Saturday, February 14, 2004
Canada takes some lumps
I arrived late at a meeting the other day and walked in to see that the two people due to speak after me had already done their thing. I said that I felt a bit like a government IT project - late in arrival and even then of questionable benefit. It occurs to me that the word "government" is unncesssary in that sentence, it's just that the PR around government projects is usually more widespread than that around private sector projects. So it was with interest as I was scanning Google's e-government news last night that I saw several reports on some problems in Canada with their "Secure Channel" project. This is a kind of catchall project for Gateway-style authentication, civil service directories, email exchanges with citizens and what not. I've talked to folks from over there a couple of times to compare notes with how we do it in the UK and the approaches are similar in style but with a few major differences - one of which is that the Secure Channel would route transactions to the target department for authentication there, rather than hold it centrally. I'd always been told that this was up and running - so maybe that was a pilot version back then. The articles, like this one, refer to a recent report by Canada's Auditor General (who, by the sounds of it, is a particularly tough, but straight-forward, in-your-face kind of person) saying that: Canada's Government On-Line project is already missing deadlines, costs for completion are unknown and specific department plans are short on details there's no umbrella organization coordinating all efforts, which makes it difficult to manage GOL initiatives as an integrated program One key project, the Secure Channel â€“ which the government considers critical to GOL because it gives citizens a secure, responsive system â€“ is estimated to cost $604 million, yet implementation of some parts of this project are behind schedule and obstacles like long-term financing have emerged Of all the question marks surrounding the government's ambitious proposal, Fraser suggested one of the biggest unknowns is GOL's cost. She said by December 2001, $880 million of new funds was provided for GOL, but since then the government has announced no new direct financing. She predicted that delivering such an array of online services will cost much more, because goverment departments are spending large amounts of money on internal online projects. Moreover, Parliament needs more up-to-date reports on the status and funding of GOL. I have no idea about the detail of all this. One of the things that I do know about audits though is that they are usually very good at identifying the facts, but that the opinions expressed can sometimes be a little off the mark - sometimes nececssarily so because the audit process is a snapshot in time and doesn't cover the whole history of a project, but sometimes because there's a difference in opinion over what the real impact of data is. What's interesting about the points raised here though is that I imagine they're common to any and all government IT projects and probably just as many commercial IT projects. How much is it going to cost? How is it all going to be integrated? Who is joining it up? In an organisation where lines of control are tight to the top, these probably don't exist as issues, but in a distributed and complex organisation - i.e a global corporation or a set of government departments, they're every day problems. Interestingly, one of the harshest criticisms appears to be that the Secure Channel project was due to cost $57 million (Canadian dollars) and is now looking to come to over $600 million. That's quite a change - the missing facts though are, I imagine, has the scope changed, does it include departmental integration costs now? After all, I can't imagine an extra 1/2 billion sneaking up on people without anyone noticing! Canada has, pretty much since the beginning, been one of the leading lights of e-government and I'm hoping that it will stay that way. The ideas that they have fully realised - including consistent UIs for their websites (like this one and this one) - are on many people's agendae, and have resulted in good takeup so far. In the end, every government is struggling with authentication technology, aligning it with existing business processes and creating new, better business processes to support it. I don't think anyone has cracked it yet - and the more that people look at initiatives that might be said to have failed, probably the further we will get from actually cracking it. But likewise, the more people look at technology to be the solution - whether it's smart cards, retinal scans or whatever - the less likely we are to crack it too.
Posted by Alan at Saturday, February 14, 2004