Is disputing data dirty politics?

Jarrod Gilbert is tweeting that he has an OIA showing I asked for information to attack him. This is not true. I asked for information to dispute his data, and he seems unable to see the difference. This is the legacy of the Hager book, that people now call anything they disagree with as Dirty Politics. In fact the record will show I have never attacked Dri Gilbert or called him names, while he in fact has called me a range of names of which quisling is one of the politer.

However I can understand he is upset becaise at the end of the day he was right in the assertion he was making, and I was wrong to be dismissive of it. I get things wrong sometimes.

Anyway I thought it would be useful to detail this rather sad saga from the beginning. It started when Dr Gilbert said Anne Tolley’s speech on gangs was incorrect, and he said he would eat a suitcase of carrots if she is correct.

In my experience Ministers tend to be pretty careful about using correct figures, and the nature of Dr Gilbert’s promise to eat carrots attracted my attention, so I had a look at the data under dispute which was that:

4,000 known gang members in New Zealandar responsible for 34 percent of class A & B drug offences; 36 percent of kidnapping and addiction offences; 25 percent of aggravated robbery/robbery offences; 26 percent of grievous assault offences..

Why I thought these figures were credible as that the number of Class A & B drug offences, kidnappings, robberies and grievous assaults is quite low. I was focused on the number of crimes side, not the number of gang members side.

I should point out that no one suggested to me to write my blog post. I didn’t talk to anyone before I wrote it. I saw Dr Gilbert’s post, and thought that the number of crimes in those categories made it very possible gang members were responsible for the proportions quoted.

My blog post here said:

Aggravated Robberies and Robberies

There were 2,032 robberies (both types) last year. 25% would be 508. That seems a credible number for 1,620 to 4,000 gang members to do.

Kidnappings and Abductions

There were 198 kidnappings and abductions last year. 36% would be 71. That seems a credible number for 1,620 to 4,000 gang members to do.

Grievous Assaults

If you add up the 17 assault categories that mention GBH, there were 500 offences last year.  26% would be 130. That seems a credible number for 1,620 to 4,000 gang members to do.

Class A and B drug offences

There were 16,070 illicit drug offences in 2013. They are broken up into specific drugs and it would take a long time to do an exact count. But a previous Stats report is that less than 10% are Class A and B. So a fair assumption is 1,607 Class A and B drug offences last year. 34% would be 546. That seems a credible number for 1,620 to 4,000 gang members to do.

So this was me responding to a post on data, with data. How this is dirty politics I don’t know. After I did the blog post, a staffer from the Minister’s office e-mailed me and said:

Hi David – I have a breakdown which backs up the stats. Let me know if you want them.

Again, nothing extraordinary. In fact sensible pro-active work. I understand the data was also provided to Radio NZ. I did not ask for the data, it was offered to me.

I updated my blog saying:

I have been sent the actual stats the Minister was relying on, which are for the first quarter of 2014. They are:

• Class A/B drug offences total 218 out of 649

• Kidnapping and abduction 16 out of 44

• Aggravated robbery/robbery 72 out of 284

• Grievous assault 130 out of 506

So I made it very clear I had been sent the stats, and implicitly of course from her office.  Nothing secret.

I commented:

I look forward to the Herald covering the Jarrod Gilbert eating his carrots.

Jarrod then responded with this post:

What the 28 percent prison number represents is gang members as well as gang associates in prison. This makes a massive difference. A few years back The Police Association said gangs and associates numbered 60,000. Associates are difficult. Am I an associate? Is a guy with a brother in the gang an associate? Associates are an arbitrary measure that can capture people not connected to a gang in any meaningful way. Therefore, if you are talking about gangs as ‘membership’ numbers or gangs as ‘members and associates’, then you are talking about very different things. That’s a 4,000 to 60,000 difference. While I think the latter should be avoided, you need to at the very least use one definition or the other and remain consistent. Apples with apples, as they say.

My focus up until now had been on that the statistics around number of offences were correct. I had not focused on the issue of whether those responsible were gang members or merely gang associates.

It seemed unlikely to me that government agencies would be using different definitions for different stats. As it happened they were, and Dr Gilbert was quite correct. I had no way of knowing so I did what was responsible and asked the Minister’s office:

Any useful data to respond to this with?

Note the use of the word data. That is and was my focus, not attacking Dr Gilbert. I regret he feels I was. I reject the idea though that me asking a Minister’s office for data is in anyway inappropriate or dirty.

I don’t have a record of an e-mail in response, but think there was a phone call which said something along the lines of we trust the data from the Police.

I then did this response in which I said:

Now sadly Dr Gilbert won’t accept he was wrong, but is now trying to argue that there is a difference between gang associates and gang members. So he is not at all disputing that gangs are responsible for 25% to 36% of kidnappings, robberies, grievous assaults and serious drug offences. He is now just saying that the crime figures may include associates, not just gang members.

Is Dr Gilbert Saying the Corrections Department is lying when it says 28% of the prison population are gang members? They supplied the data, and I see no reason why they would make it up.

So it is a technical argument over definitions. I don’t care what you call them.

The Police Association are not an official source. The Police are. They say there are 4,000 gang members. I don’t know if they includes associates in that. I presume Police and Corrections are using the same definition.

So I’m glad Dr Gilbert is not disputing that gangs are responsible for 25% to 36% of those four serious crime categories. He is disputing whether they are all done by members, or associates. I doubt that matters to the victims of the grievous assaults, robberies and kidnappings.

Now I’m not proud of this response. I focused on the fact that I was right in saying “gangs” (if you include associates) were responsible for the proportions of crimes. I said that I wasn’t very interested in the difference between a member and an associate. This partly reflected my views that most serious crime in NZ is committed by not that many people and I took the difference between membership and an associate to be matter of degree. In hindsight I should have dealt more seriously with the point Dr Gilbert was making, which was that there is a difference between 4,000 gang members and 60,000 associates. I was trying to win a debating point rather than being fair (I blame the OU Debating Society). I do regret that.

I also honestly didn’t think the Police and Corrections would be using different definitions, so I tended to discount the probability of this.

I also made a mistake in not understanding the wide difference between an associate and a member. I tended to see an associate as someone on their way to being a member, or someone involved a gang if not a formal member. Again I should have recognised that this is an area Dr Gilbert is an expert in, and given more due to that.

Dr Gilbert responded to my blog post with his one here.

Then on 24 September the Minister’s office e-mailed me saying:

Police have released info which says their stats also include associates and family (Which they didn’t tell us at the time). Jarrod will no doubt blog.

My reply was:

Ouch.

What really hit me was that the Police figures included family. As someone who puts a focus on individual responsibility, I was a bit stunned that merely having a family member in a gang would get you classified as a gang associate. That made me realise I was wrong in not giving credence to what Jarrod had said. I acknowledge that even without that info, I should have treated his argument better, as I will in future.

I then did a blog post saying Jarrod was right, and I apologised to him.

While I think my second blog post on the topic was not something I’m proud of, I reject that I was seeking to damage or attack Jarrod. My focus was on the data. It should be apparent from my blog posts that I blog on data all the time. What was meant to be a fairly friendly exchange on offending stats, become more than that. I accept my responsibility for that, but it was never my intent to have Jarrod feel slighted. I would point out again that I have never resorted to name calling, and while I appreciate that some of the comments about Jarrod by some of the commenters here were not pleasant, the same goes for comments on Jarrod’s blog about me.

Again I repeat my apology to Jarrod, but at the same time I reject his assertion that I was conspiring to attack him.My aim was to attack his data and arguments, not him.  Regardless I will in future tread more carefully.