Policy for justice founded on numbers that don't add up: The Home Secretary admits his department's crime figures are 'next to useless'. Heather Mills reports

'THERE are a lot of bad statistics in this world but there are few that are more unreliable than crime statistics or anything related to it,' Kenneth Clarke said yesterday on BBC Radio.

The Home Secretary's condemnation of crime figures - relied upon for decades for the development of criminal justice policy - comes at a time when his officials have been relying upon them before the cross-party committee of MPs investigating juvenile crime.

While police and both Labour and Conservative MPs have been reporting an increase in youth crime, Home Office statistics show a sharp drop in the numbers of young offenders.

Department officials say that in 1981 there were 230,700 young people aged 17 or under convicted or cautioned. By 1991 the number had dropped to 149,000.

In contrast, the Association of Chief Police Officers says that in the last 10 years, the rate of juvenile offending has gone up by 54 per cent - it is only the detection rate that has dropped.

A survey by the Independent supports the claim that clear-up rates are dropping, and without identified culprits it is impossible to to say how many crimes are committed by young people.

But there has been a marked increase in burglary and car crime, usually blamed on the young. Of the 5.46 million offences reported in the 12 months ending June 1992, 1.29 million were burglaries, 931,000 were thefts from cars and 572,000 were thefts of cars - all up by more than half since 1989.

It has been suggested that the apparent drop in the numbers of juvenile offenders is due to increased informal cautioning by police - which would not show up in statistics; cases being thrown out by the Crown Prosecution Service before trial; petty motoring offences being no longer recorded; and a sense of disillusionment and apathy on the part of police officers who believe the courts do nothing when young people are brought to court.

Home Office officials accept that their figures need to be treated with caution because of the fall in the number of young people in the population and a possible increased use of informal warnings. But they maintain government policy since 1981 has led to a drop in juvenile offending.

But the debate about the merits of crime statistics goes much wider. The British Crime Survey - generally regarded as more reliable - suggests that most offences are unreported. For every crime recorded, two are not.

Further, bulges in crime statistics can occur every time police in one particular area target a particular type of offence. For example, recorded sex crime in Thames Valley soared when police decided to clamp down on homosexual activity in a public lavatory in Slough.

Professor Norman Tutt, co-author of Children in Custody, said that figures collected within the criminal justice system itself were in general accurate - for example how many people were imprisoned and for what, how many were cautioned and passing through the courts. 'The problem is that the criminal justice system has little connection with criminal behaviour on the streets,' he said. Those cautioned or convicted accounted for a tiny proportion of crime.

'If Kenneth Clarke is accepting that, it puts him in a serious predicament. It would prove that his policies of tightening up on sentencing would be doomed to fail. Locking up a few does not stop the bulk of offending or lessen crime. It might suggest there needs to be new policies.'

Yesterday Tony Blair, shadow Home Secretary, said: 'Mr Clarke appears not to know whether to defend his own figures or denounce them. It is his reponsibility to ensure the Home Office conducts some real and in depth research into juvenile crime.'

Leading article, letters, page 22