Written By:
Marty G - Date published:
9:34 pm, May 21st, 2009 - 50 comments
Categories: crime, employment, mt albert -
Tags:
Duncan Garner had to do a balance piece against David Shearer because he has been at the forefront of covering the implosion of Melissa Lee’s campaign. Garner says someone “deep within” Lee’s campaign pointed him to an interview where Shearer says keeping employment high, particularly among migrant communities is important because when people don’t have work it is more likely some will turn to crime. Except he didn’t put it that clearly.
Actually, he was very inarticulate, which is a strange trait for a man with his background to have.
As gotcha politics though, this fell pretty flat. Garner called it an attempt at “payback for Lee’s awful performance so far” (payback? they’ve got no-one to blame but Lee). Garner acknowledged what Shearer was trying to say is true, unemployment does lead to crime, but his “comments are confusing and might be taken the wrong way by some”.
Might! Some! Oh noes.
The important thing is that Shearer, unlike Melissa ‘crims from South Auckland can’t use off-ramps’ Lee, is actually right. Crime is a symptom of poverty. Unemployment is a prime cause of poverty. Check out the numbers from StatsNZ.
0.92 correlation. I’m actually amazed it’s that high. (It also has some pretty worrying implications for crime if unemployment reaches 8% during this recession).
Incidentally, I wanted to do a graph to test Lee’s motorways stop crime claim but I couldn’t even think of a remotely sensible way to that.
–Marty G
The server will be getting hardware changes this evening starting at 10pm NZDT.
The site will be off line for some hours.
Certainly not much of a “gotcha” when Shearer had already put out a press release on the subject that day:
http://www.scoop.co.nz/stories/PO0905/S00285.htm
Duncan Garner somehow forgot to mention that in his report.
Poor Nats. It’s OK to be desperate – but the trick is not to look desperate.
Was this an excerpt from a Monty Python episode?
Of course there is a link between unemployment and crime. Your calculated correlation shows this clearly.
I am not sure what Norman was on. This is unfortunately a familiar feeling. Shearer’s comments may not have been dynamic sound bite stuff but they were accurate and a proper representation of the issue.
Is this somehow balance? One person maligns whole geographical and ethnic sectors of our society, the other makes a perfectly appropriate comment based on decades of research and understanding.
The real difference is that I do not know one person who is proud to be unemployed. I know plenty of proud South Aucklanders and Pacificans.
Correlation does not mean causation. They could both be caused by some third (ungraphed) factor.
BTW thanks for at least allowing the possibility that some crime is not caused by unemployment and/or poverty.
You’re quite right, Anita, that correlation does not necessarily mean causation but I would be very interested to know what you think the third factor might be. I think its well understood that poverty increases the likelihood that a person will commit crimes.
You’ll note that the intercept is above 0. That is, even if unemployment were 0, crimes per person would be about 0.8, assuming the relationship holds.
I think though, that your particular angle is that domestic crime is not caused by unemployment. True, it’s not a sole cause but I would have thought that unemployment as a cause of poverty, depression, substance abuse etc is certainly a factor in a lot of domestic violence.
Note, I’m not saying that there isn’t domestic violence in wealthy families.
Hi Anita
It is a theoretical possibility but are you able to name one factor that could account for this (increased crime rate)?
I guess the abundance of time and the complete lack of financial resources along with the harming of one’s sense of worth are the real cause of increased crime amongst unemployed and not being unemployed per se but the three factors follow unemployment so strongly that for all intents and purposes perhaps we should treat the situation (unemployment) and the consequences (time, lack of finances and lack of self worth leading to increased crime) as the same?
I should try to write shorter sentences …
1) I am not saying that there is no causal relationship (in fact let me state for the record that I think there is), simply complaining about your implicit “there is correlation therefore there is causality” argument.
2) You have actually hypothesised unemployment –> poverty –> crime. I’m not sure the first pair are exactly correlated, so your argument appears to be:
I believe B–>C
I believe A–>B
I can show correlation between A and C
Therefore B–>C and A–>C
3) I am pretty sure that one could show a correlation between motorways and crime. For a start it seems that crime has fallen and motorways have increased (in length and probably in total number of offramps). So if we got the total length of motorways in the greater auckland metropolis (or number of off ramps) and the total crime stats and cherry picked the date range and adjusted the axes I reckon it would look convincing.
4) Are you arguing that poverty is the determining factor for the level of domestic violence? You would be way wrong.
Oops, attached to wrong comment, actually a reply to Marty G’s comment.
P.S. mickysavage, I should learn to writer shorter comments 🙂
“4) Are you arguing that poverty is the determining factor for the level of domestic violence?”
No. A factor among others that increases the odds of domestic violence. Like I specifically said because I feared you would choose to read words other than what I wrote.
3) Yeah but that only works when crime is going in one direction over the sample period.
2) I think it’s a bit strange you choose to frame unemployment leads to poverty, poverty leads to crime in such disparaging tones. It’s hardly controversial.
1) When there is strong correlation, a logical reason why there would be causation and no other apparent reason for the correlation Occam’s razor says we ought favour the notion there is correlation rather than choose to believe in mysterious unknown common causes or mere coincidence.
Don’t confuse causation with X being the sole cause of Y. In just means that a change in X will lead to a change in Y. It doesn’t exclude that Z (and A and B and…) is also a causative factor.
The thing is, if unemployment leads to poverty and poverty leads to crime why did crime not go up during the depression?
Other factors such as a sense of continuing community togetherness (when many are in the same boat) and or egalitarianism (generous provision) also play a part (a protection against crime).
This is why crime rates explode amongst minorities experiencing economic disadvantage – they feel a disconnect from the haves and no sense of community with them (this speaks to immigrant groups and is a separate issue though similar to Maori alienation after the land loss).
Which speaks to the need “close the gaps” – where there are gaps there is a multiplier at work.
I’m sure there are studies which look at this in more detail, and give a better handle on the causation/third factor issue (longitudinal studies maybe?). Anyone interested in looking into this?
I can’t stand how Garner frames arguments. I never have. One reason I don’t watch Tv3. Don’t have to put up with that smug twit. It doesn’t matter what he is reporting on or who he is reporting to. The guy is an ass.
yes.
Agreed. If there is one thing we can take from this story it is that Duncan Garner is mentally deficient, but that was quite clear already. I suppose it comes with being a journalist.
you’re using two different scales on the y axis.
Which disguises the fact that while unemployment dropped 50%, crime only dropped 10%.
Clearly unemployment is a possible factor but definitely not the major factor in all crime. Although Shearer seems to think it is with Polynesian and immigrant comunities.
Of course I’m using two different scales. Crimes per person is around 0.1. Whereas the unemployment rate is around 0.05. Crime varied about 16% for a change in unemployment of nearly 50%. Put them on the same scale and you don’t see the correlation.
You know what correlation means don’t you? It measures the likelihood that for a given change in X there is a given change in Y. Those changes don’t have to be of the same magnitude. For instance, if for every 1% taller a person is they are always 3% heavier, that is a correlation of 1, the size of the changes are different but the link between them is perfect. In the crime/unemployment correlation, the relationship is not perfect but it actually amazingly close for a pair of social indicators (anything over 0.2 is usually regarded as indicative of some, perhaps causative, link in social indicators)
I chose to show a line graph because people are more used to it but I’ll put in a link to the scatter graph for you too if you like.
I’m not sure where you get that interpretation of Shearer’s words. Care to quote him? Because in the video he says to Garner that unemployment leads to some people getting into crime. He doesn’t say all unemployed people become criminals or that unemployment is the only reason people become criminals.
Of course the axes are different, what’s being measured is different.
Of course the proportions are not the same, what’s being measured is different.
There is a correlation (and causal relationship) between ambient air temperature and tomato plant growth. Does it matter that the temperature at which tomato plant growth reaches 0cm/day is not 0º (on any scale) and that one is distance/time and the other is degrees something? Nope! 🙂
visual representation of data is as much a part of statistics as the maths. Any stats course would teach about taking care not to persent data in a misleading fashion.
it would be less misleading to convert both data sets to percentages. Then overall trends could be compared.
But to go back to the facts – unemployment dropped 50%, crime 10%.
It’s probably because there are different types of crime commited by different types of people.
Shearer has taken the view that Polyensians and immigrants are more susceptible to crime when unemployed. He didn’t produce any facts to back that up and I can see why some in those communities could take offense.
Shearer has taken the view that Polyensians and immigrants are more susceptible to crime when unemployed.
Bullshit he does. He is saying that the some communities are more susceptible to becoming unemployed, and that we should do something about that.
from the link:
He does not say what you claim he is saying. At all. Pretty shameless for someone that’s nitpicking about graph axis.
The time on their hands comment is pertinent*
Back in the 90’s* there was the advocacy of work for the dole (despite the cost of the programme people not on it found paid work more quickly) – so by the end of the decade people were advocating job club (which in its American origins was basically keeping a record and reporting in daily about it of the employers/businesses annoyed that day).
Training is much more effective – the problem is the cost of retraining (which is why the Jobs Summit idea of a 10th day in paid traiing was abandoned). Also existing training providers cannot just rapidly expand places (invest money) and then retrench back to lower levels when the economy recovers (they lose their investment and possibly their original business). Which is why government funding is vital.
PS* The only thing of value in being unemployed is having free time – which is why those who would operate a low minimum wage economy would also run work for the dole and job club (and if they could term limits to welfare) to maintain demand for work.
So he just happend to talk about migrants and Polynesians being unemployed and then straight after talks about unemployment leading to crime – along with the “idle hands” theme. Where have we heard that one before.
Obviously Shearer isn’t racist but he has strayed from his comfort zone and said something quite stupid. No motorway invloved but it’s much the same ill-informed and somewhat dangerous musing on ethnicity and crime.
He should be pinged for it just as much a Lee. But just as there were some on the Right who denied there was any issue with Lee’s musings so it’s the same on the Left with Shearer.
FunWith Spin
Have you read Shearer’s press release?
There is no comparison with Lee’s blurt.
If I had converted the figures to percentages we would have lost the information about the levels of crime and unemployment, which are also of interest.
It’s clear that the percentage changes are different. Anyone can see that and I did not argue otherwise.
Next you’ll be insisting that an axis always start at zero, even when one is looking at trends in a value that doesn’t approach zero.
“It’s clear that the percentage changes are different.’
you just happened to choose the scales for the two sets of data so the graphs appeard to be on top of each other. Which gives no visual indication of the percentage changes.
You keep avooiding the fact that while unemployment fell by 50% crime only fell by 10%. So the relationship bwteen the two is a little more complex.
But it is true that unemployment increases some types of crime in some population groups. Shearer made the mistake of claiming unemployment led to higher crime in immigrants and Polynesians. Maybe he has evidence for that.
“Shearer made the mistake of claiming unemployment led to higher crime in immigrants and Polynesians. Maybe he has evidence for that”
It is the other way around.
He said they are groups with increasing unemployment, and made the link with increasing unemployment and crime. There sure is evidence of that – unemployment and crime do have a clear link. that is what this post illustrates despite your best efforts to not see it because you do not understand what correlation is (for starters, you seem to insist that there must be a 1% change in one factor leading to a 1% change in the other factor for there to be a correlation. You know that if a 1% change in one factor always leads to a 10% change in another factor that there is still a very strong correlation? obviously not).
For someone making a fairly irrelevant hair-splitting point about numbers, you play fast and loose with words. And you get it wrong with both.
WTF is Norman on? Why is he jumping into this kind of crap story? Loads of Green people I know hate the way he is conducting himself.
Incidentally, if we’re on the subject of 3 News shoddy reporting, they told us the guy running off with 10 million bucks from Westpac is Korean. That’s false.
As this story has got vastly more attention and interest from the public than anything David Shearer might say, I reckon Melissa Lee might want to lodge a complaint!
Excellent post.
The reason you can’t think of a remotely sensible way to test the proposition that motorways “stop crime”—I think the phrase you’re looking for is “alter the geographic distribution of crime”; that’s what Lee was talking about—is that it’s a *much* more subtle claim than positing an unemployment-crime link. It’s not _quite_ as implausible as many make it out to be: motorways alter the goegraphic distribution of almost all other economic activity; there’s no good reason property crime (burglary/home-invasion, theft) should be any different. Whether or not it actually trends in the same direction Lee said it does is another matter entirely.
(Note: I stipulate that Lee’s comment was stupid. That doesn’t make the question any less interesting from a more academic point of view.)
The difficulty is the implicit assumpon that there is fixed population of people called “criminals” who do all the crime. (It’s not as tautological as it sounds. How many crimes must you commit before you are a “criminal”? How long must you spend as a completely law-abiding citizen before you are no longer a “criminal” for these statistical purposes?) Correcting for this is not straightforward, but there have been—mainly inconclusive, mainly anecdotal—attempts to do it.
There’s no simple test for the proposition: you need to borrow methods from criminology, epidemiology, economics, and probably even fluid dynamics (to model traffic flow of “criminals”). Even then, I doubt you’d find conclusive evidence either for the proposition or against it.
CraFarms the experts in dirty dairying show a link between business and crime.
The Serious Fraud Office was established because of white collar crime.
Is there any disparity in rates of spouse abuse between middle class Pakeha families and others (to the import of economic stress as a cause there might be some)?
My point is that one can use correlation to negatively profile people based on employment status (or low wage levels) – from there one profiles the whole working class (higher skilled workers excepted) as a group associated with crime risk.
Most poor people are as law-abiding as the rest of the community.
That said, crime does rise when there is economic distress – finance companies fleece investors, people of the middle class put their hand in the till, attempts at fraud by well presented conmen increase. Businessmen try and lay off workers after they have diverted money for redundancy payments elsewhere … . Loan sharks move in, even on some of the middle class.
It is true though as the FBI BSU showed with an international survey back in the 80’s
1. there was no link between crime and adult porn.
2. there was a link between crime and economic disparity disadvantage experienced by an ethnic minority (little crime in mono-cultural egalitarian community and presumably multi-cultural egalitarian community where this was achieved).
Which is why when you have ethnic groups with higher unemployment, it is unwise to hold benefits down too low – and closing the gaps is wise policy.
“Incidentally, I wanted to do a graph to test Lee’s motorways stop crime claim but I couldn’t even think of a remotely sensible way to that.”
Incidentally, there is research on the patterns of crime, and the relationship with roads and other geographic and architectural features. Have a look at this interesting read:
http://www.newscientist.com/article/mg19826541.000-sin-cities-the-geometry-of-crime.html
Of course these things are a lot more complicated then what was suggested by a certain MP, and don’t involve prejudice about where people are from.
If this is the best that the Nats can come up with then Labour should be fine. “Sources deep in the National Party” seem to be saying that unemployment isn’t a factor in crime statistics, when it is. Dumb stuff.
Desperation stuff innit.
A great post.
Shearer was spot on. Can’t see how anything he said was clumsy. A bit too honest maybe but not clumsy. Even my ten year old son could see that unemployment may contribute to crime doesn’t mean that by becoming unemployed makes you a criminal
Garner did him a huge favour by mentioning the tip off from National- desperate was to word coming to mind
Garner did a beat up on Shearer last night too when the one excerpt screened of Shearer speaking at Auckland uni was the one answer he gave poorly. Admittedly his performance wasn’t stellar, but Shearer did ok (Norman did slightly better but he was still pretty limp), certainly not as poorly as Garner’s piece implied.
Once again Garner wants to be the man to manipulate New Zealand politics at the expense of his viewers.
WTF is Norman on? Why is he jumping into this kind of crap story? Loads of Green people I know hate the way he is conducting himself
Tthat sentiment is expressed a lot on this blog,
However I know a huge amout of Greens and have never once heard it expressed
and on the contary, are very supportive of Rusell I gueess iI’ll find out for sure at the conference the weekend after next.
I think he’s actually performing very well by far the most relaxed, confident and articulate candidate but he does have something new in Green MPs, a willingness to get into a dirty attack, like this one. I don’t know why he would choose to be part of a pile on against Shearer when he must surely agree with what Shearer was saying because it’s true.
Why is questioning what Shearer said a “dirty” attack – consummate politicians would welcome the opportunity to clarify what they said – and if they did not want to highlight the issue and talk further about it, why would they have raised it?
Its actually providing Shearer with an opportunity to become part of some interplay, so the centre and left can dominate the by-election agenda.
I know some very senior Greens who have expressed this sentiment OOB. But you are right, the proof will be in the pudding. I would advise you to also take note of who turns up and who doesn’t.
I love Garner’s reference to his source as “deep within” National.
Actually, the original source is an interview on public television in Auckland. What Garner’s really saying is “Here at 3 News we can’t even be bothered to watch the candidates on TV, we just wait to be told about it by the party spin doctors.”
What a joke.
(if he had any interest in real journalism rather than waiting for his stories to be delivered on a plate, he’d follow up this issue with Ravi Musuku. Now that would be worth seeing).
Balance my fucken arse
1) It was not a lead story
2) He specifically says Pacific Islanders, Lee did not, only Cunliffe said that
3) He says Lee had an awful performance – all he says about Shearer it is confusing
4) While Lee stumbled once in five minutes, and Garner himself was unable to correct her it lead the news, he ignores Shearers almost constant waffling
Where are the stats in the graph from?
Note the post says they are from StatsNZ.
Much has been said about Ms Lee’s communication skills.
How about this deep perspective from Mr Shearer:
If anybody can successfully interpret what that means, then I think they deserve a medal.
Mr Shearer seems to be making a habit of waffle and bland, nonsensical replies.
So let’s make him Prime Minister!
Or at the very least an MP ……. surely waffle and bland non-sensical replies is a prime requirement for question time and most addresses within the house.
A quick update: despite the best efforts of the wingnut bloggers, this Shearer non-story has been ignored by all media except Duncan “Dubya liked my suit” Garner. No surprises there.
And it’s about to be wiped clean off the radar by Melissa Lee’s extraordinary confession this morning that she “hopes to come second.”
The media (and the Greens, and ACT, and the Bill and Ben Party …) are going to be all over that one.
yeah i heard that interview.
what kind of complete fool says in a campaign they hope to come second? it’s mind blowing.
and how profoundly incompetent are National’s minders and handlers that they keep letting a candidate behave like this? the depth of their stupidity is really quite stunning. where did they learn their craft – the John Key School of Allocution? The GW Bush School of Higher Thought?
i can’t believe Key continues to employ these munters – they reflect very poorly on his judgement.
The story is here:
http://www.stuff.co.nz/national/politics/2432518/Lee-hoping-for-second-place
My computer illiteracy prevents me from embedding it. Cut ‘n’ paste, boys and girls.
Oh look, it did it for me. Cool.
And what comes next? Yep, the “clarification” … a.k.a. more digging.
http://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=10573871
Oh dear.