A record-high turnout of early voters suggests a 'youth-quake' is happening that would wrong-foot polls that are weighted using turnout rates at previous elections, Bernard Hickey reports.
Currently, the two major published polls put National ahead of Labour by eight or nine percentage points, but those polls have assumed that turnout rates among the young will be the same as for previous elections. Polls in Britain and the United States were wrong-footed by a surge in turnout among Brexit and Trump supporters that meant those demographics of older and poorer voters in rural areas were under-sampled.
As of 2pm on Wednesday, 806,043 people had voted, which was almost double the number who had voted at the same stage of the 2014 election. On the face of it, the very high early turnout suggests that young voters have voted in droves, mostly for the Jacinda Ardern-led Labour Party. The high early turnout suggests almost half the electorate, and well over one million people, will have voted before Saturday.
But analysts of the Electoral Commission's enrolment figures have challenged the idea of a youth-quake, pointing to figures up until September 19 showing 222,906 or 28 percent of the eligible population of 799,610 had not enrolled to vote by September 19, which was higher than the 16 percent of the population that had not enrolled by the end of the electoral process for 2014.
However, they had assumed that all of the young voters who enrolled and voted at the same time were included in the enrolment figures.
Electoral Commission spokesman Richard Thornton said by email that the enrolment numbers currently on its website were enrolments that had been processed so far and are not the final figure.
"There will be late enrolments that haven’t been processed yet, including enrolment forms coming in from advance voting places, and I can’t provide a figure for how many of those there are," he said.
"We will update the figures on the website at 6pm tonight, and we’ll put up the final figures up on the website by the end of next week, once all late enrolments that came in before the cut off time have been processed. People can enrol right up until midnight tonight."
The early figures suggest that enrolment rates among the young are higher.
Thornton said in the two weeks to September 17, numbers on the electoral roll increased by 44,361, a 55 percent increase on the same two week period before the last election in 2014 when 28,629 were added to the roll.
In that same two-week period, the number of 18-24 year olds on the roll increased by 9494, which was 43 percent higher than the 6668 who enrolled in the same period in 2014. In the 25-29 age group, it increased by 7397, which was 51 percent higher than the 4887 that enrolled in 2014.
The question is what has happened to enrolment and voting rates between September 17 and today. The figures are not in yet, but the anecdotal evidence of queues of young people suggest there has already been a 'youth-quake'.
So the polls might be wrong
If the young have voted at greater rates, that would mean the polls are likely to have underestimated the Labour and Green votes because they assume the same turnout rates for various demographics when they do their sampling and re-weighting of results. Polls have shown young voters are more likely to vote Labour and Green.
Even though the two big polls rely on landlines for their sampling (100 percent for Colmar Brunton and 75 percent for Reid Research) they keep calling until they reach their required number of young voters, so they don't miss out the youth completely, as some might think.
Some might argue that young people contactable by landline may be 'different' to those not contactable by landline. If this is true and there has been a change in turnout among the young, then the polls could underestimate both the scale of the vote and would have missed who they would vote for.
The polls show that the young are mostly voting for Labour and the Greens, while women are more likely to vote for Labour and men more likely to vote for National.
A Horizon Poll of 846 voters taken last week (and published yesterday) showed those leanings, as have others in previous weeks. The Horizon poll found 42 percent of enrolled voters between the age of 18-24 years who said they would vote also said they would vote Labour, with just 27 percent saying they would vote National. The portion of 25-34 years who would vote Labour was 45 percent to National's 21 percent.