Perceptions

How can we reconcile the fact that crime rates have been falling for the past seven years with the current perception (targetted by Brash) that it is increasing? The answer is right there in Brash's speech:

Every day, the media carry stories of horrendous crimes...

Our perception of crime is shaped primarily by the media. The problem is that the media's presentation does not nexcessarily reflect the facts. Michael Moore touched on this in Bowling for Columbine, where he noted that, in a period when crime had fallen by 20%, media stories about crime increased by 600%. Sociologist Barry Glassner pins this on the television editor's maxim of "if it bleeds, it leads". Crime coverage is both exciting and cheap - all you need is a scanner and a camera to get hours of flashing lights, sobbing relatives, perp-walks and grim-faced police officers. No investigation required.

This isn't just happening in America. A recent study compared newspaper crime reporting in 2001 with coverage from 1992, and found that:

[i]n 2001 crime news (as a proportion of ‘hard news’) had risen to 21.6% in The Dominion (compared with 15.79% in 1992), to 18.83% in the (then) Evening Post (compared with 13.90%), to 24.24 in the New Zealand Herald (21.05%), to 15.17% in The Otago Daily Times (13.79%) and to 21.37% in The Press.

Professor McGregor notes that the general increase in media coverage of crime occurs at a time when recorded crime is at its lowest rate for over a decade, dropping by 12.7% since 1996-7 (Department of Police Annual Report 2000/2001).

I haven't seen any similar studies for TV - any of our media geeks know of any? - but the same trend seems to be operating there.

As for the effects, a 1999 survey by the Ministry of Justice (Attitudes to Crime and Punishment: A New Zealand Study) found that the public

tended to have an inaccurate and negative view of crime statistics and to underestimate the lengths of sentences imposed on offenders. Survey respondents perceived there to be higher levels of crime than national figures suggest. The overwhelming majority (83%) of the sample wrongly believed that the crime rate had been increasing over the two years prior to the survey.

Survey respondents substantially overestimated both violent crime and property crime statistics. Two-thirds believed that at least half of all the crime reported to the police involved violence or the threat of violence, yet police statistics show that the figure is nearer to 9%. Two-thirds of those surveyed overestimated the likelihood of a New Zealand household being burgled. Only 15% came close to the actual figure of approximately one in every 14 households burgled annually.

Generally speaking, it is a bad idea to base your policy prescriptions on misinformation. Unfortunately, Don Brash and his friends in the Sensible Sentencing Trust are advocating exactly that.