Uncertain meaning of crime statistics

A leading criminologist has claimed that a "Drop in crime reports 'shows trust in authorities'" (November 9). It might, and then again, it might not.

It might well mean that people have lost faith in the authorities and aren't bothering to file reports any more. This is acknowledged some way into the article, but only as a reluctant afterthought. The write-up also carries a tabulated statistical analysis of crime reports over a four-year period but this comparative exercise uncovers some inadvertent flaws in the criminologist's post-hoc reasoning.

The "drop" he speaks of is largely limited to reports of property damage, murder, attempted crime and theft.

Reports of injury, domestic violence, drug use, forgery, fraud, sexual offences and violence against public officers bucked the trend and actually experienced an upsurge. It is ironic to note that reports of illegal immigration have gone down a good 48 per cent during the same time frame; a bit of news which I expect will not go down well with those who so hysterically have been proclaiming the contrary.

Will people go along with these numbers? Will they be critical of them? Will they even care?