Saturday, 6 May 2017

Perceptions Of Violent Crime

I spent last week as a parent helper at my child's school camp at Tui Ridge near Rotorua. I found it exciting and challenging in various ways, not least just having to exercise social skills non-stop for five days solid. I spent a fair bit of time socializing with the other parents. One conversation came around to the topic of violent crime, and some of the other parents talked about how violent crime is getting worse and there seemed to be a murder every day in Auckland. I ventured to suggest that violent crime, especially murder, is declining here (and elsewhere), to general skepticism. Unfortunately I didn't have data memorized and I drowned my phone recently, so I left it at that.

On returning home it didn't take long to confirm my point. For 2014 (the latest year for which I can find data) the police reported 66 "homicide and related offenses" nationally. That is the lowest number since that dataset began (1994), when the numbers were much higher. (I don't know why numbers for 2015 and 2016 aren't available in that dataset, but I guess bureaucratic inefficiency. There appears to be more recent data at, but it's behind a Flash app so I can't be bothered.) The idea that "a murder occurs every day in Auckland" was off by an order of magnitude, in 2014 at least, and there seems no reason to believe things have gotten worse.

Wikipedia has a good discussion of how in New Zealand (and other Western countries) people tend to believe violent crime is increasing when in fact it has been generally decreasing. I've read all sorts of theories and surveys of theories about the causes of this decrease, from "increased abortion" to "declining lead levels", but it appears no-one really knows; there are probably multiple causes. It's certainly welcome!


  1. There are a lot of rationalizations why public perception is so skewed. It seems reasonable to accept that people are more influenced by the which direction change is occurring, either increasing or decreasing compared with absolute figures. Crime has certainly been increasing in the last few years (with police funding being slashed), but that has only been to pause the overall general decline.

    Petty crime and resolution statistics definitely haven't improved as much, and people have the tendency to extrapolate. If people perceive it not to be constantly decreasing, surely it must be doing the opposite. Media reporting has a big impact on perception, and better reporting and the highlight of increases in youth crime make for more shocking headlines.

    Throwing out another crazy theory: How about a plot of violent crime vs time spent indoors the last several decades?

    1. > Crime has certainly been increasing in the last few years (with police funding being slashed)

      What crime, and where? What data are you looking at?

      > How about a plot of violent crime vs time spent indoors the last several decades?

      Yeah. I think "angry young men are spending increasing amounts of time playing video games instead of making trouble" is as plausible a theory as any.

  2. Not only has reported violent crime been dropping, but the definition of violent crime has been tightening. As recently as the 1990s it was acceptable for a teacher to assault a child with a weapon, while domestic violence was disregarded and drunken brawling was a routine fact.

    It could be that people are more scared of being beaten up because the have become so unused to it, though I think the perceived murder rate is due more to it being good cheap news. There is the event itself, the funeral, the police investigation, the anonymous court appearances, the revealing of the accused's name, the court case, profiles of those involved, and some years later a series of parole hearings. It just gives and gives. Each murder provides dozens of headlines over stories requiring very little work.

  3. I'd say it's a mixed bag. Looking at


    Both very serious, both sadly increased. You also should keep in-mind that unreported crimes are also a factor in any society. In my nation (UK) you'd find it easy to be lulled into a sense of propriety and well-natured population, but it varies by location.

    A few months ago a group of thugs attacked and racially abused a local business owner. The police didn't act. A few weeks ago
    what can only be described as a feral gang of brats broke our block door to retrieve a football. Within the past 6-months I've seen road-rage violence, public intoxication leading to violence, brazen drug-dealing and youth gangs (between 11 and 25); there has been a rise in acid-attacks, terror attacks.

    I think it's easy to be dismissive of people, which is why there are such problems.

    1. I'm sorry to hear about such incidents, but statistics trump anecdotes.

      You're right that reporting rates (and crime definitions) make tracking these numbers over time difficult, but that's exactly why people studying crime often focus on homicide rates --- relatively easy to define, relatively difficult to conceal or ignore. On the other hand, harassment and sexual assault are exactly the sorts of crimes whose definition has changed over the years and that are highly subject to changes in reporting behaviour.

      Given you looked at statistics for lots of different kinds of crimes over just a couple of years, I'm not surprised to see increases in some categories --- any long-term trends are hidden by noise.