}

Saturday, December 12, 2015

The winning alternative

The results of the first flag referendum, released last night, show that there was far more engagement in this referendum than we had been led to believe there would be. It was also more evidence, as if any more was really needed, that scanning social media commentary is a really terrible way to understand what the general public thinks about elections or public issues.

Comments on social media were mainly pretty unpleasant, with people constantly mocking and dismissing designs they didn’t like. This was done by everyone, including me: Toward the end of the process, I often called the “Koru” design “monkey butt”, as many others did. It was the only design I mocked.

However, in the main, the snarky social media comments that I personally saw were far more numerous coming from the fans of Red Peak than any other design. This may have been because the design only got on the ballot because of a campaign waged online through social media, but whatever the reason, the importance of the pro-Red Peak chatter was clearly exaggerated.

As the results above [SOURCE] show, Red Peak was a pretty distant third on first preferences. In fact, the two Kyle Lockwood designs were far out in front from the very beginning, and the other three designs combined had considerably less than half the first preference votes of either Kyle Lockwood design.

My first preference—“Silver Fern (Red, White and Blue)”—was in the lead the first two rounds, and it was only in the third round that the eventual winner, “Silver Fern (Black, White and Blue)” pulled in front, but only by some 5,000 votes. It eventually won on the fourth round with 50.53% of the votes to my choice’s 49.47%—a VERY close result.

The flag I that I ranked fifth—“Koru” was eliminated in the first round, which I felt was a good result. Although at first I thought the design was interesting, I grew to loathe it, and felt it by far looked the worst when hanging limp on a flagpole. Even so, it got more than 40% the number of first preference votes that Red Peak had.

I ranked Red Peak fourth. Unlike so many folks I read on social media, I just didn’t see anything special about it. The fact that it had a white chevron was interesting, but not in any way compelling to me.

There were two things that put me off the Red Peak design. First, I didn’t see anything representing New Zealand in that design—it could be for nearly any country. The truth is, the only thing connecting it to this country was the story made up for it. Second, the entire campaign for it felt manipulative, especially the campaign to get it on the ballot. I don’t like being manipulated, and I just couldn’t shake the feeling that it was all P.R. and spin, and for a design that just didn’t do anything for me.

I ranked “Silver Fern (Black and White)” third, even though I didn’t like it, simply because I liked it more than Koru (which isn’t much of a compliment…), and I wanted it ahead of Red Peak because of what I said in the previous paragraph. Together, it and Koru had more first preference votes than Red Peak had.

Now, what all this says is that New Zealanders who voted were overwhelmingly in favour of one of the Kyle Lockwood designs, and only a minority wanted Red Peak, and even smaller numbers wanted either of the other two designs. After months of constant hype about Red Peak, constant discussion on social media (and some mainstream media, too), RedPeak ended up in third place. This alone is why social media is such a terrible predictor of actual voter behaviour (as we should have learned from the last general election, or even the one before that…).

The preliminary results showed that 1,527,042 votes were received, including 148,022 informal votes (9.7%) and 2,476 invalid votes (0.16%). An informal vote was one in which it couldn’t be determined which design was the voter’s first preference. An invalid vote is one where the ballot itself cannot be processed, like if a voter made the QR code unreadable, their ballot had been replaced by a new ballot, and so on. There were more informal votes than Red Peak ended up with when it was eliminated.

Voter turnout was 48.16%, which is right in the middle of the turnouts for other recent postal ballot referenda: The 2013 referendum on asset sales had a turnout of 45.07%, and the 2009 smacking referendum had a turnout of 56.09%. Turnout is calculated by taking the total votes received as a percentage of the total number of voters enrolled. For this referendum, as at 19 November 2015, the date enrolment to vote in the first referendum closed, there were 3,170,726 enrolled voters.

Many people on social media claimed that they’d “spoil” their ballot, such as by putting an “X” in all five boxes, and they further claimed that such a huge number would do so that the eventual winner would only have a tiny number of votes. Others vowed to refuse to vote at all, so there'd be a tiny turnout. Social media got all that wrong, too.

In fact, the informals were quite high for this referendum, but, even so, it was less than 10% of the turnout, so I really don’t think it can be seen as any kind of “message” to John Key, about the process, or the desirability of keeping the current flag, particularly when there was no clear intent of people casting informal ballots, and it's impossible to know what reason(s) they had (apart, apparently, for negative feelings about something). However, we can see that the vast majority of people who voted took the process seriously and made their choices. This also suggests to me that the result of next year’s referendum is also not as sure a thing as many people have suggested.

Like many people, I believed the declaration that most voters will pick the current flag next year, even as I reminded myself that such polling was only speculative since there wasn’t yet one alternative design: Asking people to choose between the current known flag, and some hypothetical alternative couldn’t possibly produce reliable predictive data. Now that there’s one alternative, voters can make an actual choice about which flag they want, and I suspect the poll numbers will tighten up because of that.

The theme running through all of this is that, like so many other people, I was hoodwinked by social media. By pointing out how wrong it was, I hope to remind myself to never again pay any attention to it for elections, and to treat it as nothing more than entertainment.

Still, even with all this evidence of how lousy a predictor of voter behaviour social media truly is, there will be plenty of people who will continue to rely on it, as if the chatter they see and participate in actually means something—hint: It doesn’t, except as mere entertainment.

I must keep reminding myself of that fact—and so must we all.

No comments: