Dangerous Enthusiasms: E-government, Computer Failure and Information Systems Development
Robin Gauld and Shaun Goldfinch
Otago University Press, $39.95,
Anyone who follows IT news knows that many large-scale projects end up as catastrophic failures. Even when the projects do get finished, their backers often find the technology is late, over budget, or doesn’t deliver on its promises. Despite this, New Zealand businesses and government departments still look to massive IT projects to solve their problems. Dangerous Enthusiasms lays out in detail why that’s usually a bad idea, and anyone thinking of embarking on a large-scale IT project would do well to read it first.
For decades, governments and businesses have thrown money at these projects, and managers still rush in doe-eyed and hypnotised by the potential of the promised technology. The authors estimate that trillions of dollars have been poured into public sector ICT initiatives, with little real benefit. In some cases the new systems were worse than the old. This money would have been better spent, they say, on more police, or hospital equipment, or saving the environment. The same mistakes keep getting made, but public agencies don’t seem to be learning from them. The latest mania is for “e-government”, which will overhaul government processes along technological lines, making them more efficient and more democratic – if it succeeds. Just don’t hold your breath, because as the authors warn, “large projects almost always fail”.
Dangerous Enthusiams looks at three huge projects in New Zealand over the past decade. The Police’s INCIS project and Health Waikato’s aborted installation of the SMS software package were utter disasters by anyone’s criteria, and the other, Land Information’s Landonline, blew out its budget and deadlines to the point where it could only be considered a partial success at best. Robin Gauld and Shaun Goldfinch examine these three projects in detail, using news reports, government accounts and first-hand interviews. Even if you followed their progress in the media, you’re sure to learn something new.
INCIS is the most notorious of the three. It was an extremely ambitious project to link thousands of police stations with each other and a central crime database. Its backers promised it would cut costs, reduce paperwork and solve crimes. But the system never materialised, leaving the government more than $100 million out of pocket when it was finally abandoned in 1999.
So why do these massive projects often fall apart? It’s a complex issue, but the authors, both academics at the University of Otago, point the finger at several key factors. Some are obvious, like long timescales that make the original technology all but obsolete. Others are well-known but continually missed, like staff’s resistance to change. After all, there’s no point in spending millions of dollars on new software if no one actually wants to use it. But organisations keep pushing ahead with these projects without asking staff what they want, or whether they even want it.
Goldfinch and Gauld also point out other, less obvious causes. The dangerous enthusiasms of the title are behind a lot of failures, and this “pathological” eagerness comes from technology vendors and managers alike. The vendors often oversell their goods, which the authors call “Lomanism”, often building penalties for delays into their tenders. Managers themselves have fanciful, even ludicrous, ideas about what technology can do. The authors also show how optimistic the reports coming from the front-line were, even when the project was completely off the rails. Problems were hushed up, and delays glossed over with claims it would be “right on the night”.
When things do go wrong, the thousand clauses in the signed contracts usually don’t protect the government from getting burnt. Human frailties also foul things up. Managers often don’t understand the underlying technology and won’t admit it, while in-house IT staff think technology can solve what are really problems within the business. Most consultants and contractors will tell you that’s in the past, and that new management or programming techniques can take much of the risk out of projects nowadays. But the authors say if that’s true (which they seem to doubt), no one’s found them yet.
The current solution is to break large projects into small parts, but Gauld and Goldfinch doubt this makes things much better. In the case of Landonline, PricewaterhouseCoopers won the tender to design the project and the subsequent tender to build it. The company then claimed its own design couldn’t be built. The authors conclude, rather depressingly, that partial successes like Landonline might be as good as large IT projects can get. The complexity of the projects makes them incredibly difficult, if not outright impossible, to control. And there’s still no reliable way to estimate how much a large project will cost, so large projects will continue to blow out their budgets in the near future at least.
But it’s not all bad news. The New Zealand public sector is doing no worse than the private sector or other governments. Gauld and Goldfinch aren’t Luddites either; they just argue for a strong dose of realism, even pessimism, before an organisation decides to go ahead.
So how does a project succeed? First of all, by being truly necessary. The police did fine using the Wanganui Computer (that INCIS was supposed to replace) for a few more years. There was no crimewave. If a technology project isn’t really necessary, there’s probably a better way to solve the organisation’s problems. A good project has modest aims too, with a healthy scepticism for the enthusiasm of consultants and contractors, and from within the organisation. And it uses existing technology, not bleeding edge stuff.
Despite the daunting title the blurb on the back says the book is written “for the general audience”, but any book that uses words like “disaggregated” has a tough time convincing me of that. The tone is formal, and the negatives often double. Still, it’s not a difficult read. The authors don’t expect you to wade through details about the underlying technology. Anyway, the technology is often irrelevant: projects usually fail because of people, they argue. Bad planning, unrealistic expectations, personality conflicts – these are the constants of failed IT projects. Dangerous Enthusiasms sounds a note of warning that anyone involved in large projects will recognise. Let’s just hope the Government is listening.
Reuben Schwarz is a technology writer with The Dominion Post, and this year published his first children’s book, The Whizbanger that Emmental Built.