The Australian Electoral Commission released the final data from the Eden-Monaro by-election earlier this week. This data breaks down the two-party-preferred preference flows between Labor and Liberal based on the primary vote of the voter. This data allows us to look at how voters for various minor parties and independents marked their preferences.
This data is not usually published for state elections, forcing us to rely on the distribution of preferences, which doesn’t distinguish between primary votes for a candidate and votes they have picked up along the way. The AEC in recent years has even begun publishing these statistics at the level of the polling place.
Antony Green and Kevin Bonham have both published posts analysing the preference flows at the top level, but I thought it’d be interesting to look at how the flows for the three biggest minor parties varied across the electorate.
The Nationals’ Trevor Hicks came third with 6.4%, with the Greens’ Cathy Griff close behind on 5.7% and Shooters candidate Matthew Stadtmiller in fifth place on 5.3%. Almost 26% of the primary vote was cast for candidates other than the Labor and Liberal candidates, and about two thirds of that vote was for one of these three candidates.
The Nationals preferences strongly favoured the Liberal candidate, although not as strongly as they did in 2019. Shooters preferences slightly favoured Labor, while Greens preferences overwhelmingly favoured Labor.
The following table breaks down the results based on the geography of the seat, using the same breakdown i used in my by-election guide. For my post-election summary I also divided up the numerous pre-poll centres between these same regions. For this table I have lumped in pre-poll booths with the surrounding election-day booths. “Other” refers to postal votes, pre-poll votes cast in Canberra and other special vote categories.
The first three columns refer to what proportion of each party’s primary vote flowed to Labor as a preference ahead of the Liberal candidate. The last three columns refer to how many more votes flowed to Labor than Liberal.
|Voter group||SFF flow||GRN flow||NAT flow||SFF margin||GRN margin||NAT margin|
For context, Labor won Eden-Monaro by 735 votes. If 368 voters changed their mind and preferenced the Liberal candidate ahead of Labor, that would have changed the result. With such a close result, many factors could have flipped the result.
You do notice some differences in how preferences flow in different parts of the seat, although they are never too dramatic. The Shooters preferences only range from 54% in the north to 60% in the west (excluding the other votes – the Shooters preferences favoured the Liberals in the postal vote and in the Canberra pre-poll vote).
The Greens preferences did weaken as you move deeper into the country, with just 85% of Greens preferences favouring Labor when you cross to the west of the great dividing range. This could reflect the Greens’ ability to hand out how-to-vote cards in this area. This compares to 94% on the coast. It’s worth noting that the Greens vote was very low in the west and south, so an increase in the preference flow to 90% would have netted Labor just 21 more votes.
The Nationals preferences to the Liberal candidate tended to be around 75%, ranging from 74.4% in Queanbeyan to 77% in the north. The one exception was the south (the snowy mountains, Cooma and surrounding areas) where they flowed at 86.1%. If preferences in the rest of the seat flowed at this rate, that would have netted an extra 482 votes for the Liberal candidate, which would have been decisive.
Finally, here is my map showing preference flows for these three parties across the ordinary election day booths. I’ve defaulted to the Shooters – you’ll see that most booths favour Labor but some small booths do favour the Liberals. You can toggle to see the Nationals and Greens maps.