John Key and David Cunliffe both spent much of the election campaign talking about the dreaded “things that New Zealanders really care about”. But Key, under direct attack, was much more disciplined about sticking to those things. The metacampaign, Dirty Politics, and the Dotcom Bomb were worth nothing more than haughty dismissal. At the time this seemed arrogant and ill-advised — how could he just shrug off such scandal? But he did. The National party ran an orthodox, modern campaign. They stuck to their guns amid all the madness, and the result was triumphant.

The poster child for this campaign was candidate Chris Bishop, who ran an old-fashioned shoe-leather campaign in the Labour stronghold of Hutt South and pushed the party’s strategic genius Trevor Mallard to within a few hundred votes of losing the seat he has held for 20 years. No stunts, no social media hype, no concern for the wittering about his being a former tobacco lobbyist, he just talked to the people and listened to the answers that came back.

But how did they know?
A few days before the election, Russell Brown was polled by Curia, and posted the questions. Russell noted that “these questions aren’t really geared towards firmly decided voters” — and this is key. They don’t presuppose a strong position, they’re simple and to the point, and they lead respondents through a narrative about what leadership is about and what government is for. And they seem as if they would yield an enormous amount of actionable data, not only about what voters value, but about why they value it. On election night, Key singled out Curia pollster David Farrar for particular praise:

“To the best pollster in New Zealand — and don’t charge us more for it — David Farrar, who got his numbers right! And who I rang night after night, even though I was told not to, just to check.”

John Key could afford to dismiss the metapolitics because he had plenty of good data telling him that people didn’t care about it, and to the extent that they did care about it, it favoured him. The single most evident difference between the campaigns is that When John Key said “the things New Zealanders really care about” he actually knew that these were the things that New Zealanders actually care about. The National party ran a reality-based campaign, not a hype-based, or a hope-based, or a faith-based campaign. In this they mirrored the most famous hope-based campaign of all time — Barack Obama’s — where the breezy, idealistic messaging was built on a rock-solid data foundation.

Key seems to have been the only party leader who was really secure in this knowledge. The Greens and Labour did seem to want to stick to their guns, but their data was evidently not as good, and they bought at least some of the hype that Dirty Politics and the Dotcom Bomb would bring Key low. So did I. But nothing much is riding on my out-of touch delusions. But opposition has a responsibility to be, if not reality-based, then at the very least reality-adjacent.

Play, or get off the field
I am a terrible bore on this topic. I have been criticising Labour, in particular, since at least 2007 on their unwillingness or inability to bring modern data-driven campaign and media strategy to bear in their campaigns — effectively, to embrace The Game and play it to win, rather than regarding it as a regrettable impediment to some pure and glorious ideological victory. Mostly the responses I get from the faithful fall under one or more of the following:

• National has inherent advantages because the evil old MSM is biased
• the polls are biased because landlines or something
• the inherent nature of modern neoliberal society is biased
• people have a cognitive bias towards the right’s messaging because Maslow
• it inevitably leads to populist pandering and the death of principle
• The Game itself devours the immortal soul of anyone who plays ( which forms a handy way to demonise anyone who does play)

But data is not a Ring of Power that puts its users in thrall to the Dark Lord. And, unlike the One Ring, it can’t be thrown into a volcano and the world saved from its pernicious influence. Evidence and strategy are here to stay. Use them, or you’re going to get used. The techniques available to David Farrar and the National party are not magic. They are available to anyone. Whether Labour has poor data or whether they use it poorly I do not know. It looks similar from the outside, and I have heard both from people who ought to know. But it doesn’t really matter. Data is only as good as what you do with it. Whatever they’re doing with it isn’t good enough.

The best example from this campaign isn’t Labour, however — it’s Kim Dotcom. He said on election night that it was only in the past two weeks that he realised how tainted his brand was. He threw $4.5 million at the Internet MANA campaign and it polled less than the Māori Party, who had the same number of incumbent candidates and a tiny fraction of the money and expertise. Had he thought to spend $30,000 on market research* asking questions like those asked by Curia about what New Zealanders think of Kim Dotcom, he could have saved himself the rest of the money, and saved Hone Harawira his seat, Laila Harré her political credibility, and the wider left a severe beating.

That is effective use of data: not asking questions to tell you what you want to hear, but to tell you what you need to know. This electoral bloodletting is an opportunity for the NZ political left to become reality-adjacent, and we can only hope they take it. Because if they don’t, reality is just going to keep winning.

L

* In response to this figure, UMR pollster Stephen Mills tweeted “$1000 would have been enough”.

I wrote most of this before DPF’s post on the threshold, including his link to Chris Bishop’s handy paper on representation and stability went up, so read that first. In fact, you’d also do very well to look over BK Drinkwater’s series comparing electoral systems: noise, wastage, proportionality, and a critique of some critiques of SM, although note that the SM numbers assume a 70-50 electorate-list split as per our MMP system at present. This is good from an apples-apples perspective, even if it’s not an option that’s actually on the table. I also wrote it before my more-recent post on the topic, for which some people have begun offering their preferred electoral modifications. Wonderful!

Much like the subtext to the s59 referendum question was ‘do you like the anti-smacking bill?’, the question above lurks behind the forthcoming debate on MMP, for which the troops are currently massing.

The likeliest contender, in my view, is the retention of MMP as we have it now, with a 5% threshold and a 70-50 split. Other less-likely contenders, again in my view, are as follows:

• STV, as employed in some local body elections and for the Australian Senate.
• SM, as apparently favoured by National and employed in the Republic of Korea.

Note that FPP isn’t in this list. I don’t think NZ would go back. Modified MMP also isn’t in the list; not that I think it isn’t a credible contender, just that the way the process is structured (referendum: MMP yes/no; if no, referendum on alternate systems) doesn’t seem likely to permit it. There are lots of other peripheral options, such as open list; run-off or instant run-off; or any number of other possibilities. Feel free to argue your corner.

But what sort of system do people actually want? As I see it, within a centralised democratic structure such as we have, relevant factors include the following:

• Transparency. Results in transparent electoral systems are clear and obvious; how a particular candidate, party or government was elected is reasonably self-evident. FPP is very transparent. STV is very opaque.
• Simplicity. Simple systems are easy for people whose political engagement stretches to ticking a box or two every three years to understand. Again; FPP is very simple, STV is not at all.
• Proportionality. Proportional systems elect candidates from parties according to the party’s share of the vote. FPP is not proportional at all. STV is often claimed to be proportional, but it’s really fauxportional, often producing results which seem proportional but were arrived at by non-proportional means. Open list is (in principle) perfectly proportional. Thresholds in proportional systems and the number of electorates in mixed systems are also relevant to this question.
• Representativeness. Similar but orthogonal to proportionality, a representative system contains mechanisms to guarantee certain segments of the electorate representation. This is a complex notion; geographical electorates are such a representative measure, ensuring that people from the geographical margins are represented, when a non-geographically-determined system (such as purely proportional open list) might marginalise them. Reserve seats for tangata whenua or other groups are another such form of representativeness.
• Low wastage/regret. Conventional wisdom is that the prospect of a wasted vote depresses turnout (or changes behaviour) among voters who believe their vote might be wasted, which is a self-perpetuating cycle. This is most evident in FPP, but is also present in proportional systems to an extent, due to the effects of a threshold.
• Decisiveness. Decisive systems produce strong, stable executive governments with few constraints on their power. FPP, except in the rare case of a hung parliament being elected, is decisive, while proportional systems which elect a number of parties and rely on coalitions are less decisive.
• Small size. Self-explanatory. Any system can be made large or small, but this frequently has huge impacts on other factors.
• Durability. Durable systems are not prone to future governments tinkering with, amending or replacing them. FPP was extremely durable. MMP has proven fairly durable. This is a meta-factor, in a sense; it seems like anyone valuing this factor highly should lobby for one of the less-extreme systems; a second-best choice, rather than a perfectly proportional system or a highly decisive system, since ‘pretty good’ is less likely to be overturned.

In principle, the relative importance a person assigns to of each of these factors should point to that person’s ideal electoral system. Could be programmed into a handy poll in the leadup to the referendum; in fact, I bet it already has been, I just haven’t found it.

There are other relevant electoral changes, as well. Here are a few; please add your own:

• Size of parliament and division of seats. Yeah, I listed it above – what I’m referring to here is the electorate-list split in mixed systems; the North/South island and rural/urban splits, that sort of thing. Also the vexed question: how many MPs overall?
• Allocation of seats. Historically, the One True Way in NZ was for seats to be allocated along population-geographic lines. Nowadays it’s a mix of population-geographic and party allegiance. But what other means of allocating seats are there? What would happen if seats were allocated according to social class? Income? Level of education? Ethnicity? Religion? The history of democracy contains precedent for all these things in one way or another.
• Decentralisation. Federation of micro-states? Balance of central and local government power? How does one dismantle centralised democracy using democratic mechanisms?
• Electoral term and other constitutional institutions. Our three-year term is quite short, and there are few checks on the executive ability of governments – as long as they have a parliamentary majority, there’s little they can’t (and won’t) do. Do we need a second chamber? A longer term? Should one go along with the other?
• Referenda and non-electoral plebiscites. What should their status be? Other representative mechanisms, such as citizens’ juries?
• The big one. What difference would becoming a republic make anyhow? A better question: if people knew that NZ would become a republic in the near future, how might their electoral preferences change?

Please, answer the question. What do you actually want in an electoral system, and why? And more than that — what do you want, and what do you think is (even remotely) plausible?

L

Update: Scott Yorke has a few choice words on the topic, as well.