Tuesday, June 05, 2007

The Ghost Cabinet on IT Procurement

I'm far from sold on this piece from the Ghost Cabinet 'Technology Ghost'. A few particularly problematic paragraphs stand out:

"No-one is looking at the use of technology across the whole of government."

Utterly untrue. The National Audit Office should do some more work on examining the government's record in delivering major Information Technology programmes but there are definitely organisations looking at the government strategy for IT in the aggregate. There is the Office of Government Commerce and the Cabinet Office's Delivery and Transformation Group (formerly the e-Government Unit) both of whom are responsible, within government, for government IT implementation. Then, in Parliament, there's the Public Accounts Committee which launched a report just yesterday on the lessons from successful projects for government IT procurement. Finally, from the private sector, the consultancy Kable is focussed on public sector IT expenditure.

"This is a historical mess. At some point in the past, each government department went and did their own thing, spending loads of money implementing essentially the same solution multiple times, and now we’re looking at spending loads more either doing it yet again, or joining them all up. No-one was empowered to put their hand up and say, ‘Hang on, does anyone think that a costly, massive database of everyone’s names, addresses and other information might be useful across government?’ And if they did, no-one was made to listen."

This paragraph is the biggest flaw in the piece. The whole problem with the NHS National Programme for IT (NPfIT) is that it attempted to impose a single solution across too large an organisation. Different sections have different legacy systems all of which need to be integrated into the new project or replaced and the data moved. Everyone has different objectives which need to be accounted for and which cause massive creep in the project's scope. Pretty soon you get the over £10 billion (that's a LOT of money) overspend that has marked the NPfIT. Doing this over all of government would take a big risk of facing the problems that have plagued the NPfIT again but on a much larger scale.

A rational, problem-solving, mind will always find centralisation tempting. It is similar to the logic that a major supplier like the NHS can be more efficient than an insurance based system as it can buy in bulk and spread fixed costs across a larger health service. The problem is that the centralisation creates complexities, variety and scale too large to be managed. The impossibility of managing the lumbering behemoths that centralisation has created leads to the inefficiency that characterises today's government departments.

It seems pretty clear that, instead of the centralisation that the 'ghost' is after government IT projects need to be more decentralised, more often It is far easier to manage and control small projects that the stakeholders can be properly kept in touch with.


Devil's Kitchen said...

Actually, by far the best thing to do would to have people in charge of IT who actually know something about IT; even a low-level programmer who keeps up to date would help.

For the main problem with the English NHS spine (as opposed to the Scottish NHS one, which cost practically nothing and works) is that they chose entirely the wrong solution.


Ghost Cabinet. said...

I've encouraged the technology ghost to pop by. I believe that he thinks that you have the wrong end of the stick, but thanks for promoting debate.