Pockmarked by political flak and lumbering behind schedule, the New Zealand Police force’s embryonic new computer system is certainly copping a fair share of controversy.
There may be more to come. The rationale for INCIS (integrated crime information system), presently costed at about $200 million over eight years, is that the gains will far outweigh the costs. And there is a curious arrangement between the government and the police regarding how the benefits of the system--assuming there are benefits, something that many are now questioning--will be distributed.
Those putative benefits (administrative efficiencies, cost savings and more time on the beat for officers) from the project will be split between the government and the police, with the police taking the lions’ share. Deputy police commissioner Barry Matthews says the tentative division is for the government to take 30% of the benefits and the police to take the remaining 70%.
“We argued we should keep all the benefits,” says Matthews. “We eventually came to an arrangement where the police would keep the majority of the benefits--albeit to be negotiated further.” The 70/30 split is to be re-negotiated shortly, as a more detailed analysis is completed of what the benefits might be.
But hang on. Aren’t the police part of the government?
Well, yes and no. More so than other departments of state, the police operate at considerable arm’s length from the rest of government. Primarily this is for what boils down to constitutional reasons--so police minsters can’t set police officers on to people they don’t like. A consequence of this is that the police minister’s role is effectively reduced to an advocate at the Cabinet table for more money--a role some have played with more enthusiasm than others.
And some play up the connection more than others. Current police minister Jack Elder, for example, was not prepared to be interviewed for this article. His press office made it clear he is less than keen to talk in public on the subject of INCIS. He’s clearly taking an interest in the project, though, and has asked for fortnightly reports on progress. It is not clear how much say he has to do with the project--which was what Computerworld wanted to ask him.
The division of benefits is going to be an issue particularly as it does rather beg the question of what the gains will be.
And there’s the rub. The underlying assumptions on what INCIS will bring to policing stand or fall on projected administrative efficiencies. The system will allow police to cut down on the time they spend on administrative work and spend more time on the beat.
“We wanted to provide an electronic or computer system,” says Matthews, “together with all the other things that go with that--a distributed network with a whole lot of PCs on it that can download and manipulate data, provide front-line officers and support officers with the information to effectively predict trends, analyse data, manage cases, minimise paper flows and reduce the amount of administration required to manage the volume of information and paper through the organisation and thereby to increase productivity, in more hours out on the line in terms of average officers’ time.”
While the number of police jobs--mostly administrative--eliminated by the system is publicly stated to be in the vicinity of 540, the initial business case for INCIS actually says the increased efficiencies will be worth 800 front-line staff. The difference--between the 800 and 540--again reflects the 70/30 split in distribution of benefits.
Both the police and Treasury hired independent consultants to analyse the business case before it went to Cabinet, and both, says Matthews, suggested the police estimate of likely benefits was conservative and they would in fact be higher than projected.
The Police Association, initially happy with the reductions, is now not so sure. At the time INCIS was approved, in early 1994, the government of the day asked whether the association was happy about the change.
“We went to the service organisations,” says Matthews, “and they agreed they wanted INCIS, they accepted the reduction in numbers. Now I should imagine that government would feel a little bit let down on that. I think that’s unfortunate, really.”
It’s unclear whether the change is due to growing scepticism about the value of INCIS or pressure from members because of unhappiness about the reductions already made--before INCIS is even working-- coupled with increased workloads. Police Commissioner Peter Doone conceded to last week’s Justice and Law Reform Committee at Parliament that police morale at present is not good.
INCIS has thus far been a series of delays. The first phase of the rollout was due last month, but will now be in December. While hardware will be delivered around July, the first phase of applications will not be available until the end of the year, and the second phase is not due until December of 1998.
A series of different events have contributed to this. Cabinet approved the deal in June 1994, but contract negotiations with IBM took longer than expected, and work didn’t start until September that year. The usual pressures of any information technology project--“mission creep” or added functions to the original brief, plus changes in technology--have also played their part.
“Firearms, family violence and traffic offences have all been added,” says Matthews. “Bear in mind when we first went with the business case we hadn’t merged with Transport, so we had to modify that part way through. We didn’t want to revamp the business case because that would have put us way behind in terms of time and budget. So we noted there were some additional aspects to be incorporated and we’ve done that.”
Technology substitution also played its part. There have been several, such as TCP/IP for SNA, and an upgrade to the microwave wide area network, but most controversial--certainly within the development team itself--was the change from OS/2 to Windows NT on the desktop, which, says Matthews, “we agonised over for quite some time”.
“So that’s been an aspect of the delays, and of course cost rises too.”
A further hiccup--though not conceded as such by the police--was the recent departure of director of information technology Greg Batchelor.
However, Matthews believes the cost is not significantly different from the original contracted price--just under $90 million for the capital part.
“We believe it’s probably going to be--I’m talking about the IBM price--$95 million to $100 million. We’ve still got aspects we can minimise. You’re probably looking about 11% over the whole project. In other things we’ve being able to wind back the cost.”
Matthews admits to being uneasy talking about the figures.
“We’re in a little bit of a difficult position because there’s all sorts of figures being bandied about in the market and at the same time we have to honour our arrangements with IBM that we will not disclose figures. The 11% figure seems to be appropriate overall given the scale, scope, time and some of the technology substitutions. And there will be increased functionality with those costs.”
The budget breakdown thus far is just short of $100 million on capital investment--the IBM part of the contract. IBM has subcontracted a considerable part of this to EDS, which is doing the data conversion work from the old Wanganui Computer. The remaining $100 million is operating costs, spread over eight years. Thus far $32 million has been spent.
There are plans to sell INCIS offshore.
“If we could have--and we tried to--got a shrink-wrapped product we would have. We looked all around the world--at that, and at buying from two different companies and cobbling the systems together. That would have had far less risk than building something that’s right out in front. But we couldn’t.”
The police are banking on other police forces around the world being in a similar position, and see a ready market there for INCIS.
“If they approach it like we’ve approached it, and said is there a system already there that fills our needs, even if it’s 90% of it, would they buy it? I would.”