New attack on stabbing statistics

The government claimed a knife crime crackdown was working

Statistics chiefs have renewed their criticism of government knife crime statistics - including a claimed 30% reduction in violence at Halloween.

Home Secretary Jacqui Smith has already apologised for the selective use of statistics in a December press release.

She admitted figures suggesting a 27% fall in stab wound hospital admissions had been released too early.

But further criticisms have now emerged - including the claimed reduction in violence during "Halloween week".

Sir Michael Scholar, head of the UK statistics authority, mounted an unprecedented attack on the government's presentation of crime figures when the press release and an accompanying two page fact sheet were published in December.

Sir Michael accused the government of releasing the figures without proper checks and of manipulating them for political ends.

He accepted Ms Smith's apology for being "too quick off the mark" in releasing the figures but said the problems with the figures ran much deeper than that.

'Unsubstantiated claims'

Now the UK Statistics Authority, which was set up to improve public confidence in official figures, has published a detailed breakdown of mistakes and distortions in the December press release.

It highlights the "selective or otherwise inappropriate comparisons," lack of contextual information, the drawing of "inappropriate conclusions" and "unsubstantiated claims," including "those caught with knives are now three times more likely to be sent to prison".

The original Home Office release claimed the results of a knife crime crackdown launched in June in 10 police areas were "already clear".

It listed the results as "fewer teenagers carrying knives, fewer young people carrying knives, fewer offenders getting off with a caution, more people going to prison".

The publication of our new Code of Practice for official statistics marks the beginning of a new chapter for the Statistics Authority Sir Michael Scholar

But it contained no information about the "source and quality of statistics nor the use to which the data can legitimately be put" and mixed references to "young people," teenagers and "youth violence" without saying what it meant by such terms, the statistics authority says in its report.

The release also failed to take into account seasonal factors when claiming a reduction in violence and its claim that "over 105,000 stop and searches for offensive weapons were carried out" was published without saying whether numbers had gone up or down or how they compared with other areas.

The statistics authority also takes issue with the fact sheet's claim there was a reduction in violence at Halloween.

The report says: "Youth violence is reported as being 30% lower in Halloween week than in the previous year.

"Halloween week is not a recognised period for statistical comparisons. And no evidence is given about the reasons for this change - it could be because of the weather or other external factors".

The Metropolitan Police have stepped up patrols over Halloween in recent years due to a reported increase in robberies by gangs posing as trick or treaters.

Code of practice

Government figures which fail to comply with the UK Statistics Authority code of practice, which was published in full on Tuesday, will in future be refused a UKSA stamp of approval.

The UKSA also published a list of 340 sets of official statistics which are not covered by the code and it thinks should be.

Sir Michael said: "The publication of our new Code of Practice for official statistics marks the beginning of a new chapter for the Statistics Authority.

"We now have the guiding principles in place against which we will assess official statistics to determine whether they meet the standards necessary to be labelled as National Statistics.

"Compliance with the Code allows official figures to carry the National Statistics label, and the assessment process helps to ensure consistently high standards of service, as well as identifying areas that need attention.

"The Authority wishes to see the National Statistics label recognised as an assurance that the statistics have been produced and explained to high standards, and that they serve the public good.

"By looking at whether the statistical system is meeting the needs of users of statistics and wider society, it will help, over time, to build trust in official statistics.

"I believe the new Code of Practice will help the many producers of UK official statistics to collect, explain, and release those statistics in ways that build public trust and confidence in the statistical system as a whole."