Kiwiblog

Len Brown writes in the Herald:

The new Auckland Council will have a lot of work to do to help create a safer environment – police statistics show that Auckland City has more reported crime per 100,000 residents than any other police district in the country, and Counties/Manukau the fourth highest rate.

Now I always get interested when anyone starts talking about the overall crime rate for it is almost a meaningless statistic, as it treats all crime as equal. A brutal homicide is the same as jaywalking (well that is not an offence, but imagine the most minor criminal offence there is).

I always pay most attention to the violent crime rate, and sexual crime rate, as they are the one people worry about most in terms of safety.

So Len Brown is right that in terms of overall crime, Auckland Police District has a higher rate that Counties/Manukau. There were 55,898 and 54,409 reported crimes respectively, which is a rate per 10,000 population of 1328 and 1094.

However if we look at violent crimes, the respective rates are 128 for Auckland and 195 for CM – CM is 52% higher.

Much the same for sexual offences – the rates are 7.8 and 9.8 – CM is 26% higher.

The category that makes up most of the overall crime stats is dishonesty, and the primary offence within that category is theft.

So remember whenever anyone talks about the overall crime rate, ask about the violent and sexual crime rate also. Otherwise a 2% fall in thefts may be masking a 50% increase in violent assaults and rapes.

NewZblog tries very lamely to blame National for the high migration in October 2008, arguing it should have reduced because National was ahead in the polls.

Now I think one doesn’t even have to try and respond to the lame attempt to blame National for migration figures that are for the month before the election. But it does raise the issue, when should you start holding a new Government accountable? Is it the Nov 2008 figures? Dec 2008?

A tend to to think 9 – 12 months is the usual lead in time you look for, to expect a Government to be “accountable” for various statistics. Accountable doesn’t mean they have a hell of a lot of control over them, but means that the country is now operating under the new policy settings.

Very few people or companies change behaviour just because a Government changes. Employers don’t go out the day after the election and start advertising new jobs they were holding off on. Expats don’t suddenly pack up their homes and get the next flight home.

What they do respond to, is policy changes. And generally it takes nine or more months to pass a law, and you have to wait until the budget in May next year for changes in government spending and taxation – and those changes may be spread over the next year or so also.

Another key thing to look for is what was the trend up until the change of Government, and did the trend then change? For example In 1990 unemployment was skyrocketing andat the end of the year National came into power. Unemployment kept increasing for nine months, and then it turned about and declined for the next five years. I blame the increase from Dec 1990 to Sep 1991 on the mess inherited from the former Government.

Some stats will not turn around for years and years, if at all. If you accept part of the solution to outwards migration is wage levels, and you also accept that the solution to wage levels is increased productivity growth, well it is a very long-term project to increase productivity growth. So at first what you might look out for is not a reversal in the trend, but even a slowing.

The other thing to look for is not just measuring actual vs actual, but actual vs projected. For example National will have a bigger deficit in next year’s budget than Cullen will have this year. The correct comparison to look at will be the size of the deficit compared to the projected deficit from the DEFU next month.

1. Busted Blonde has a post on domestic violence and how she spent several years in an abusive relationship. Go read it, and make sure you show it to any friends who need to read it.
2. Adam Smith blogs on the Giles cartoons. As a child I loved Giles cartoons, and every year could not wait for the annual. Grandma was my favourite character. For those who never saw them, you missed out on classics.
3. Matthew Hooton makes the case for an Upper House.
4. Steven Price reviews two decisions by the Advertising Standards Authority.
5. Matt Nolan wants some better statistics from the Government.

Clinton Smith at The Standard responds to my post on the 11% increase in violent crime with what is really a nasty smearing post. I avoid responding to most of his stuff, as it reflects more on him than me. But as he basically calls me pro-rape, there is a limit. I’ll go through his post in turn:

David Farrar really is a disgusting person at times.

Yes how disgusting it is to point out violent crime increased 11% last year.

He writes that we shouldn’t look at overall crime rates (although he made a big fuss when the overall recorded crime rate went up because of changes in recording practice in 2006)

Smith is wrong. I have been making the point about violent crime, not overall crime rates, for years and years. To be precise:

1. 3 March 2008
2. 2 October 2007
3. 2 April 2007
4. 2 October 2006
5. 4 April 2006
6. 6 July 2004

Pretty much all of these said the same thing – the overall crime rate is not a very useful measure. I welcome people wanting to debate it is – but funnily enough few people ever engage that argument on the merits.

Instead we should just look at violent crime, after all “having 1,000 less [sic] cannabis crimes and 500 more rapes is a net decrease in crime, but would be a worrying trend”. He then goes on to act all shocked over the recorded violent crime stats. First, he’s trying to make you think there are more rapes

Yes that should be fewer. But I doubt anyone would read what I said as stating there were 500 more rapes. It was obviously an example given the round numbers used.

in fact, the number is down and he’s hardly one to fret about rape when he’s good mates with a pornographer).

Now I can not even describe how offensive Smith is with that statement. He basically says I don’t care about rape or rape victims. I can think of few things more defamatory let alone downright nasty to say about somebody. Next time Smith complains about the standards of a right wing blog, people should remember he considers accusing people of being unconcerned about rape as a legitimate tactic.

Smith deosn’t even try to debate the merits of violent crime stats vs total crime stats. He just decides to smear me. As I say that reflects far more on him than me.

Secondly, this professional statistician is deliberately trying to mislead you regarding the change in the crime stats. He knows that the explanation from the professionals is reporting of domestic crime is up but he hopes you don’t.

I am aware the Police provide excuses for why violent crime has gone up 11%, saying it is all due to increased reporting of domestic violence. However like all government agencies they of course try to put a favourable spin on the statistics. Does Clinton attack Dr Cullen when he disagrees with Treasury advice? More to the point when you look at specific categories such as:

• Homicides from 96 to 112 – 16.7% increase
• Wounding with Intent from 635 to 697 – 9.8% increase
• Injuring with Intent from 996 to 1,180 – 18.5% increase
• Assault with a weapon from 2,738 to 3,007 – 9.8% increase
• Assaults on Police from 367 to 439 – 19.6% increase

You realise the explanation is not that credible.

I submit that these groups of crimes are not ones that would be greatly affected by increased reporting of domestic violence. Wounding with intent is the most serious assault possible, short of attempted murder. Assaults with a weapon don’t tend to be the exclusive domain of domestic violence, let alone assaults on police. And homicides speak for themselves.

Crime is down because poverty and the conditions that breed crime have been reduced. At the same time, reporting of crimes is up. Thefts, car thefts, and burglaries (together, the largest group of crimes) have halved in the last ten years. Homicides are down.

1. 98/99 – 94
2. 99/00 – 104
3. 00/01 – 97
4. 01/02 – 112
5. 02/03 – 108
6. 03/04 – 100
7. 04/05 – 80
8. 05/06 – 105
9. 06/07 – 96
10. 07/08 – 112

That is the homicides for the last ten financial years. The latest year is the highest equal of the last ten. Note when I point out an error Clinton makes I don’t feel the need to call him disgusting and other names.

In fact, all classes of crime are down except two. Recorded violent crime is up because reporting is up. Recorded property damage offences have climbed in the last three years thanks to the moral panic over tagging. But those rises are attributed (not by me, by the Police) to higher reporting.

Recorded violent crime has increased almost 50% since 1999. Clinton attacks and smears anyone who points this out and dares to disagree with the assertion it is all because of increased reporting of domestioc violence.

Farrar needs to misrepresent the statistics to create the impression of a wave of crime where there is none. It’s all part of National’s ‘New Zealand sucks’ campaign. It disrespects his readers and the victims of crime. Disgusting.

I regard a 11% increase in violent crime in just one year as very significant. The fact almost every category of violent crime has had a significant increase makes it highly unlikely it is just a matter of increased reporting of domestic violence. I think one can debate these issues without calling peoeel disgusting and suggesting they don’t care about rape. Sadly Smith is unable to do so, he is so caught up in his hate towards others.

But in the end, I think his posts reflect far worse on himself than his intended targets. It may appeal to the highly partisan, but it tends to be counter productive with most people.

The Police released the latest crime statistics this week. I tend to focus on violent and sexual crime for two reasons.  Firstly because it tends to be the most serious and damaging offending. The overall crime rate holds little meaning as very minor offences such as cannabis possession are given the same weight as a serious assault or worse.

The second reason is that violent crimes especially are the most likely to be reported, and less susceptible to changes in policing practice.  Changes in the level of “victimless” crimes are almost totally related to the amount of targeting the Police put into those areas.

There are four distinct trend periods in the graph above.  Firstly violent crime did rise from 1990 to 1994. They were tough times for a number of reasons. But from 1994 to 1999 the level of violent crime basically stayed cnstant, and even declined in terms of the population rate.

Then from 2000 to 2004 there was a modest level of growth in the violent crime level and rate. And from 2004 to 2007 there has been basically massive increases in violent crime levels.  The number of record violent crimes is 44% higher than in 1999, and the rate of violent crime is 32% higher.

Now the Police have said the latest increase is mainly domestic violence, and may be as a result of publicity encouraging greater reporting. I’m somewhat sceptical that an increase in recorded violent crime of 6,250 is solely due to increased reporting than an increased actual level of violent crime.

So let us look at the sub-categories of violent crime and how each has grown in the last year:

1. Homicides dropped 10 (10%) from 98 to 88. Less full moons last year I guess
2. Kidnappings dropped 6 (2%) from 263 to 257
3. Robberies dropped 191 (7%) from 2801 to 2610
4. Grievous Assaults increased 715 (17%) from 4,116 to 4,831
5. Serious Assaults increased 2,894 (16%) from 17,729 to 20,623
6. Minor Assaults increased 1,321 (10%) from 12,700 to 14,021
7. Intimidations increased 1,410 (11%) from 12,567 to 13,977
8. Group Assemblies increased 119 (26%) from 457 to 576
9. Wounding with Intent increased 168 (30%) from 554 to 722
10. Injuring with Intent increased 182 (20%) from 895 to 1,077
11. Assaults on Police increased 79 (23%) from 342 to 421

Now remember this is just the increase in the last calendar year – it is not the increase since 1999.

It is worrying that the greatest percentage increases are in the more serious assault categories. And they are the ones most likely to not have been under-reported in the past. Minor assaults may have been not reported as often as they should, but if someone has actually stabbed you with a knife, that tends to get reported I suspect.