It’s a crime

36 comments on “It’s a crime ”

• Comments are now closed

1. The Sprout 1

   2 February 2009 at 11:03 am

   agreed, that article’s well off form for Armstrong. his health is deteriorating, maybe that’s got something to do with it.
2. BLiP 2

   2 February 2009 at 11:32 am

   What is up with Armstrong? Has he been got at by the Goober’s traitor journo’s turned spin doctors? Do they have something over him that has resulted in his latest effusive National Party knob-gobbling?

   I’m wondering if the Gallery has decided to give the gNats a free run for the first 100 days and then will resume their role as commentators and not cheerleaders, and fact checkers not policy pushers.

   Has anyone seen http://www.mediamatters.org ? We are starting to need something like that for our own media – would require heaps of work, though. Anyone interested?
3. Greg 3

   2 February 2009 at 11:37 am

   Yeah but isn’t violent crime rising?
4. Lew 4

   2 February 2009 at 11:40 am

   Greg: Yeah but isn’t violent crime rising?

   Reported violent crime is rising. The police themselves say this is to do with higher reporting of domestic violence.

   L
5. Lew 5

   2 February 2009 at 11:49 am

   BLiP: Aside from your use of terms like `knob-gobbling’, there’s a credibility problem:

   Media Matters for America is a Web-based, not-for-profit, 501(c)(3) progressive research and information center dedicated to comprehensively monitoring, analyzing, and correcting conservative misinformation in the U.S. media.

   Partisan media agencies don’t have very much. The world doesn’t need a left-wing version of Fox News. There’s plenty of misinformation being put about by the so-called left as well – and far more genuine error and incompetence and overwork. Read Nick Davies’ Flat Earth News.

   I might be interested in being involved with a bipartisan analysis and fact-checking agency, but you’re right – it’s a simply insane amount of work, and thankless at that. Even in a tiny media ecology like NZ’s.

   L
6. @ work 6

   2 February 2009 at 12:15 pm

   Maybe he means Asian crime?

   (sorry, couldn’t resist)
7. jbc 7

   2 February 2009 at 12:55 pm

   “Reported violent crime is rising.” I find it curious how the violent crime stats are often brushed off with this. Some even go as far as suggesting that this is actually a good thing – but remember that this is not the first time this explanation has been used. I vaguely recall it being trotted out in the 90’s too. Do the police ever offer that explanation when reported crime is falling?

   The brits have their own twist on this problem.

   Surely at some point the rise in reported violent crime will become a real problem though. Sometime within the next 3 years I suspect.

   Bottom line is that I’d trust the police crime figures about as much as I trust a fund manager’s advertised returns.
8. Lew 8

   2 February 2009 at 1:29 pm

   jbc: Bottom line is that I’d trust the police crime figures about as much as I trust a fund manager’s advertised returns.

   Yes, the age-old conundrum of how to measure crime rates. If not the police, who would you trust? Who doesn’t have a dog in this fight?

   L
9. Tim Ellis 9

   2 February 2009 at 2:12 pm

   Serious violent offending has risen. Armstrong just doesn’t spin a story in the direction you would like it to lead, SP.
10. Lew 10

    2 February 2009 at 2:44 pm

    Tim: Serious violent offending has risen.

    No, reported serious violent offending has risen. If you want to argue that it’s not for the reasons the Police claim (higher awareness, better responsiveness, etc), then it’s incumbent upon you to demonstrate why the police are wrong. JBC has started, and it’s certainly an arguable case. So argue it.

    L
11. Jimbo 11

    2 February 2009 at 3:07 pm

    Lew,

    I don’t understand why you’re so excited about the “reported” aspect. So what? Even if you’re right that the stats only show a greater level of “reported” crime, why are you so happy about it?

    “Reported” crime being up ALSO means that the problem is BIGGER that we previously thought and needs additional measures to deal with it. Or is my simplistic mind missing something?

    (As an aside – how do you know that the “unreported” element has not also increased, in real terms? Let’s face it, saying that 40% of domestic violence (or whatever) now gets reported is total guesswork – by definition no-one knows how many “unreported” incidents there are!)
12. Draco T Bastard 12

    2 February 2009 at 3:17 pm

    I don’t understand why you’re so excited about the “reported’ aspect. So what?

    I was going to answer this but you answered it yourself in the last paragraph. Which just shows that you’re not ignorant of the facts but that you don’t want to believe them.

    In answer to your last question in your last paragraph – statistics. You go out and you make surveys and then extrapolate those results to the rest of the population using a process known as statistical analysis. It’s not perfectly accurate but it’s not far off and it’s certainly better than nothing.
13. Rex Widerstrom 13

    2 February 2009 at 3:44 pm

    BLiP:

    The trouble with mediamatters.org lies within its mission statement:

    …systematically monitor a cross section of print, broadcast, cable, radio, and Internet media outlets for conservative misinformation… [my emphasis]

    There’s something not quite credible about pointing out bias in the media from a position of… well, bias. I’m not denying Media Matters does a good job – but it only does half of one.

    If someone, of any political persuasion, truly believes that the media need to be held to account for inaccuracy then they should have the integrity to do so even when the correction disadvantages their “side”.

    And besides, imagine a “Media Matters” run by commenters who contribute regularly here, for instance. The standard of proof required to get everyone to agree that a correction be issued would be so high, its accuracy would be unquestionable 😀
14. Rex Widerstrom 14

    2 February 2009 at 3:53 pm

    Lew points out:

    Yes, the age-old conundrum of how to measure crime rates. If not the police, who would you trust? Who doesn’t have a dog in this fight?

    Like jbc I don’t trust the Police’s statistics. That’s because I have personal experience (as a journo) of one district under reporting violent crime so as to make their clean-up rates look better. They’ll massage them to tell whatever story is most advantageous at the time: Copping a bit of flak? Reduce the stats and we look effective. Want more power? Increase them and we’ll get it to fight “growing lawlessness”.

    Just like unemployment figures based on WINZ’s criteria are unreliable, so most people use the Household Labour Force Survey as a measure. Statistics NZ doesn’t have a dog in any fight.

    It would be a simple matter for the data collection instruments already used by Statistics to collect a variety of information to have added to them some questions about crime.

    Have you been a victim? What sort of crime? Did you report it? would give us vastly better data than we get from the Police, and more questions would of course allow even better analysis.

    So why don’t we?

    Because that would help end the “law and order auction” that – despite recent calls for it to end – neither National nor Labour are willing to stop in their chase for the fear vote.
15. Steve Pierson 15

    2 February 2009 at 4:02 pm

    jimbo. the issues around crime stats have been covered and re-covered, see our archives.

    what lew says is what the experts say, and i tend to believe them (if their arguments make sense) rather than people trying to make reality justify their iedology.
16. Lew 16

    2 February 2009 at 4:03 pm

    Jimbo: I don’t understand why you’re so excited about the “reported’ aspect. So what? Even if you’re right that the stats only show a greater level of “reported’ crime, why are you so happy about it?

    Oh, no, I’m not happy – I just think it’s important to understand the difference between the two things. All statements like `crime is up’ and `crime is down’ are actually based on evidence which says something else. How the evidence becomes the statement is a matter of explanation – or you might call it `spin’. Certainly the police have a motive to make reported crime look higher as a result of their actions while arguing that the actual crime rate stays the same or drops, just as other groups have motives to explain it in other ways.

    So I’m arguing that if the police say reported crime is up due to higher reporting rates, then that’s a good thing – IF you believe the police rationale. If you don’t, there’s an onus on your to provide an alternative (more credible) rationale. This is by no means impossible – depending on the numbers in question it might even be easy – it can’t just be stated blankly as fact without any supporting evidence as Tim has. The police, for all they have a reason to make themselves good, do a huge amount of work with crime figures, and are subject to a whole lot of oversight on matters like this. That lends them credibility.

    “Reported’ crime being up ALSO means that the problem is BIGGER that we previously thought and needs additional measures to deal with it. Or is my simplistic mind missing something?

    Well, it means the problem is bigger IF we previously presumed that we knew the full extent of the problem. Agencies who deal with crime statistics on a habitual basis tend to be extremely cautious about presuming such things. Technically, reporting rates are best understood as a lower bound.

    L
17. BLiP 17

    2 February 2009 at 4:05 pm

    Lew and Rex

    Yep – fair enough, lads. I concede, Media Matters is biased and probably not a good model for a credible check on the NZ MSM. A bi-partisan approach would be best.

    I wonder where it comes from that the left feels just as agrieved with apparent MSM bias as the right. In some ways I think the MSM is trying to have its cake and eat it as well – attempting fairness but actually annoying everyone.

    I dunno. Now, where did I put my thinking cap . . .
18. Lew 18

    2 February 2009 at 4:09 pm

    Rex: Good, you proved my point – a case, if somewhat conspiratorial, against the police explanation.

    Yes, I agree that there are better ways and more credible authorities – university criminologists, for instance, do a great deal of research in the way that you describe. As to why we don’t do more – I guess (I don’t know) it’s for logistical reasons. It’s incredibly time-consuming and expensive to conduct research of this type, and methodologically more complex than you make it seem. There are also political problems with entrusting this sort of things to `leftist academics’, which in the context of the law and order auction are non-trivial. However I agree that that’s what we need more of.

    L
19. Lew 19

    2 February 2009 at 4:12 pm

    BLiP: The Center for Public Integrity isn’t quite what you seek (it deals in original long-term in-depth investigative journalism, not tactical fact-checking), but it’s a useful organisational model in principle.

    L
20. BLiP 20

    2 February 2009 at 4:31 pm

    Jimbo Said

    ” . . . I don’t understand why you’re so excited about the “reported’ aspect. So what? . . . ”

    As I understand it, the events leading up to and including the shooting on Western motorway, for example, would have been reported by heaps of people – the way the police gather their statistics, each report is counted so you get a situation where one event attracts mulltiple reports so, hey presto, “rising crime statistics”.

    Further clouding the data are, for example, the advertisements encouraing greater community action in the quest to eliminate domestic violence. In the past a “domestic” would have gone largely unreported but nowadays, I understand, there could be three or four reports for the same incident.

    I like the idea of putting the collection of crime data into the hands of the professionals at the Department of Statistics (is this the only “Department” left in government these days?) who have nothing to gain from providing anything other than accuracy.
21. Jimbo 21

    2 February 2009 at 4:31 pm

    Lew,

    I’m not sure why the onus should be on the public to prove or disprove anything about the crime stats – you’ve said that a couple of times in your posts above. It’s the people who run the “reported” line (i.e. police, etc) that need to do the justifying, surely?

    Step 1: Crime figures come up, and show larger number of convictions, investigations, etc.

    Step 2: Police (or whoever) says “Don’t worry, actually only “reported” crime is up” (and actual crime has fallen?)…

    Step 3: I say “How do you know that? What methods are you using the work out the etimated “unreported” crime figures? Isn’t this total guesswork since “unreported” crime is, by defintiion, something we don’t know about.”

    Step 4: ???

    The crime stats are of course not a perfect measure. But they ARE the measure we have and they DO show an increase in (reported) crime. If, however, the police believe that crime is in fact FALLING because unreported crime is falling (something that is pure hypothesis), surely they’re the ones who have to prove it…?!

    How does the Police estimate the “unreported” crime, and what makes them think that this sub-set of crime is now smaller than it used to be?
22. Jimbo 22

    2 February 2009 at 4:35 pm

    BLiP – that’s crazy if one crime gets counted multiple times. Certainly should be changed if that’s the way they currently do it. Agree with your suggestion that Dept of Statistics should collect (or at least audit) the stats.
23. Lew 23

    2 February 2009 at 5:07 pm

    Jimbo: I’m not sure why the onus should be on the public to prove or disprove anything about the crime stats – you’ve said that a couple of times in your posts above. It’s the people who run the “reported’ line (i.e. police, etc) that need to do the justifying, surely?

    On their own, those figures don’t tell us very much – methodologically speaking, you can’t validly look at an increase in reported crime and assume that there’s been a commensurate increase in committed crime; there are a lot of other factors in play. There are a bunch of possible explanations for the change, and `crime has gone up’ is certainly one. In a way it’s like assuming from a 10% increase in vehicle traffic that the price of fuel decreased by 10% – there’s probably a link between the two figures, but it can’t be relied upon without further evidence.

    In any case, the Police do justify their arguments. You can disbelieve or query or argue the toss, but fundamentally if you aim to discard their explanation, then without an explanation of your own you’re saying nothing more than `I reckon …’, and that’s noise, not signal.

    BLiP’s also right – there are a lot of methodological problems in play. Yeah, they’re the figures we have, but rather than just pretending their imperfections don’t exist, it’s best to try to mitigate against those imperfections.

    L
24. Lew 24

    2 February 2009 at 5:08 pm

    I should add that while the police do justify their arguments, they also trade heavily on their credibility as the police force – which isn’t ironclad.

    L
25. Rex Widerstrom 25

    2 February 2009 at 5:43 pm

    Perhaps those here who deal more in statistics than do I could comment on whether a “snapshot” would be of much use. Like, for instance, a section on crime in the Census.

    While not perfect, would this not allow us to do two things:

    1. Compare the accuracy of other statistical measures, most notably those of the Police; and

    2. Track change across the entire populace over time, albeit a very intermittent sample rates.

    The only cost I can see in this – which I’m in no way offering as a perfect solution – would be in a bit more analysis and reporting, which surely wouldn’t be huge?
26. Jimbo 26

    2 February 2009 at 7:35 pm

    Lew,

    You raise some fair points, but I think you overstate the “vagueness” of the figures.
    It’s not at all like your “10% increase in traffice MIGHT lead to 10% decrease in fuel”. The stats we’re talking about are direct stats about the point at issue. Police are asking us to read/interpret those stats with reference to another factor (unreported crime) for which they cannot give us any information at all…!

    When most people think of crime rates, they think “how many times have the police had to deal with a crime?”. The Police obviously have stats on this, and that’s what gets reported to the public at large.

    There is no possible way to track every crime committed in NZ – surely looking at how many the primary crime-enforcing government agency deals with is a pretty meaningful proxy?

    What we’re now being told (and what Steve said in the original posting) is that “any journalist with a modicum of professionalism KNOWS” [emphasis added] crime is falling.

    If the stats show an increase, but we’re supposed to read that as an overall fall, then unless I’m missing something the police should really be proving to us why they reckon “unreported crime” has dropped by a greater factor than the increase in reported crime.
27. Lew 27

    2 February 2009 at 10:12 pm

    Jimbo: You raise some fair points, but I think you overstate the “vagueness’ of the figures […] Police are asking us to read/interpret those stats with reference to another factor (unreported crime) for which they cannot give us any information at all !

    Yeah, it comes down to a judgement call, and it’s a very complicated multivariate problem. Your argument is logically valid – you can’t count what you can’t know. But I’m inclined to believe the police on this one because the thing about the It’s Not Ok campaign is that the police will have been consulted on and planned in advance for its launch, and will have (reasonably, because the effects of media campaigns are somewhat predictable) been expecting an increase. When they got the sort of increase they expected, it makes sense to call that a correlation. Unless there were other factors which might have themselves led to an increase in the incidence of (rather than the reporting of) violent crime (in particular domestic violence), then I’d say it’s a fairly strong case regardless. Weaker cases are accepted all the time, absent a better explanation.

    For what it’s worth, crime reportage is a pretty good (and pretty standard) measure of a community’s confidence in a police force; a better measure of that, I would argue, than of crime incidence (though they’re dependent variables). Suppose a government starved the police of resources such that they couldn’t respond to some crimes: people would stop bothering to report them. You’re arguing that since the police have been involved in a campaign which improves awareness of certain crimes, and improves their ability to respond to them, that the incidence of those crimes has increased. Does it therefore follow that if a police force became less competent or trusted and crime reportage went down, you’d argue that the incidence of crime had likewise decreased?

    I work with vague figures on a day-to-day basis (I sometimes, facetiously, refer to my business unit as the Department of Meaningless Numbers) and I’m generally pretty cagey about anything where a correlation isn’t pretty clear; I believe that conservative assumptions about data minimise rash responses. Notwithstanding Rex and jbc’s objections, it seems to me a less-rash assumption that the police know what they’re doing with their statistics than that every extra reported crime means an extra crime committed.

    But it’s something about which reasonable people can disagree.

    L
28. jbc 28

    3 February 2009 at 1:53 am

    Lew, “Who doesn’t have a dog in this fight?”

    [excuse the huge pause – I hadn’t meant to do a “hit and run” comment – lots of interesting and good points on this diversion from SP’s post]

    Much has been said on this and from my own point of view I’m more interested in the honest reality than any particular angle. More than that, I hope that whatever is presented in the media is the honest reality. That “rising tide of Asian crime” article a couple of years back really was a low point – because it was clearly untrue.

    In this case I’m not sure if “rising tide of crime” is accurate – but then I’m equally unsure that crime is decreasing as some would like us to believe. I think we have no option but to trust the police numbers .

    One plausible explanation for the overall decrease in reported crime is that there is increased apathy in reporting minor dishonesty and property crimes. Certainly not a stretch to believe that.

    During the 1990’s my house was burgled twice, my car was broken into several times, stolen once, neighbour assaulted (detectives woke me up with a knock on the front door), and I witnessed an assault on a taxi driver in the city (I called the police and stood by the taxi until they arrived – they were fantastic). The last year before I left NZ my house was tagged once per month on average. All of this in sunny Pt Chev.

    All but the tagging was reported to police. The latter I handled myself (with industrial solvents, gloves and a wire brush) because:
    a) didn’t want to waste police time.
    b) unlikely that it would resolve the problem.

    Based on my own experience I wouldn’t call police for minor crime unless it was required for an insurance claim that was worthwhile making. Police roaming the house at night leaving fingerprint powder everywhere is just not worth the hassle. [yes, in the 1990’s police actually visited on the night of the burglary]

    Earlier today I looked at the police stats and dug a little deeper with the same stats on the statistics.govt website. Fiscal years 98/99 – 07/08. Total violent crime rate rose from 105 per 10,000 to 138 per 10,000 population. About a 30% increase.

    Increased reporting of domestic violence? Perhaps – that’s one category of violent crime that I can understand being kept silent. The thing that gets me is that this increase is spread quite broadly: robbery, grievous bodily harm, etc. It just leaves me a little skeptical of the “increased reporting” line. Why wouldn’t you report a robbery in 1999?

    Perhaps I’m just getting suspicious and over-skeptical as the years tick by – although I think I’ve always been suspicious of people wielding graphs with a vested interest in the numbers.
29. Draco T Bastard 29

    3 February 2009 at 2:11 am

    Police are asking us to read/interpret those stats with reference to another factor (unreported crime) for which they cannot give us any information at all !

    http://www.courts.govt.nz/media/fact-5-3-april-2006.pdf

    They carry out surveys and have a reasonable account of how much crime actually happens in the community. Reported crime does not equal the actual crime rate – It will be less. So, when the number of crimes reported goes up and the surveys say that the amount of crime in the community stays the same or goes down then the police can say that the reported crime has gone up but crime has stayed the same or gone down. They actually do have the numbers to back up what they say.

    But it’s something about which reasonable people can disagree.

    No there isn’t. They could argue methodology and numbers but the overall result will be the same.
30. jbc 30

    3 February 2009 at 2:35 am

    DTB: “They carry out surveys and have a reasonable account of how much crime actually happens in the community”

    Courts know about unreported crime? Hows that?

    But I couldn’t find that from your link. Digging further on that survey gave me this:

    New Zealand Crime & Safety Survey 2006 (NZCASS)
    Key Findings report

    News media fact sheet 1: Responding to crime trends

    1. Is crime going up, down, or staying the same?

    The NZCASS cannot provide a definitive answer in respect to all forms of crime. This is due to significant methodological changes in the survey design from the two previous surveys (the NZ National Survey of Crime Victims) in 1996 and 2001.

    and this:

    2. Why can’t you say what the crime rate is?

    The survey cannot say what the crime rate is because it does not cover all the crimes that occur.[…]
    At the same time, police figures do not tell us what the crime rate is either. The Police only record offences reported to them.

    Which doesn’t actually shed much illumination on the comments above. It’s a light with a broken filament… and flat batteries 🙁
31. Lew 31

    3 February 2009 at 9:53 am

    jbc: In this case I’m not sure if “rising tide of crime’ is accurate – but then I’m equally unsure that crime is decreasing as some would like us to believe. I think we have no option but to trust the police numbers .

    This is almost my argument – qualified by the idea that if anyone is more credible or has a more sound explanation, let them come forth.

    Perhaps I’m just getting suspicious and over-skeptical as the years tick by – although I think I’ve always been suspicious of people wielding graphs with a vested interest in the numbers.

    That’s not over-skeptical, that’s prudent.

    The NZCASS cannot provide a definitive answer in respect to all forms of crime. This is due to significant methodological changes in the survey design from the two previous surveys (the NZ National Survey of Crime Victims) in 1996 and 2001

    Yes, this is a problem with regard to long-term crime statistics – when you change or improve or tune your methodology, results are often no longer comparable.

    DTB: They could argue methodology and numbers but the overall result will be the same.

    In cases like this, there is no `right’ answer upon which we can agree. Methodology determines results to a large extent.

    Jimbo and I disagree over points of methodological validity and interpretation – I accept that his arguments are valid, he seems to assume that at least some of mine are valid, though we both prefer our own. That’s the definition of reasonable people differing.

    L
32. Draco T Bastard 32

    3 February 2009 at 11:01 am

    But I couldn’t find that from your link.

    I used that link more as an example. It was late (early?) and I should have been clearer.

    The point is surveys are taken. From those surveys you can extrapolate out what the general crime rate is and compare with reported crime. If reported crime goes up but the general crime rate as determined through surveys stays the same then there’s only one conclusion.

    Courts know about unreported crime? Hows that?

    By taking surveys. Here’s the link to the full report.

    It measures the amount of crime in New Zealand in 2005 by asking people directly about crimes they have experienced. The survey includes crimes not reported to the Police, so it is an important complement to Police records.

    The new survey leaves out a few crimes but I’m sure the police and the courts would have access to those statistics as well.

    The NZCASS covers reported and unreported personal and household crime experienced by those aged 15 years or over, but not commercial crime, crimes against people living in institutions, nor ‘victimless’ crimes such as drug and alcohol offences. It involves a survey of 5,416 people. The data relates to the 2005 year.

    Collate all that data from police and courts and you will end up with reasonably accurate crime statistics.
33. Lew 33

    3 February 2009 at 1:09 pm

    DTB: If that indeed gives us `reasonably accrate crime statistics’, then you’ve just solved one of the fundamental problems in the discipline of criminology and the sociology of deviance!

    The trouble is that, while multiple sources and types of data certainly result in more information and can indeed result in better information, they only rarely result in more accurate information without very careful handling. Collating and correlating information from diverse sources is methodologically far from trivial, as you make it seem. In principle, what you describe is fine. Practice is somewhat different.

    L
34. Draco T Bastard 34

    3 February 2009 at 2:20 pm

    If that indeed gives us `reasonably accrate crime statistics’, then you’ve just solved one of the fundamental problems in the discipline of criminology and the sociology of deviance!

    I didn’t suggest the needed questions or what their weighting would be so, therefore, I can’t have done that.

    I didn’t imply that it would be easy either – that’s you reading into what’s written that which is not there. My implication was that the collation was needed for better understanding.

    Even then it doesn’t remove the fact that the data does exist, contrary to some claims, and it’s on this data that the police claims are being made. So far I haven’t seen anyone bring up anything to counter the police data or it’s interpretation.
35. Lew 35

    3 February 2009 at 2:27 pm

    DTB: Yeah, we mostly agree.

    So far I haven’t seen anyone bring up anything to counter the police data or it’s interpretation.

    Nor I. Except Rex, who doesn’t offer an alternative.

    L
36. Rex Widerstrom 36

    5 February 2009 at 9:26 pm

    Sorry for the huge delay in responding and I guess no one will read this now. Bugger.

    With respect Lew I’ve offered two alternatives, perhaps not very clearly though:

    1. A dedicated survey similar to the Household Labour Force Survey. The Household Crime Experience Survey, if you will… a weighted sample of households asked about their experience of crime.

    2. An add-on section in the Census (not great, but a measure against which Police stats could be compared over time or correlation or lack thereof). About the only advantage over the HCES idea I can see would be lower cost.

• Comments are now closed

• Log in
• Donate
• Contribute post

• Articles RSS