How special votes could change the election
On the provisional result, the 2017 election is New Zealand’s closest MMP election yet in terms of the gap between the Left and Right blocs. With a record 384,072 special votes (around 15 percent of the total) still to be counted, the final result will be even closer. While the special votes will not change the fact that either bloc relies on NZ First to form a government, they could have an important impact on negotiations – primarily for the prospects of a Labour-Green-NZ First coalition. They are also important to the public discourse around coalition options.
Special votes have consistently favoured the Left in recent times, particularly the Green Party. Special votes include overseas votes, dictated votes, and votes cast by people who enrolled during the advance voting period. The latter category is especially important to watch this election due to an apparent late surge of predominantly younger voters (the fabled “youthquake”).
The provisional election result has National on 58 seats, Labour on 45, the Greens on seven, New Zealand First on nine, and ACT on one. Projections by Graeme Edgeler and Chuan-Zheng Lee show that if the special votes follow a similar pattern as in 2014, National will lose two seats: one to the Greens and one to Labour. This will put Labour and the Greens on a combined 54 seats: very close to National’s 56. A Labour-Green-NZ First coalition would then command a comfortable majority of 63 seats.
This result appears likely, but how confident can we be? Curious, I made a spreadsheet tool to test out different special vote scenarios. It has had a bit of use after I shared it on Twitter, and I have seen a few wildly optimistic scenarios plugged in. I’m a bit concerned I have created a denial mechanism.
After spending far too long playing with the spreadsheet myself, here are five scenarios for how the special votes could change the election result. You can explore them visually in this interactive infographic:
Scenario 1: 2014 pattern
This scenario matches the projections described above. It assumes that parties underperform or overperform in the special votes relative to the ordinary votes by the same proportion as they did in 2014. For example, National’s share of special votes was 17 percent lower than its share of ordinary votes.
I’ve rounded the numbers to a whole percentage point, which gives the following shares of special votes: National 38 percent, Labour 41 percent, Greens 9 percent, NZ First 6 percent, and all other parties 6 percent. (I’ve grouped all the other parties including ACT together here as it’s extremely unlikely they will impact the seat distribution.) As mentioned, this leads to 56 seats for National, 54 seats for the Labour-Greens bloc, and NZ First sitting unchanged on 9 seats.
This scenario serves as a good baseline. The remaining four scenarios deviate from this in different ways.
Scenario 2: National comeback
In this scenario, National manages to hold onto its 57th seat rather than lose it to Labour. This would happen if Labour were to lose 1 percent of special votes to National and 1 percent to the Greens compared with Scenario 1, leaving it tied with National on 39 percent. There are also other ways that this could happen: the key condition is that National gets an equal or higher percentage of the special votes as Labour does.
As a relatively small shift from the 2014 pattern, this could definitely happen. However, it would go against a trend of National performing increasingly worse in the special votes relative to the ordinary votes over the last three elections (and vice versa for Labour and the Greens). This seems particularly unlikely to occur if most of the increase in special votes this year is indeed from younger voters.
Scenario 3: Keen on Green
In this scenario, the Greens take a second seat off National and the Left and Right blocs draw level at 55 seats each. The first key to this happening is the Greens getting around 13 percent or more of the special votes. The second condition is that Labour needs to be around 6 percent above National, otherwise the Greens would take a seat off them instead. This would require National to drop back to around 34 percent, with Labour holding fairly strong on at least 40 percent.
Clearly, this is a big stretch. The Greens did manage to get 15.4 percent of special votes in 2014, but that was from a much stronger provisional result of 10 percent (compared with 5.9 percent this year). For this to happen, the youthquake would need to be high on the Richter scale, and it would need to heavily favour the Greens over Labour.
Scenario 4: Jacindamania
In this scenario, it’s Labour that manages to take another seat off National. To do this, Labour would need to beat National by around 12 percent in the special votes. They’d also need to avoid pushing the Greens below about 7 percent, otherwise the Greens wouldn’t pick up their eighth seat. This could require Labour to get around 46 percent of special votes to National’s 34 percent.
This is very similar to Scenario 3 but with a redistribution of votes between Labour and the Greens (together they have 53 percent in both scenarios). It seems equally if not more unlikely, unless young voters did show up in droves mainly to vote for Jacinda Ardern.
Scenario 5: The Dark Horse
In this scenario, NZ First manages to take a seat off National. The party would need to get around 9 percent of special votes to do this, requiring an extra 3 percentage points compared to Scenario 1. In this scenario those extra votes all come from National; if some came from Labour instead then NZ First’s extra seat would also come from Labour.
This scenario would buck history more than any of the others. Across the last three elections, NZ First has done worst in special votes relative to ordinary votes out of all sizeable parties except for the Conservatives. For them to go up 1.5 percentage points this year would be utterly remarkable – but then again, is anyone willing to bet against Winston?
Could National hang on to all 58 seats?
There are ways this could happen, but it is very unlikely. One way is if the Greens received less than 7 percent of special votes and National beat Labour by at least 5 percent. Alternatively, National could take a seat off Labour if it were ahead by at least 11 percent. Either of these would require National to get at least 44 percent of the special vote – well above what it received in any of the last three elections.
What these scenarios tell us
The five (and a half) scenarios above present a wide range of possibilities for what could happen with the special votes. Of course, we could come up with even more exotic scenarios, but these ones already push the boundaries plenty.
Based on past patterns, National is almost certain to lose a seat to the Greens, and it seems more likely than not that they will also lose a seat to Labour. The chances of any further changes are pretty remote, but can’t be completely ruled out. To take another seat off National, Labour and the Greens would need to get an additional 3 percent of special votes beyond what they are projected to receive. Furthermore, these would need to be optimally distributed between the two parties as seen in Scenarios 3 and 4.
Whatever happens, none of this will change the big picture outcome of the election. However, it’s important for the public to understand – and for the media to report – that the National and Labour-Green blocs will likely be just a couple of seats apart once the special votes are counted. We’ll have to wait until October 7 to know for sure. And so will Winston.
We value fearless, independent journalism. We hope you do too.
Newsroom has repeatedly broken big, important national news stories and established a platform for quality journalism on issues ranging from climate change, sexual harassment and bullying through to science, foreign affairs, women’s sports and politics.
But we need your support to continue, whether it is great, small, ongoing or a one-off donation. If you believe in high quality journalism being available for all please click to become a Newsroom supporter.