System Error: fixing the flaws in Government IT

A few months ago, I was interviewed by researchers from the Institute for Government, as part of their work on the report that arrived, embargoed, in my email yesterday.

I’ve now had a chance to read it. I’ve heard a number of opinions on the project over the last few months — I have to say, mostly not very complimentary ones — so I was glad to see what they finally decided.

On the whole, I think it’s quite good, as far as it goes. It’s definitely saying the right kinds of things about the sort of change that’s needed: notably, choosing commodity options when they’re available, and using agile processes to innovate and develop bespoke solutions when they’re necessary. Good stuff. The report includes a now very-well-rehearsed explanation of the problems with government’s traditional approaches which is spot on, and some good recommendations for Government – it will be interesting to see how they respond.

On a minor point (at least, for me) it is very technology focussed and doesn’t much discuss the wider social context that technology needs to serve.

More importantly, I would have liked to see more detail, in general. There’s a lot of discussion of open standards, buying commodity services, and coordinating activity between departments. But it feels a bit like the report skips over the fact that these things are really hard.

Agreeing standards is hard, as is implementing them correctly. Standards for the web have taken >10 years to develop and mature, and in many respects are still not very well embedded: Microsoft have really only just got there with IE9, and that remains to be seen. And this is in an industry where the incentives to make everything work are huge. I’m really not at all sure that the incentives to use open standards for the NHS spine and people’s tax records are even nearly as strong, where suppliers may be reluctant to facilitate the involvement of others.

Commodity services can of course deliver immense value, and the success of Amazon’s Web Services is a testament to their usefulness. But even the private sector is only just getting there. Computing resources are now effectively commoditised, as is — perhaps — the purchase of hardware… and that’s… it. There’s no commodity market for VOIP services that I’m aware of. There’s no commodity market for payroll, workflow management or document storage. For all of these things, we’re still in a world of competing products which are mostly not interoperable. The report actually highlights Microsoft Word as an example of a commodity application, which made me choke on my coffee: it’s quite the opposite. In fact, it’s a dominant product in what is all but a monopoly, and — as anyone who’s tried to use OpenOffice in anger will know — it’s a long way from being interoperable with anything.

Sensibly coordinating activity between departments has been advocated as good for so long that it’s practically axiomatic, and we still aren’t anywhere near it.

Perhaps more importantly still, the report really doesn’t address procurement nearly carefully enough. It does present some interesting suggestions about how to squeeze agile projects though existing OJEU and OGC processes. But I think procurement is at the root of the whole problem, and needs wholesale reform. The main reason that Government’s incumbent suppliers are so dominant is that the barrier to entry for new businesses is immense. And procurement constitutes a very large portion of that problem.

The report even advocates the use of framework agreements as a cost-cutting measure, when — in fact — the use of framework agreements is one of the prime culprits for the lack of any agile suppliers in the market able to compete for the large projects that are currently solely delivered by the Usual Suspects.

In summary – I think this is a good start. The direction of travel is spot on, and I hope it is warmly received and acted upon by government. But it does skim over some terribly difficult problems which we’ll have to confront if we want this to work.