Original URL: http://www.theregister.co.uk/2006/11/17/nao_celebration/

NAO gets loved up on gov IT

Takes pet projects for walk in the sun

By Mark Ballard

Posted in CIO, 17th November 2006 11:27 GMT

The National Audit Office (NAO) has broken with its tradition of testing the veracity of government activities with the publication of a report about some IT projects the government has managed successfully.

Yes, you heard it right. There are, apparently, successful government IT projects. In its 23 year history the NAO has never set out to find shining examples of government administration to be celebrated, instead of testing and probing to see if there are any holes it should report in keeping with its purpose as independent auditor.

So, the NAO decided that disasters and waste have given government IT such a bad name that it ought to do something to remind us that our civil servants do sometimes get it right and that there is something we can learn from our successes as well as our failures. (HM Public Relations presumably doesn't do enough already in this respect)

Just five months since the NAO was criticised for its gushing report on the alleged commendable achievements of the NHS modernisation project, the National Programme for IT (NPfIT), a report that followed months of hair picking with a truculent Department of Health, an NAO spokesman said this latest ray of sunshine was not designed to do any favours for the government. It just says what a grand job it's been doing.

Those projects stamped with NAO gold stars included the Department for Work and Pensions' £824m automated bank transfer benefits system, the Pension Service £297m pension credit system, and the environment agency' £0.2m fishing rod licences project. [Is it just us, but are these projects rather small?]

It is worth noting that these 24 shining examples (14 if you exclude foreign and private sector projects) were all nominated for medals by government and industry. They are all written up as glowing case studies in an accompanying report and are endowed with an extra sheen from having been written by NAO cynics instead of the usual soul-sold copywriters (we presume).

Just what the government has been doing so well in its good IT projects, said the report Delivering Successful IT-enabled Business Change, can be condensed into three recommendations: get management buy in, co-operate with your supplier, and have a clear idea what you are trying to do.

These recommendations have all been recommended before, time and time again. The world renowned expert on the subject, The Standish Group, which has analysed over 50,000 IT projects since 1994, charts a top 10 reasons why IT projects succeed.**

These appear in the body of the NAO's report, along with some tailored tips for UK government IT bods. First and foremost among them is the insistence that government IT projects will fail while there is not proper oversight to see they are done properly.

This gets to the real reason why government IT projects continue to fail despite the numerous checks and measures put in place over the last five years by the Office of Government Commerce, the Treasury's procurement sheriff, and the unending attempts by private suppliers to show they are doing their bit to make it work.

Without transparency and accountability, says the report, there can be no way of ensuring the checks and measures are followed - and they are not, as this and previous reports have shown.

The reason for such failures in the private sector should be obvious to anyone. Competition forces companies to cut corners, budgets and time, so that many projects fail. In the public sector the same happens because things are run to a political timetable.

So Gateway Reviews - the OGC's panacea for shoddy IT - tend to get shoved in a drawer and the series of reviews tail off before they have done their job. There's so much to do and so little time.

The efforts begun in 2004 by Ian Watmore (now the head of government "transformation") to improve professionalism and street sense among government IT bods will help see that these and earlier recommendations are better followed eventually. But there is something very peculiar about government IT that makes it a little harder to get right than anything ever attempted in the public sector.

That idiosyncrasy may explain why the thing that by Standish's measure is the number one reason for IT project success, is by the NAO's reckoning worth burying in the body of the report: that is, user acceptance. It's basic human psychology that if you try and turn someone's world upside down without their agreement or "buy-in", they'll kick up a stink.

So any private firm worth its salt will include intended users of a computer system in its design from the outset, because it usually involves radical change. Most computer systems, by the extension of their application, have users who exist outside the firm - those people who are the intended recipients of the benefits of computerisation. It's harder to get their involvement in the design and development of an IT system, but it's easy to tell if they like the change: if they don't like it, they can take their business elsewhere.

The trouble with many government systems, however, is that the users don't have much choice in what's foisted on them beyond a vote every four years for a hotch-potch of manifesto commitments that have to be lumped like a bargain bag bought lock stock at a car boot sale just for the sake of that one shiny bauble nestled on the top of all the tat.

A shining exception to this rule is the DWP BACS system, which the NAO said was well liked by benefits claimants, despite strong initial resistance to the idea. The abysmal failure of the Child Support Agency is a prime proof. The Identity Card system will be very interesting for the same reason. It is being foisted on its "users". (Perhaps they'll learn to love the convenience it brings).

This unfortunate reality makes a nonsense of this government's talk of citizens being customers, no matter how much choice it creates by getting the private sector to deliver public services. ®

* £1.5bn of 14bn being spent on UK government IT projects this year

** Standish Group's Top 10 Reasons for Success:

  1. User involvement
  2. Executive management support
  3. Clear business objectives
  4. Optimising scope
  5. Agile process
  6. Project manager expertise
  7. Financial management
  8. Skilled resources
  9. Formal methodology
  10. Standard tools and infrastructure