The (white) elephant in the room

Jan 28, 2009 08:56 · 619 words · 3 minute read government opensource tories

[![Elephant] Politicians and IT usually go together like fish and bicycles.   The story of public-sector IT in the UK is generally one of grandiose over-budget failures at the top end of the scale, and low-level outsourced inadequacies at the other.  The same is also true in the private sector, of course, but the public sector situation gets a higher profile by the fact that it’s “our” money that’s being wasted and it’s more difficult to sweep the disasters under the carpet.

The litany of disasters is a long one - the Child Support Agency’s IT system was “one of the worst public administration scandals in modern times” according to the Public Accounts Committee, while the NHS’s Connecting For Health project is seven times over-budget and more than two years behind schedule.  And that’s before we start trying to keep track of every email sent and webpage visited in the mother of all surveillance databases.

So why do we keep letting this happen?

Earlier this week, the Conservatives launched the results of a study they commissioned last year from Dr Mark Thompson of the Judge Business School in Cambridge.   The headlines were startling - government could save at least £600m a year, according to the report, and it promises an end to “IT white elephants”.

[The text of the report itself doesn’t seem to be available via the Tories’ website, but I’ve got a copy which I can email on if you’re interested.]

There’s also some very interesting detail buried away below the headline figures.   One of the key recommendations is that there should be a cap on project size - no contract should be worth more than £100m.   Personally, that seems a fairly arbitrary figure to me - I’m not convinced that knocking a nought of the end of the contract value will make them any easier to deliver - but it does point to an increasingly widely-held belief that the days of huge projects are over.

Of course, a recession is going to make justifying telephone-number budgets very difficult even if big projects were always successful - but history suggests that when projects fail, big projects fail bigger.   I think the trend over the next few years is going to be towards projects on a much, much smaller scale.  Rather than taking years to spend millions on massive systems implementations that attempt to solve every problem simultaneously, instead organisations are going to try smaller-scale point solutions that are focussed on improving the way people actually work together.

I’m also beginning to think that there’s actually a finite limit to the size of a project, beyond which it’s impossible to achieve the stated benefits.   Some systems are just too big, and too complex, to be controlled by human brains which have been wired up be evolution to make a series of “can I eat it, or do I run away from it?” decisions.  And many situations just don’t lend themselves to a one-size-fits-all solution, particularly if the solution is being defined in isolation from the actual end-users who will be subjected to it - something that seems to sum up the IT programmes in the Health Service.

Of course, changing the culture of Government IT is going to be a huge job given the massive vested interests at play - the huge suppliers that seem to be a key part of the problem are not going to just walk away from their potential revenues.   Bureaucracies move slowly, and government bureaucracy moves slower than most.   But it’s encouraging to see signs that the status quo is being seriously questioned by a potential future ruling party - and hopefully it might influence the current decision makers in the meanwhile.

(cc-licensed photo by huangjiahui)