Curmudgeons may bemoan the death of British inventiveness, but the greatest modern breakthrough in the fight against crime was made by an Englishman, Alec Jeffreys, in 1984. Jeffreys wasn't even looking for DNA fingerprints when he discovered them. As with all the best stories about scientists, he had a blinding flash, a "Eureka!" moment, when he realised that the blood samples he was examining for evidence of inherited diseases had much more to tell him.
Because of his discovery, the police can now convict or release a man by examining a few cells in a hair, and immigration and paternity cases can be settled conclusively. When the Americans wanted to know whether a dirty old man in a hole was Saddam Hussein, or the Russians wanted to know if remains found in a grave near Ekaterinberg were those of Nicholas II, DNA fingerprinting provided it.
This is how technology is meant to work: bringing certainty where once there was doubt; cracking cases which would once have been insoluble; offering the possibility of a world where the guilty are always convicted and the innocent always walk free.
Unfortunately, Jeffreys's triumph is something of a rarity. No sector of public life has been as infested with techno-utopianism as the criminal justice system. As with DNA, the appeal has been to make new connections. Modern societies are saturated with information, the reasoning went. If only it could be collected and sorted, then justice would be done. It sounded a reasonable aspiration. Since the dotcom boom, the developed world has been obsessed with the benefits that the explosion in computing power would bring. Yet, when the Home Office tries to fight crime, it achieves little beyond wasting hundreds of millions of pounds of other people's money. The courts, prisons, immigration and probation services have had systems imposed on them which have had the remarkable effect of leaving their crime-fighting capacity worse than before.
Crams - Case Record and Management Systems - was meant to provide probation officers with intelligence on offenders and assessments of the risks they were likely to pose. Alas, the software refused to recognise aliases or villains whose first names echoed their surnames. This caused terrible trouble in Wales. "We've got loads of Jon Joneses and Dai Davises and Tom Thomases down here," a Pontypool probation officer told me. "Every time we put their names in, it throws a complete wobbly." To reach the information it did recognise, wretched probation officers had to go through 11 stages before they could access files, a process that could take 15 minutes or longer. It was scrapped along with all the others.
Part of the explanation for recurrent failure lies in new Labour's reliance on management consultants. Neither politicians nor the civil service seem to grasp that consultants have an incentive to provide bad advice, because they come from a rapacious culture which aims to get their bills as high as possible. The sensible course for any organisation, public or private, is to buy an established system whose problems have been sorted out, and adapt it. Law enforcement agencies around the world would be happy to rent the Home Office their software under licence.
Indeed, much of what government wants technology to achieve doesn't even need a specialist system. In essence, there's no difference between a programme that allows court officials to book a defendant's appearance in the dock and one that allows you to book a hotel room or flight. Management consultants rarely point out these obvious truths. Instead, they send their number of billable hours heavenwards by recommending that Whitehall creates new systems from scratch, with all their costs and glitches. Presumably, one day we'll have a government that sees through this racket, but it may well face more serious constraints.
A disaster story which never received the attention it deserved was the failure of the Criminal Records Bureau. It was meant to provide copies of criminal records or letters of good character to anyone who applied for a job. The idea had to be abandoned because detailed inspection of the records showed that far too many were hopelessly wrong.
If enough time and money were spent, criminal records could be tidied up, but would it be worth it? The same question haunts the government's identity-card scheme. If the public is going to pay at least £6bn, wouldn't that money be better spent employing more police and intelligence officers? My guess is that this is the question that will kill the scheme. Like so much else the Home Office attempts, it wastes energy by trying to monitor the law-abiding as well as the lawless, and thus falls foul of the law of diminishing returns.