Peter Graham, a director of large projects at Fujitsu, was good-natured with critics who questioned whether Libra, a single national system for UK magistrates courts, would ever work. "I can absolutely guarantee," he said, "that the system will work. The amount of effort and resources being applied to this is beyond my experience."
Since that comment last summer Graham has left Fujitsu and the company has been relieved of its contractual obligation, under the Private Finance Initiative (PFI), to deliver the core "Libra" caseworking software for magistrates courts in England and Wales.
Today Libra's financial sponsors, the taxpayer and the Lord Chancellor's Department, are on the verge of funding a new attempt to provide a standardised system to support the preparation of court cases, and the scheduling and results of hearings. It will be the fourth attempt at building a national caseworking system in 12 years.
Some of those who watched leaders of the Libra project come and go believe that Graham is one of the most gifted experts in the IT industry, which raises a question no one at the Lord Chancellor's Department seems to have dared asked in earnest: is it possible to deliver a single national system without first overcoming managerial and cultural barriers that have prevented the department simplifying its dated and overly-complex business processes?
Specialists in criminal justice systems have no doubt that it is technically possible to build national caseworking software to replace the three incompatible legacy systems that have been in use in hundreds of magistrates courts since the 1970s and 1980s. The question is: will it ever be possible to overlay a new national system on antiquated business processes and practices?
When the Passport Service introduced systems from German supplier Siemens in 1999 it met disaster, in part because its business processes and practices - and staffing levels - were not adapted to take into account the fact that the new system was more secure but slower than the one it replaced. In the end the systems worked well, once new working practices were brought in alongside the technology.
In another Home Office department, the immigration service, a national caseworking system, also being built by Siemens, was abandoned as too difficult to implement. Instead the supplier helped civil servants to modernise their business processes.
Within the magistrates courts, however, there are no signs that any large-scale simplification to paperwork or business methods will precede or accompany the next attempt to introduce a national system.
One senior specialist in criminal justice systems said, "If nothing else the number of disasters in the department should have taught all concerned that the technology has little to do with it. Projects have failed in the past because the managerial and cultural barriers have been too big to surmount."
He added that that the Lord Chancellor's Department has not seriously asked itself, nor satisfactorily answered, crucial questions on project management that have little to do with technology, before embarking on a new iteration of national system. These include:
- How do you get buy-in for a new system from thousands of employees in magistrates courts who have grown comfortable with systems and practices that have evolved separately and inconsistently over decades?
- How do you get staff to accept that paperwork, procedures and working practices need to be first simplified and then computerised when employees will (probably correctly) see the Treasury as requiring the upfront costs of modernisation to be paid for by cuts in the number of employees?
- How do senior officials and ministers gain support for their plans from staff whose experience of the Lord Chancellor's Department has been that it waved aside constructive criticisms of the department as being based on ignorance, or worse, it attacked critics as motivated by malice or stupidity? The department has responded to media criticism of its IT decisions in the past, for example, by sending reassuring circulars to justice chief executives.
- How is it possible to instil a fear of failure in the department's senior officials when their careers suffer no damage, no matter how many disasters they preside over; and why should suppliers fear failure when they get paid regardless? In the Libra project Fujitsu will be paid £49m more than agreed when the original contract was signed, and it will deliver only one third of the its objectives.
Some people fear that these big issues will not be addressed in audits of the failures of the department's IT projects, which may instead focus on the minutiae of project management: whether too much or too little risk was transferred to the suppliers, whether the supplier had allocated enough people with the right skills to the project, or whether the department was over-prescriptive in its specification or not prescriptive enough.
This could mean that the major lesson, the need for structural change, will not be learned by the time the contract for a national system is signed - which could be in the next few weeks.
Computer Weekly has learned that the department is negotiating a deal for a national system with STL, which supplies some of the legacy systems to magistrates courts. It has also emerged this week that a separate integration contract will be signed with one of three shortlisted suppliers - Unisys, Logica or Accenture - to implement STL's system in all magistrates courts.
Under the original PFI contract, signed in 1998, Fujitsu was obliged to provide office automation, caseworking software, and systems integration - all for £183m. But Fujitsu is to be paid more than £232m for delivering only office automation. The department refuses to say how much the contracts with STL and the systems integrator will be worth.
Come what may, it appears, a contract for national system number four is due to start shortly, without any structural change in the workings of the magistrates courts or the approach of the department.
Clearly senior officials and ministers believe it will work. But some specialists think that, given the cultural and managerial problems facing the department and magistrates courts, any new attempt at implementing a national system will be doomed, just like its predecessors.
A Lord Chancellor's Department spokeswoman was unable to say whether the department had considered whether it should simplify paperwork, antiquated business practices, procedures and practices in magistrates courts before computerising.
What is the Libra project?
The main purpose of the Libra project is to speed up the criminal justice system, helping Labour to achieve one of its manifesto pledges.
The first major contract for a standardised national system in magistrates courts to replace three ageing and incompatible legacy applications was signed in 1992 with Price Waterhouse but ended in legal action, which the Lord Chancellor's Department won.
After a second failed attempt at a national system, a third set of contracts was signed in 1998, this time with Fujitsu, worth £183m.
The main aim of Libra was to enable case files to be transferred electronically between courts, the police, the probation service, prisons and other criminal justice agencies. The National Audit Office reported in 1999 that "efficient case progression is dependent on the different organisations providing and exchanging information accurately and on time".
Libra's core application lay at the heart of the Government's manifesto pledge to speed up the time courts take to deal with offenders. Fujitsu was contracted to deliver a technical infrastructure comprising Microsoft Office automation and networked PCs to support an Oracle-based core "casework" application.
Now Fujitsu will deliver only the office automation and infrastructure, not the core software. Instead the department is reverting to a plan which dates back nearly 10 years - which had been rejected - to enhance existing legacy applications to provide one standardised system.
Meanwhile case files continue to be exchanged between courts, the police and the Crown Prosecution Service by means of fax and post.