It took all night, but early this morning it became clear that Labor had won the Eden-Monaro by-election, when the last of the pre-poll booths reported their results, and a first batch of postal votes were far better for Labor than postal votes usually are.
In this post I’ll show you a map of the results and swings, and break down how each part of Eden-Monaro swung, and how turnout shifted away from election day and towards pre-poll voting.
First up, this table shows the two-party-preferred vote for Labor and the total formal votes cast in each part of the seat in 2019 and 2020. These are based on the five regions I defined in my seat guide.
|Voter group||Labor 2PP||Labor 2PP 2019||Swing to ALP||Formal votes 2020||Formal votes 2019||Turnout change (%)|
I have also separated out the pre-poll booths based on what part of the seat they lay in. The regions don’t necessarily map perfectly onto the pre-poll areas: in particular the “north” includes the former Palerang council area, but there were no pre-poll booths there. It’s more likely that those voters might have gone to Queanbeyan, not to the Yass booth which is ostensibly in the same zone.
On the election day vote, Labor gained swings in the east, south and west. The swing in the east, which covers the coastal communities including Bega and Eden, was large, while it was much smaller in the other two areas, which tend to be the most conservative parts of the seat.
The Liberal Party gained swings in the north and in Queanbeyan, which remains the best area for Labor but not by as wide a margin as it once was.
The swings in the pre-poll booths have some similarities: Labor did very well on the coast and not so well in Queanbeyan. Labor picked up a swing at the Yass booth, and lost ground in pre-poll in the south and west (at the Cooma, Jindabyne and Tumut booths).
The table shows how the turnout changed, in terms of formal votes. The total turnout in ordinary votes dropped by between 7.5% (in the south) and 22% (in Queanbeyan). There were big increases in pre-poll turnout in Queanbeyan and the east, but turnout actually dropped at the Cooma pre-poll booth, which outweighed a slight increase at the other south booth in Jindabyne. Turnout at most local pre-poll booths jumped, although once you factor in absentee pre-poll votes in 2019 (which don’t exist at a by-election) the total turnout is roughly even. It will still likely be a larger proportion of a smaller overall turnout.
Overall there weren’t dramatic shifts in the pre-poll voting patterns. Labor’s two-party-preferred vote in the pre-poll dropped from 49.4% to 49.3%. The big change was in the postal vote.
We received the results for the first batch of the postal vote late in the night, oddly via tweet from the AEC. The actual figures won’t be in the tally room until tomorrow, and we don’t have primary vote data, but it showed Liberal candidate Kotvojs winning a sample of almost 5000 postal votes by just 70 votes. That’s 50.7%. In comparison, the Liberal Party polled over 57% of the two-party-preferred vote amongst postal voters in 2019.
We should have expected a change in the composition of the postal vote – the total volume of votes has clearly increased substantially – less than 6000 postal votes were cast in 2019, compared to almost 5000 counted just last night, and up to 16,000 total (although it seems more likely that around 13,000 will end up being returned).
We know that Labor generated a lot of postal vote applications, whereas they generated practically none in 2019. And you would expect that a certain kind of voter who traditionally would have voted on election day but switched to postal voting due to the pandemic wouldn’t look like the typical postal voter.
So what’s left to count? We now have all the election-day ordinary votes and the pre-poll ordinary votes, although of course they will need to be checked and there will be some small changes in the totals. We also have about 4900 postal votes. There were 16,840 postal vote applications, although not all of those will turn into votes. My best guess is about 13,000 postal votes will be returned, which means there’s about 8,000 outstanding. There will also be a small batch of provisional votes, but they shouldn’t make much of a difference.
Labor’s Kristy McBain currently leads by 1648 votes. If you assume there are about 8000 remaining votes, Fiona Kotvojs would need over 60% of the remaining votes to win, which doesn’t seem plausible.
Finally, here is a map showing the two-party-preferred vote for each election-day booth, and can be toggled to show the swings instead.