Speaker by Various Artists


The silent minority

by David Hood

I've been playing with the New Zealand Electoral Survey data for the 2011 election and what interests me more than arguments about voters and parties and alliances is what we can see in the data about why people vote, and what the data points to as what we as individuals can do about the declining voting rate.

Looking through the Election Survey data, you can ask the computer to build a model of important predictors of voting. Then you spend a while going “no, don’t use that variable” and redoing the model (then repeating the excluding of variables).

There were a number of reasons I was removing variables – for example, that they were too specific to a party to be of general use. (As a technical note, this is just playing with the survey results themselves, without working through the weightings to make them representative of the general population.)

Early on in the process, the computer recommended that, in 2011, liking Phil Goff was a strong predictor of voting.

What was important was not that you liked Phil Goff, it was having any opinion at all. Do know (99) matched to a 66% likelihood to vote, while having any other opinion at all matched to between 85% (right in the middle, so knowing you neither like nor dislike vote) and 93% for the highest voter turnout. What it was saying is that having an interest in politics is important – it was just saying it in a way that was party specific, and I didn’t want that party based view.

It also told me age was important, which I also discarded.

Basically, the 18-26 year olds (as I divide it up in this graph) are not-voting at about twice the rate of the next cohort. But age is not something you can do anything about, so it is not actually use for guiding individual behavior (at a government level, the voting age could be lowered and students enrolled at school, but that is not an individual action).

Another thing that it showed, and I discarded, was going to church.

While there is a 14% non-voting rate among those declaring No religion, and a 9% non-vote rate among those declaring themselves Very Religious, telling people to get a religion to increase the voting rate is a non-starter. Anyway, I suspect it is putting the cart before the horse, as other concepts tied in with “dutifulness” have shown up as well.

So, after checking, and discarding 35 or so variables, what the computer was suggesting to me was a fairly straightforward model.

Let's go through what the variables are and what it means in terms of actions people can actually take.

jknow  is top of the list, and is "Political Knowledge Scale". It is representative of all of the “knowing things about politics” measures.

While it is not wonderfully clear in a bar graph, there is a steady progression in political knowledge and voting.

So, the single biggest thing people can do, acting in a private capacity, is increase other people's knowledge of politics. If, through talking to them, you increase their knowledge of politics from 0 to 0.5 (nothing to half informed) you increase their chance of voting from 67% to 89%, and they are a more informed voter (but that is more about how they are voting than participating in the process).

I think this is what was behind discarded variables like “Liking Phil Goff”. If you have enough knowledge of politics to form an opinion about the then leader of the main opposition party, you were highly likely to vote.

jduty is the response to the proposition “it is a citizen's duty to vote”:

Rendered as a line plot of the proportions this looks like:


This initially seems not something we can do much about in a liberal democracy, without resorting to clichés like “Make young people do national service” (or, indeed, as an entire generation missed it, “Make everyone do national service”) and maybe in a hand-wavy unexplained way they will get a sense of duty.

But I want to make the point that the duty criteria is applying to people who slip past the knowledge one – so knowledge trumps duty. Now, they may be interlinked (I haven’t explored that) and knowledge could lead to a feeling of duty, but regardless, duty doesn’t matter if people have enough knowledge to vote.

I also suspect that this ties in with variables like religiosity – people who are dutiful about attending church bring that duty to the voting process, regardless of political knowledge.

There are plenty of other posts that could also be made about increasing people's sense of duty/care for society/involvement in society, but you won’t find that data in the Election Survey. The practical step is to increase knowledge (talk to people).

Jdiffvot is short for jdiffvoting and is based on the question “does voting make any difference to what happens?”, so is very much in the engagement/enthusiasm/knowledge area.

This is a rising pattern with not a lot of difference between extreme and mild answers.

Now actual disengagement/rejection of the political process is a hard one for individuals to do much about (parties can clearly offer alternatives and carry them out), but this does only kick in for those who aren’t voting due to knowledge or duty, so it is making less of a difference.

jmpspaci is short for jmpspacific, the question that asks “how many Pacific MPs should there be?”

This sounds like a hot-button topic, but bear with me on this. The actual percentages are:


There are two groups not-voting here – the Don’t Know, and the Fewer. The “Don’t Know” can be fixed way up the chain in the political knowledge section. And as for the “Fewer”…

Well, to be honest, if someone who doesn’t know much about politics and doesn’t feel it is a duty to vote, feels that voting can make a difference, and feels (despite having a low political knowledge) that there should be fewer Pacific Island MPs, I personally am not that motivated in finding a way to get them to vote.

Now, there are other groups where things effect voting – for example English language fluency, but in terms of a simple model that makes the most difference the computer recommended this one (which suggests that those groups are smaller in number and their needs are more specific). While a different exercise could be conducted to find a set of needs and independent interventions. A quick easy thing you can do, if you want to increase election turnout is talk politics- at the hairdresser, in the supermarket, on the bus or train. The question is, do people want to increase turnout enough to do that?

Finally, in a show your working kind of way, if anyone wants to review my R code, it is here on Github



by Rob Salmond

The annual Wellington on a Plate restaurant fortnight ended yesterday. I didn’t have a single dish. That’s because the annual Burger Wellington competition also ended yesterday, and I had eleven of those instead.

Nearly 100 restaurants put up an offering in Burger Wellington, mainly matched with hipster-level craft beer from Aro Valley’s Garage Project.

I refused chicken and vegetarian burgers on principle. This principle is so powerful that it’s not necessary to talk about it. Among the remaining surf and/or turf selection, amusing titles sent places up my list, while pretentious descriptions sent them sliding off the bottom, no matter how amazing their restaurant is otherwise. If you’re going to call your fries “WBC chipped potatoes,” or say your salad greens are “Cuba St Fruit Mart vegetables,” I’m going to call your competition. But if you’re serving up an “Aporkalypse Now” or a “Ba ba baaa ba ba bar lamb,” I’m in.

Of course, being foodie-bait, some of the burgers came in inexplicably overpriced. Some burgers, with no fries and only a couple of exotic things, were going for $25. My Stateside drive-thru go-to, the In’N’Out Double Double Animal Style would pull a solid top half finish in Burger Wellington, but it costs $3.50.

Through my two weeks of gluttonous endurance, I think I’ve learned a few things about what makes a great, classic burger.

A steamed bun tastes luxuriously soft, but it turns very quickly into a soggy bun, which tastes like tramping. All buns, even steamed ones, are better with a bit of crunch, from a walk under the grill or across the griddle.

Meat patties need some burny crunchy bits on the outside, even when they’re beautifully pink in the middle. And unless you’ve really got tartare quality meat, two thin patties are better than one thick one. More patties, more crispy bits.

Bigger isn’t always better.

Even in a world of truffled brie and cave-aged blue, there’s a special place for a strong, thick, melty slice of cheddar.

The veggies that pop are a little different, like fried pickled onions or beetroot relish.

If you’re going to pimp-up your special sauce, be aggressive. Nothing’s as underwhelming as a not-tangy-enough sauce.

Always serve some pickle. On the side, sliced in, it doesn’t matter. Serve it up.

Last, if you’re trying to stand out, dare to be different. The burger’s a pretty forgiving platform if you’re keen to experiment.

Five Boroughs in Mt Victoria is my winner. They didn’t do a classical burger at all. They did Vadereqsue all-black dish, plate and everything. The bun was squid ink. Black garlic mayo. Charred venison. Deeply sautéed mushrooms peeking out the bottom. The only bright ingredient was the truffled brie, which they gave us lots of, but hidden under the lid.

It tasted even better than it looked, with the la-de-da fancy cheese melting through the crunchy top of the venison patty, and garlicky mushrooms soaking up any juices that spilled. The ink bun tasted just salty enough to be different. It worked perfectly.

One of my client’s staffers wasn’t a believer. She thought only proper fancy restaurants could do a proper fancy burger, and Five Boroughs was just a mid-range burger joint trying to play up a league. Then she went there, and saw the light. Several staffers made multiple trips after that.

My second place wasn’t a classical burger either. Ti Kouka café’s Aporkalypse Now, geniusly matched with a beer called Death from Above, was more porktopia than aporkalypse. There was chilli oil on the side, pork crackling floating around because, well,  why not, and an amazing pork shoulder-and-bacon thing in the middle. The only downside was a borderline soggy bun. A hamlicious, pigtacular mess.

It wasn’t only oddballs that floated to the top of my rankings – Egmont St Eatery, The Bresolin, and Bin 44 were all twists on a classic red meat cheeseburger, all done really nicely, and that rounded out my top five.

I’m counting down to the start of next year’s two week burgergasm. 350 sleeps to go.

And now… time for a run.


The problem is Serco

by John Palethorpe

They were, until a few months ago, the biggest company you’d never heard of. Well, that’s not entirely true. In New Zealand, Serco are not a big company. They have two public sector contracts, just two. Unfortunately, you already know about those two contracts. Because they’re in New Zealand’s prisons.

It would be nice to say that this was a unique set of circumstances, that the fact that Mt Eden Corrections Facility is a remand prison meant that there were always going to be problems. I say that because that is exactly what the Minister for Corrections said when he was challenged on the rates of violence at Mt Eden on TV3’s The Nation. That was in May, before the Fight Club allegations had surfaced.

It would also be nice to dismiss Serco overcharging the UK Government by NZD $164,776,181 for tagging ex-offenders who had either returned to prison, left the country or were dead. As well as their chronic mismanagement of Fiona Stanley Hospital in Australia. Not to mention forgetting that Serco only managed to make a profit in 2014 from the lucrative nature of their contracts with the Australian Government and their deeply awful refugee policy. Actually, no, it wouldn’t.

The Department of Corrections is currently investigating the fight clubs, arrests of prisoners for managing a meth ring, the guards offering ‘sparring tips’, and much more. What was supposed to take a month is now going to take four, with Phase 1 reporting October 30th and Phase 2 on November 30th. If you’re seeking the way to minimise the impact of what is clearly a catastrophic systemic failure, punting that far down the line and breaking up the report is definitely one way of doing it.

But basically, it’s just not thorough enough. Prison officers don’t have faith in Corrections to conduct their investigation with integrity. Ex-inmates fear speaking out, in case of retribution if they end up back inside. And while Serco have mismanaged Mt Eden, surely some questions need to be put to Corrections and the Minister for Corrections about exactly why they needed leaked videos on YouTube to let them know one of their prisons was failing?

And boy have they failed. They failed in the UK, they failed in Australia and they’ve failed here. But the investigation isn’t focusing on the systemic problems caused by a company who slash staff numbers to make a profit. It’s not recognising the tragic history of Serco involvement in overseas public services. The problem isn’t fight clubs, it’s not prisoners. It’s Serco.

What’s needed is an independent inquiry. Labour’s Kelvin Davis has called for one as this whole mess unfolded. David Clendon of NZ Greens has also called for their contract to be cancelled. The idea of an independent investigation is supported by the Howard League For Penal Reform, the Corrections Association of New Zealand and the PSA. Because unlike many, many other countries New Zealand lacks a truly independent inspectorate of prisons. Say No To Serco Aotearoa is joining the politicians and the unions in calling for an independent inquiry.

Despite all the failures, Serco will continue to make money from Mt Eden. Because their contract ensures they will. The fines that Sam Lotu-Iiga has talked about aren’t fines, they’re just not paying them their performance bonuses. Bonuses which make up 10% of the contract. The other 90% is, contractually, untouchable.

There’s obviously more to do here. In New Zealand Serco are not a big company. Sure, they have a $300,000,000 contract for Mt Eden. Yes, they’re expecting $30,000,000 a year in profit from South Auckland Wiri prison. But that’s just two contracts. With social impact bonds and social housing up for privatisation in the future, you can bet Serco will express an interest.

Why? Because that’s what they do. Serco don’t stop just because they’ve been a catastrophic failure in one part of the public sector. They’re dedicated, it seems, to making a profit while failing to deliver lots of different public services. Right now, we’re in a good position to ensure that Serco don’t make a first class profit, while New Zealand gets second class services.

So, while we’re asking for an independent inquiry into the Mt Eden fiasco, we’re also demanding a moratorium on Serco being considered for any further public sector contracts. It makes sense. Why would you give them a second chance in New Zealand, when the number of chances and failures globally far exceed that number?


With help from the wonderful ActionStation, we’ve started a petition calling for an independent inquiry and a moratorium on Serco bids for New Zealand’s public services. You can sign up, here. Our campaign is in its early stages, and your support is vital You can find us on Facebook here, or track the hashtag #NoToSerco on twitter.


John Palethorpe

No To Serco Aotearoa


Meet the middle

by Rob Salmond

Last week I did a post about Jeremy Corbyn, and the possible worldwide implications of his rise. There’s been plenty of reaction. At the risk of prolonging a PFJ / JPF situation, I’d like to pick up on some points people have made.


Around a third of New Zealand’s population are leftists. Same for right-wingers. But you need 50% of the vote to govern. Guess where the rest comes from? The middle.

There’s certainly a legitimate debate about how best to approach winning over the political centre. But any suggestions that elections can be won without doing well in the centre aren’t grounded in reality.

Rule 1 in politics is “learn to count.” 33 < 50.

Some think this idea – of appealing to the middle – is somehow new or untested. Stephanie Rodgers, for example, talks about it being Labour’s plan “since 2008.”

Starting narrowly, anyone who looks at Labour’s successful 2005 platform and sees anything other than an appeal to the centre is dreaming.

More broadly, this theory – the median voter theorem – has been one of the most dominant global ideas in political science and politics since at least 1948. While it applies at the party-level most directly in FPP electoral systems, its logic applies just as well at the coalition level in most PR environments, too.

There’s actually plenty of revisionism in this debate, I’m sure myself included. Some PA commenters were sure John Smith won UK Labour the 1997 election despite dying in 1994. And Mike Smith is certain NZ Labour’s supposedly sharp tack left won it the 1999 election, rather than, say, comically self-destructive opponents.


Chris Trotter made two points in his posts on this debate: first that centrists believe in nothing; and second that centrists are Nazis. Yes, seriously. I’m surprised at Chris. Everyone knows you can’t be a nihilist and a Nazi at the same time.

Chris is welcome to his self-parodies about Orwell and Hitler. It sure it a long way from his earlier suggestion that Labour chase “Waitakere Man”. It’s almost like 2015 Chris is calling 2010 Chris a sympathizer…

His one argument worthy of response was that centrists lack any political beliefs. Others made similar arguments, too. They’re wrong.

Chris and others suggests centrists aren’t actually in the middle, they just say they’re in the middle because of their confused psychology. But rule 2 in politics is “perception is reality.” That includes the perceptions of people you might not agree with. If they think they’re in the middle, then they are.

While centrists often do not have strongly held ideological views, they do have beliefs and values. They don’t wake up each morning waiting for ideologues to fill their empty heads with things to think. Centrists aren’t goldfish.


Percent agreeing…





Import controls?




More $ for health?








Dole = bludger?




The Table gives some responses from the 2011 NZES, by ideological group. It suggests:

-  Centrists think more like lefties on economic protections and public investment

-  But they think more like right-wingers on new taxes and welfare

That’s helpful to know. You’re more likely to win centrists’ votes if you emphasize the issues where you agree with them, and downplay the issues where you don’t.

It’s the old story of flies, honey, and vinegar.

“Being relevant,” however, doesn’t prevent “standing for something.” None of this is necessarily about changing policy. Labour needs proper social democratic policy in order to stay Labour. Instead it’s about – for want of a better word – “narrative.” And issue emphasis.

From the 2011 data above, if Labour had convinced the population that the most important problem facing New Zealand was lack of public investment, the left could have won. If the right convinces the public that the most important issue is avoiding new taxes, they do very well.


Danyl McLaughlin helpfully mused:

I think [Labour would] look for something new. And I don’t think it would be movement along the values spectrum. It would look, probably, like the data-driven grass-roots campaigning of Obama.

Good news! Data-driven, grass roots campaigning is exactly what Labour – and everybody else – is pursuing right now.

We’re learning more about every voter before we make contact, and talking with more voters ever before. Labour has made good progress in parts of this work, as I discuss elsewhere. In some areas, we lag behind National. In others, we’re in front.

But, as 2014 shows, you can have all the whizz-bang data and volunteers with phones you like – if people don’t like what you’re saying, it doesn’t matter.

Make your message relevant to those voters who’ll decide the outcome, and you’re a chance. Don’t, and you’re toast.


In defence of the centre

by Rob Salmond

The leadership election in UK Labour has crystalised into everyone else vs Jeremy Corbyn’s fearless jump to the left. This has led some commentators to ask whether the old political orthodoxy of “move to the middle” is, long term, a death-knell for left-leaning parties. Here’s George Monbiot in the Guardian:

Across three decades New Labour strategists have overlooked a crucial reality: politicians reinforce the values they espouse. The harder you try to win by adopting your opponents’ values, the more you legitimise and promote them, making your task – and that of your successors – more difficult.

Monbiot says Tony Blair’s three election victories, won with a strategy of pursuing the centre, were actually harbingers of a long period in the wilderness, as the Blair-era hardened centre-right attitudes across the UK. There is at least a little recent academic research supporting this conclusion, too:

Voter surveys from Germany, Sweden, and Britain show us that although uncommitted centrists initially respond favorably to Social Democratic moderation, these voters don’t stay with centrist Social Democratic parties for long and the moves to the middle also increase abstentions and defections from formerly core Social Democratic voters.

This idea is taking hold most strongly in UK Labour, with Corybn’s impending landslide, but it also has some following in the US, with hard left alternative Benrie Sanders of Vermont packing stadiums to provide a further left alternative to Hilary Clinton.

However, I’ve got three problems with this thesis. (For the record, these are my personal views only, not Labour’s views.)

First, they don’t consider the alternative. How have centre-left parties gone when they’ve tacked away from the centre? It doesn’t happen often, but when it does, it goes badly.

Michael Foot and Neil Kinnock lead the UK Labour party through seventeen years of Tory rule. It was bleak. Why did they keep losing for so long? Because under Foot the hard left got to run Labour’s policy, alienating Labour from the middle ground. And under Kinnock the hard left continued to battle prominently for policy influence, allowing Labours’ opponents to scaremonger successfully about what a vote for Labour really means.

How did UK Labour break out of this funk? Tony Blair.

And if you think that lesson, of declining centre-left fortunes when its narrative swings left, doesn’t apply for in modern New Zealand, here are two phrases you may find familiar: “Man ban.” “Sorry for being a man.”

Second, peoples’ votes are more malleable than their values. Monbiot says:

The task is to rebuild the party’s values, reclaim the democratic debate, pull the centre back towards the left and change – as Clement Attlee and Thatcher did in different ways – the soul of the nation.

The part where Monbiot is right is that the centre ground really is where elections are won and lost. (That statement is more controversial in New Zealand than it should be.) There are a ton of people there, and those peoples’ own identities are of being open to voting left or right. Below is a chart showing how New Zealanders perceive themselves, Labour, and National. Over a third see themselves as right of where they see Labour, and left of where they see National. That’s huge.

But “pulling the centre back towards the left” is massively, massively hard. You win those people over by being relevant to them as they are, not by telling them they’re worldview needs a rethink. It is just basic psychology. Tell people they were right all along; they like you. Tell people they were wrong all along; they don’t.

And if you win a majority of centrists, you win. The New Zealand Election Study series records six MMP elections in New Zealand – the three where Labour did best among centrists were the three Labour won.

That’s another message from the adacemic study I quoted above – in Germany, Sweden, and the UK, the elections where the left did best among centrists were the elections where they took power. As their popularity among centrists declined, so did their seat share.

Third, Monbiot conflates policy with competence:

Labour’s inability to provide a loud and proud alternative to Conservative policies explains why so much of its base switched to Ukip at the last election. Corbyn’s political clarity explains why the same people are flocking back to him.

Clarity is always a good quality in a politician. Saying what they’re for, and saying why they’re for it in simple, accessible language, are cornerstones of good political communication. But you can have clarity, and be competent, no matter where you stand on the ideological spectrum. “Clear” does not mean “extreme.”

Here are some of the best, clearest centre-left communicators in modern political history. Clinton and Obama in the US, Lange in New Zealand, Hawke and Keating in Australia, Blair in the UK. Clear communicators, all. Politically competent, all. Hard left? None.