computing Expo 2006: nice conference, shame about the venue

There is something deeply depressing about Earls Court Exhibition Centre. Its ugly façade, appalling acoustics, and general drabness compared to the gleaming conference centres of the US are a sad symbol of the depressing parochialism and fading glory that has characterized Britain's post-imperial decline. Quite frankly, I can think of no better place to host a conference on public sector IT.

Only joking, calm down, I'll take it back (though not the bit about Earls Court, the place really should be bulldozed).

In fact, this year's Government Computing Expo (GCE) provided IT success stories from all corners of the public sector. Whether the projects were e-procurement systems saving millions a year, Birmingham City Council's new joint venture with Capita revolutionising its public service provisioning, or elements of the NHS IT project achieving a sadly underreported 100 percent service availability, the overriding theme was that the public sector is on the up and much of the improvements are being underpinned by IT.

And yet, the overriding perception is that public sector IT projects are as likely to fail as an English World Cup campaign. Richard Granger, head of the government's NHS IT modernisation programme bemoaned this disconnect between perception and reality in his keynote speech. He claimed that just as 80 percent of the public have had good experiences of the NHS but believe they were lucky to have done so, he was finding that people now enjoy good connectivity across the NHS, but regard themselves as lucky that systems are working.

So why have such perceptions developed? Granger lays the blame foursquare on the press, claiming that their desire to report failures instead of success stories creates public pessimism. In ever combative style, he even suggested that much of the negative coverage of the NHS project that followed the National Audit Office's status report was due to the fact that some of the journalists probably had not even read it.

In fairness, he has a point. While the NAO report claims the project continues to face challenges to ensure suppliers meet deadlines and to gain wider support among NHS staff, it also offers a lot of praise for its procurement process and the manner in which many systems are up and running.

However, blaming the press is a dangerous game for anyone in public life, particularly when it cannot be denied that many public sector IT projects have failed at massive cost to the taxpayer.

The fact is, as the keynotes at this year's GCE revealed, there are systemic problems within the public sector that have given rise to both the high number of failing projects and the perception that the government's understanding of IT belongs in the Dark Ages. And until these failings are corrected the public sector is likely to face ongoing criticism of its IT initiatives.

Perhaps the first and easiest step the government should take is to ditch the management-speak. Listening to many of the GCE keynotes it was evident that in attempting to emulate the private sector many civil servants have picked up the worst aspects of commercial life without also taking the focus on metrics, streamlined management structures and accountability that make some private firms so efficient.

At one point a senior civil servant actually discussed the Transformational Government strategy by claiming: "Transformation is a big word. It's an exciting word. It means change. It does not mean doing things the same way." Yeah, thanks for that.

The second step is to change the way the government predicts costs for large-scale IT projects. It seems strange, given how consistently projects go over-schedule and over-budget, that everyone assumes the projects are being run badly but no one ever seems to ask whether the targets are realistic.

Granger hinted at this predicament when asked what his biggest regret was for the NHS project so far. "In an ideal world you wouldn't set a timetable for something like [modernising NHS IT] until you were 20 percent through and had a sense of how long things were taking," he replied. "Though that's not the world I live in."

Of course, determining finishing dates on the fly is not possible when politicians need to have an idea of costs and deadlines before they can approve a project, but there is still no excuse for the inaccuracy of many early estimates.

I agree with Granger when he says the NHS will end up with a state-of-the-art system that will revolutionise health care in this country. But regardless of its benefits the difference between the £12.4bn the National Audit Office claims the programme will now cost and the £5bn originally estimated was always going to dominate the headlines.

Whoever it was in Whitehall who seriously believed a project of this scale could be delivered for £5bn when even straightforward outsourcing deals in the private sector often cost well over £1bn really needs to be locked up. Maybe they could be joined by the civil servant who recently calculated that the Home Office can develop and run its biometric ID card database for £5.8bn.

The reason for these gross inaccuracies are perhaps understandable, when every department knows that if they go with cap in hand to the Chancellor asking for, say, £12bn to revolutionise a public service, they are not going to get it (unless they are the Department of Defense and they plan to use the cash to invade somewhere). So they lowball the numbers, get the go-ahead, get two years into the project, and then say we'll need more cash to finish.

This situation remains unchallenged because the Treasury accepts that keeping the estimates for IT projects low, however spurious the numbers may be, makes it far easier to get ultimately beneficial systems approved by Parliament.

Breaking this cycle of inaccuracy would allow the government to finally enforce the culture of accountability that is desperately needed to restore confidence in public sector IT projects.

Managers in the private sector would think twice about underestimating the cost of an IT project, knowing full well that to fail to deliver puts them in serious risk of spending a lot more time with the family.

But where is the accountability when government IT failures occur? The minister in charge always takes a pounding as they trot out the same tired excuses; the IT providers usually get a lot of flak for being incompetent, money-grabbing racketeers; but the Teflon civil servants simply keep their heads down and move on to the next project.

Without an occasional mea culpa from someone in the public sector when projects go badly wrong, how are we supposed to have any confidence that lessons have been learnt? Granger's speech – in which he dismissed concerns that NHS staff aren’t fully behind the scheme and downplayed the financial problems at some of the projects major suppliers, before moving on to list his many successes - was just as one-eyed as the negative press coverage he accused of lack of balance.

It would be nice to think this culture is changing and that the ID card scheme has been costed properly and heads will roll if in a decade's time the Home Office ends up spending more than the estimated £5.8bn. But I won't be holding my breath.

Of course to make such a fundamental cultural change in the way government IT projects are managed will require extremely strong leadership, lack of which is the final and perhaps defining cause of the government's poor record with IT.

The Transformational Government strategy to use IT to drive public sector efficiency relies on an ability to get countless councils, departments and agencies to co-operate in the way they provide services and share data on citizens. Driving such reforms requires serious political clout, but it is hard to see where it will come from. As result even the examples of successful IT projects highlighted at GCE create the impression of a confused patchwork where some councils or departments are embracing e-government and shared service centres while others continue to regard new technologies with the suspicion usually reserved for Tessa Jowell's financial affairs.

Nominally, Ian Watmore in the prime minister's delivery unit is in charge of this chaos, but while boasting a pretty impressive record thus far he seems to have more irons in the fire than a hyperactive blacksmith. A quick glance at the list of GCE speakers reveals government technology reforms are being managed to a greater or lesser extent by (deep breath now): The Cabinet Office, the Treasury, the Office of Government Commerce, the Department for Communities and Local Government – formerly the Office of the Deputy Prime Minister, IDeA, local government regional Centres of Excellence, Birmingham City Council – just one of 388 English councils, and the NHS.

Sort that little lot out and replace it with a coherent management structure, and maybe, just maybe, the public sector will be able to both guarantee IT projects succeed and ensure it gets the credit it deserves.