NAO: 1 in 5 of Whitehall's mega projects at risk of failure
£89bn at stake from 39 programmes with 'high or fairly high' chance of tanking
What you need to know about cloud backup
With nearly one in five major government projects in danger of failure, there are still significant gaps in how Whitehall oversees the delivery of such schemes, according to the National Audit Office (NAO).
There are currently 205 major projects being undertaken across government, costing £376bn in total over their lifetimes, the NAO says. Thirty-nine of them, worth an estimated £89bn, have a delivery confidence rating of red or amber/red, which indicates that they have a high or fairly high risk of failure.
The figures were published in an NAO report into major project assurance – procedures designed to assess whether large programmes being undertaken in the public sector, such as complex IT programmes or changes to public services, will be delivered successfully.
The report follows an earlier examination of major project assurance by the auditor, published in 2010, which found a number of failings, including inconsistency in how much attention departments pay to assurance and no systematical way to gather or make use of information to improve project performance.
"The government agreed with our assessment that the central system for assuring major projects was not optimal," the NAO says, and the government has since set up the Major Projects Authority (MPA). Launched last year, the MPA was intended to improve the performance of government's projects through initiatives such as building up project management skills among senior civil servants and escalating issues of concern to ministers.
"Given that the authority has only existed since April 2011, it would be unreasonable to expect the new assurance system to be fully mature yet. However, the steps that have been taken so far have enabled it to progress in the right direction," the report says.
The MPA has made project reviews more exacting, improved the data that is gathered about major projects, and strengthened the link between assurance and project approval, according to the NAO. It has also established the Major Projects Leadership Authority, which will train civil servants how to run major projects and is set to open its doors in the autumn.
"For the system to continue to benefit government into the future, it must be built to last. There must be a chain of close co-operation between the different organisations involved in the system. Some organisations are not yet engaging with the system in a consistent way, and useful information is not being shared and used to best effect," the report says.
The report notes that departments are meeting their assurance requirements with "variable quality and completeness", while the Treasury's engagement has not been as strong as expected, particularly among senior management. It also highlights that there is still no formal system to gather, analyse and share insights on project data, nor any agreement between the Cabinet Office, the Treasury and the departments on how to publish it. The organisations should agree both a format and date for public reporting, according to the NAO.
The spending watchdog also raises questions around skills, noting limitations in the capabilities and number of review staff at the MPA. If staff leave, it says, considerable skills and knowledge would be lost.
It advises the MPA to establish formal ways of planning and carry out assurance work, how knowledge gleaned from such work is passed on, and how it will improve work in the future. A better understanding of whether assurance work has improved the delivery of major projects is also necessary, according to the NAO.
"The ambition of regular transparent reporting, which has not yet been met, is crucial for encouraging co-operation and improving project performance. HM Treasury, departments and the authority must make improvements to maximise the value for money that can be achieved from the assurance system," the report concludes.
Commenting on the findings of the report, Cabinet Office minister Francis Maude said: "The NAO has rightly recognised the huge impact the authority has had in its first year. Unlike in the past, major projects are now tightly managed and the trend of running over time and budget is in reverse.
"The cost, ambition, complexity and risk of major government projects have increased hugely over the past decade – our monitoring system allows us to go out and intervene where projects need support. We publish more data than ever before and will publish the first ever annual report on major projects soon."
This article was originally published at Guardian Government Computing.
Guardian Government Computing is a business division of Guardian Professional, and covers the latest news and analysis of public sector technology. For updates on public sector IT, join the Government Computing Network here.
COMMENTS
I'll be shocked
If only 1 in 5 of these projects fail.
That would be a huge step forward.
If 4 in 5 are at risk of failure but the oversight is so weak that noone has yet realised on 3 of them then that will be more like the norm.
Re: I'll be shocked
Could be that the other four projects have already failed. Semantically, their statement would still be correct and free of lies.
How to do things right - or wrong
Let's assume that you wanted to run a project well.
You'd develop some requirements, and fix them, but with build in contractual means at scheduled intervals to manage upgrades as technology moves on. You'd hire a skilled and experienced project manager, pay them a decent wage, and give them appropriate delegated authority in decision making, hiring/firing, and even pay (to motivate/retain the best personnel). You'd also contract the project manager for the duration of the project, with major and attractive long-term incentives based on the performance of the project. You'd then back off, and let them run the damn thing, only getting involved as a senior manager to help smoothe the worst problems.
Let's assume that you're the government.
Requirements change every few months. Civil servants gain project management qualifications, but with few exceptions, their grade and pay bear scant resemblance to their levels of skill and experience. Delegated authority is non-existent, and while the ability to hire staff is merely lethargic and painful, we are nearly unable to fire. One of our current team on an un-named but many-many-million-pound project is a barely-functioning alcoholic. He has threatened that, if we try to dismiss him, he will bring a discrimination case, which, although he won't win, will cost the department hugely in both time and money. The unions will, of course, back him on this. Senior management only get involved in projects if they think it'll make them look good. Generally, they do this by taking Industry's side in any dispute, because the civil service doesn't give out directorships...
Anonymous coward because, although on any given day I find the governance anywhere between frustrating and woeful, I do actually believe that the job I do is, or at least can be, worthwhile, and I don't want to lose it.

IT infrastructure monitoring strategies
Agentless Backup is Not a Myth
Steps to Take Before Choosing a Business Continuity Partner
Enabling efficient data center monitoring
Requirements Checklist for Choosing a Cloud Backup and Recovery Service Provider