Islands in a sea of blue

Written By: - Date published: 1:59 pm, November 28th, 2011 - 25 comments
Categories: labour - Tags:

There can be no clearer message about what type of campaigning pays dividends for Labour than the images from Keith Ng’s party vote page. [Upate: I misunderstood this graphic at first – it is of electorates where Labour’s Party vote exceeded National’s.  r0b]  The islands of red are:-

  • Mt Albert
  • Mt Roskill
  • The three M’s – Mangere, Manakau East, and Manurewa
  • Rongatai
  • Dunedin North

Obviously, I’ll have to wait until December 10 and the special vote values before I get into the numbers. But it is striking.

With the exception of the three M’s, the other elections are seats that should have moved demographically over the last decades, but don’t. Lets look at the four electorates that are not the 3M’s. They are the electorates that concentrate almost all of their efforts on local campaigning organisation. These days they target the party vote rather than the electorate vote.

Retention of the support of the people that vote for your party is probably the most effective campaign tool that the Labour party has. It allows you to survive the hard times in the political swings. These are also the 4 electorates that before this election had the most canvassing information about their constituents. I’d imagine that there are others now – I know of a couple.

Mt Roskill has had the emphasis that way ever since Phil wound up in the wilderness in early 90’s under a succession of volunteers. Mt Albert had that long before I started working there in the early 90’s. Rongatai’s organisation has been strong for Annette King with the redoubtable Lloyd Faulk making sure everything worked. Dunedin North has been working for Pete Hodgson with a strong electorate team playing the numbers.

I don’t think that there is much effort wasted in waving bits of Corflute at passing motorists in any of these electorates. They concentrate on running canvassing phone banks and door knocking all the time including in the non-election years. Canvassing data is carefully conserved, punched including the date, and don’t damn well lose lose the data when volunteers leave.

They have databases of canvass data for voters that goes back decades because they keep updating it and they make sure there is sufficient continuity in the local party organisation to use it. Of course it helps that there have been a strong continuity of MP as well. But that is the case with many electorates that have gone below the blue tide. They have canvassing data on a significant proportion of their electorates and they use it intelligently.

And before anyone asks, Mt Albert is not demographically a “natural Labour seat” and haven’t been for quite some time. Probably since I was a kid there. It has gentrified long ago, even before the infill housing filled it in the late 90’s. Mt Roskill isn’t particularly either. From what I know of the Dunedin North and Rongatai electorates they have been gradually changing away from “natural Labour” electorates for a while.

In my opinion, the current database is a flawed incomplete system that hasn’t really bothered to be designed with the best of Labour’s expertise. I was wincing on election day looking at exactly the same frigging mistakes I made seven years ago when building teh first web based systems for the party. The targeting systems absolutely suck, the pages of targets were designed by someone who was more interested in conserving trees than effective results, and the voting intentions information that can be collected was designed by someone who has no understanding of how to do micro segmentation marketing. But it does collect information and hold it outside of the vagaries of volunteers lives.

It should be the kernel* for collecting and retaining information and ensuring continuity in electorate information that the four electorates holding against the blue tide got to decades ago. It mightn’t be glamorous or fast or particularly soothing for the egocentric political types. But it is the technique that works for the party over the long term.

I’d have put this observation through the party organisation in their review. However over the years I have discovered that it really isn’t worth the effort of making observations about operations. It tends to fall into a large black hole of ignored experience. The party channels tend to be far too sclerotic to pass information to electorates. Hopefully that should be on Moira Coatsworth’s review of the party organisation this year as to why valuable experience doesn’t pass crossways through the party.

I know that some in the Labour party don’t want to hear it (because I have ineffectually been yelling it for more than a decade) – but I told you so.

* So long as I don’t have to use it – it is primitive. 

25 comments on “Islands in a sea of blue ”

  • Comments are now closed

  1. mickysavage 1
    28 November 2011 at 2:45 pm

    In South Auckland there was a significant swing TO Labour. Throughout the rest of Auckland the result was not so bad with an overall swing away of less than 4% compared to the overall average of 7.1%. Christchurch and Dunedin swung heavily away. It is noticeable that the swing in Labour held seats with better organisations and in Waitakere where there was a good campaign were far better than in the National held seats.

    I agree entirely about the need for effective organisations but the problem is that voluntary organisation memberships of whatever sort are declining. Labour has had a far more effective on the ground presence than any other political organisation for quite a while but where every other area is running against you every volunteer is vital.

    • lprent 1.1
      28 November 2011 at 3:11 pm

      I had a wee discussion with Anthony in Dunedin North earlier about this. His experience is like mine. The bulk of the work can and usually is being done by quite a small active crew working slowly through canvassing over a period of time between the election cycles. That is what happens in Mt Albert, Mt Roskill, and these days in Auckland Central. And the data has to get stored centrally so that volunteers disappearing doesn’t mean that the hard won data dies.

      When you need the bodies is up towards the election. Ideally you have a few people who have run a campaign before who you can alternate rotate around elections whilst training new people up.

      But what is really required these days is some kind of online forums that cross talk can go on between electorates. We don’t have anything like it (and no I don’t have time to do it).

      The biggest shortage at present is in how to target with limited numbers of volunteers so the absolute most effective use is made of their time. I’m afraid I have been appalled over the years at how little effort has been made to maximise that resource. But that is probably the production/operations background coming out in me. The techniques prevalent in the NZLP electorates seem to have largely come from when the party was hundreds of thousands of people rather than when the number of active volunteers reduced. The Mt Albert 2009 by-election was a classic 60’s effort. But it only achieved pretty much the same effect as a smarter campaign in 2008 and 2011 with a fraction of the people.

      There are some pretty simple rules to figure out how to statistically analyse an electorate that just require enough canvassing data to feed them. You need to sample and the target for effect. We’ve been doing in Mt Albert and quite a few other electorates since 2005 because the classic red dot system wasn’t working.

      Problem is that the frigging red-dot system is still the only operational technique in e-Trac.

  2. Trix 2
    28 November 2011 at 2:52 pm

    While I understand this site being all about Labour, to say the seats represent a “sea of blue” based on the party vote is a little misleading. In many of the urban seats, Labour + Green has a sizeable majority, so could therefore be showing up as a lovely yellow colour.

    Remember this is MMP, and about proportionality, not what colour shows up for party votes in each electorate. Local candidate votes are another thing.

    Having said that, it’d be fun to see electorate maps based on additive colour coding, maybe simulating NZF’s party vote (over the 5% threshold) using saturation effects….

    • Jimmy 2.1
      28 November 2011 at 3:00 pm

      Yes I noted this too. One of the biggest shortfalls of New Zealand politics is that there is a strong hangover in mentality of a two horse race.

      National has no real opposition for right wing voters, hence the “overwhelming” support compared to left support for Labour.

      I would like to see a move away from having a government coalition where parties buy each others support and instead have each piece of legislation debated on it’s merits and voters truly represented by their MPs.

    • lprent 2.2
      28 November 2011 at 3:14 pm

      It isn’t a Labour site – it is left. I just happen to be a Labour activist, so that is what I wrote about.

      Oh I’d agree about the RedGreen thing. But I was really trying to put a burr up Labours collective arse using a very convenient map to illustrate a point.

      However I’d say that the same thing applies to the Greens as well. From my observations of their electorate techniques they are quite unsophisticated. That shows in the fluctuations of the green vote in various electorates as their volunteers change, and as Labour activists decide to include Green supporters in their targeting for election day.

  3. Carol 3
    28 November 2011 at 3:02 pm

    Acdtually this site isn’t all about Labour. it’s a left wing site and many of us voted for left parties other than Labour.

    Your point is more relevant to this particular post which is about Labour and I agree with your main point. I voted Cunliffe plus Green in New Lynn, as I have done before. National got a bigger party vote than Labour, but the combined Labour-Green vote was bigger than National’s.

    • lprent 3.1
      28 November 2011 at 3:17 pm

      But Labour is part of the left isn’t it?

      Just happens to be the bit that I know the best. So that is what I write about.

      • Carol 3.1.1
        28 November 2011 at 3:32 pm

        My comment was a response to Trix @2.52pm, but I somehow ended up in the wrong lane.

        But, yes, it could equally apply to the Greens.

  4. Tiger Mountain 4
    28 November 2011 at 3:31 pm

    The Red/Blue imaging might imply that people with four legged partners have taken over the country. But even in some of the worst offending areas such as Northland where bent ex copper Mike Sabin got 16,000 odd electorate votes, Labour and Green combined got around 11,000. If you add in the non national Northland equivalent Te Tai Tokerau votes, and I realise you can’t and shouldn’t, Sabin would not be close to heading for Wellington.

    So MMP friendly maps are more useful overall, but the map above does make a certain gruesome point.We are a post colonial nation divided and always have been, but the huge non vote, non enrollment and ‘no party badges please’ at Occupy actions show something different has to be done soon.

  5. Uroskin 5
    28 November 2011 at 3:51 pm

    Why is Palmerston N missing on the maps? Shouldn’t that be a little red dot in the deep blue sea too?

    • lprent 5.1
      28 November 2011 at 3:55 pm

      It is a map of the Party vote… (Because in the end who really cares about the electorate vote apart from the various candidates).

      National Party 13,209
      Labour Party 10,024

      and for the sake of the greens…

      Green Party 3,023

      So it’d have been mostly blue…

  6. tc 6
    28 November 2011 at 6:56 pm

    After a recent session for all us residents in waikato west coast with a geologist from a nice touchy feely Chinese mining company the blue areas represent the areas up for mining……..ah the brighter future.

    It was great to find out that it’ll be so clean you could live right next to it and well all get our roads upgraded yet no one will be monitoring the effects on local waterways, oh well it must be good because national say so eh.

  7. bob arctor 7
    28 November 2011 at 9:08 pm

    It could have been worse, Mt Albert, Mt Roskill, Rongatai, Dunedin North seemed reasonably close between Labour / National

    I guess there were quite a few that were National which wouldn’t have taken much of a swing to be Labour (Mana for example looked pretty clsoe to have a higher party vote for Labour.)

    Would be interested to see it the graph as a shade of blue and a shade of red. (i.e Mangere Bright Red, Rongatai a light pink)

    It looks like Labour still got the majority of the party vote in all the Maori seats though

  8. Waldo 8
    28 November 2011 at 9:20 pm

    In Rongotai Labour’s party vote lead has been cut from 5,000ish to about 160 (pre-specials), yet it is quite possible Annette’s lead in the candidate vote will increase from her 9,000 advantage in 2008. In the party vote National has picked up a couple of points as have NZF, the Greens have got an extra 5 or 6 points. Norman has also gained a little in the candidate vote. I am interested in your views on this electorate LPRENT.

    What do you think was the cause of this result? Why did the party vote shrink so much while the electorate vote got stronger. Did Annette’s higher profile as Deputy Leader contribute to a higher candidate vote? Why does an electorate with a relatively high household income (76,000) still vote slightly Labour (and leans heavily left when you factor in the Greens) while Ohariu with similar demographics favours National in the party vote? Is this attributable to King as an old fashioned excellent local MP?

    • lprent 8.1
      28 November 2011 at 10:23 pm

      Good local MP and electorate party with lots of canvassing data – bloody hard combination to shift. Just look at Wigram and bloody Anderton. Or Dunne and Ohariu. If the MP’s party is viable, then they can spend time targeting on party vote because that will feed straight into electorate votes as well.

      I haven’t looked closely at the numbers except for the overall trend and the electorates I’m involved in – it is mildly pointless until December 10 because with 10% of the votes in specials it is comparing oranges with apples.

    • SHG 8.2
      29 November 2011 at 12:10 am

      Got any examples of her campaign material?

      I was interested to see that by November Cosgrove was campaigning without a single mention of Labour in his material. Not on his website, not on his flyers. He had given up on the party vote and association with Labour entirely and was campaigning solely on his own name.

  9. Craig Glen Eden 9
    29 November 2011 at 8:46 am

    Sorry Guys I disagree, having stood for Labour in a Nat seat and all most taking it. The key issue imo is having candidates suited to the electorate. Waitakere is a classic example Labour got that wrong and we payed the price, Bennett should have got her arse kicked she didnt because we had the wrong candidate Maunakekei is another example of great person wrong electorate. So the first thing we need to do if we want to win seats is put the best person we have in them matched to the seat. Twyford should have been in Waitakere. Second thing is in a big Rural electorate Labour will never have the man power to get data its a waste of time, visibility is the key the candidate has to be everywhere and I mean work your arse off everywhere get signage sites go to everything sign written cars etc.

  10. Jill 10
    13 December 2011 at 4:08 pm

    Glad someone noticed something different happened in South Auckland. We actually increased our share of the Party vote and increased our electorate vote against the national trend. Not only that we had a new candidate in Manurewa in Louisa Wall. Just so nobody takes us for granted, we did adopt a more systematic, organising approach that involved a helluva lot of hard slog.
    What happened in Mangere and Manurewa, is that we mobilised hundreds of ordinary union members who live and work in South Auckland. We visited them personally at their homes and we asked them to go out and knock on the doors of people in their own streets. They each had “street captain” packs including a script, and lists of all those on the Roll in their streets from e-trac. These had been downloaded as CSV files, and converted into Excel spreadsheets so we didn’t have to chop down loads of trees. We included a notes column and some of them faithfully recorded where the dogs are and where the gates are locked for future reference.
    The street captains also delivered leaflets to their own neighbourhoods during the Election. We tracked every activity they did on spreadsheets and also entered our street captains’ details into e-trac (at least in Manurewa) though the system does have limitations in that it only allows one deliverer per block and we had up to four street captains. And we can’t tag our street captains or record their activities. But I thought it was a great tool.
    Some of our street captains came to Louisa Wall’s election night party. They were brimming with excitement as they told amazing stories of how they got people to the polling booths. We need to record these stories as they point to the potential strength of our Labour movement. Not a few Party faithful going out and doing election work as a Party machine, but a mass movement of ordinary working people who have been inspired to act!

The server will be getting hardware changes this evening starting at 10pm NZDT.
The site will be off line for some hours.