Considered thoughts on the South West Coast State election 2018, one month out

This is a long piece, but contains quite a bit of detail that - I think - rewards close reading.

If you’re interested in what might happen in the coming election in South West Coast (my old seat), persist with this piece and engage with the ideas and logic behind it.

I’m just going to come out and say it – Roma Britnell will win the seat of South West Coast at the coming election. That might sound  a bit ‘tell us what we don’t know’, but the entry of Vote1-Local Jobs Party candidate James Purcell into the race has put a different complexion to the usual on this election. Here’s why and how.

First, the obvious. The Wentworth by-election result has reinforced a prevailing notion that voters are dissatisfied with the major parties and are looking for alternatives. The South West Coast electorate is  more conservative and rusted-on than the voters of Wentworth (believe it or not); less given to change and less attracted to things they haven’t had the opportunity to become comfortable with over many years. But there could be a segment of the voting population in the South West who are ready to try something different, but most opt for safe and steady. To be fair, James Purcell is not that different. He is conservative by nature, but is not from the major parties and, to that degree, is independent. He’s also served a term in parliament and has done so in a very public way, so he has a proven track record to point to. So, he could present as a safe alternative. And, given Trump, Brexit, Wentworth, circumstances might be aligning to allow him to step forward as an electable alternative.

Of course, he’s been elected before – at least twice. He served on the Moyne Council, including as Mayor.  More tellingly, he won the fifth Upper House position in Western Victoria in 2014, with his new party polling only 1.3% of the primary vote in a very populous, diverse electorate that takes in Geelong and Ballarat as well as south west Victoria. He scored 5501 personal first-preference votes though, so we can assume that he has a solid personal base – largely in his home country of the south-west – of 5000 votes; a very good start. He also ran for South West Coast in 2010, as an independent, gathering an impressive 11.53% of the primary vote. More about that later.

One key factor is that Roma Britnell isn’t Denis Napthine. In her 2015 by-election win, the Liberal primary vote dropped by 7748 votes (17%) against Denis Napthine’s 2014 result. This may have been because there were more – and more conservative – candidates running in the by-election, splintering the traditional Liberal vote. However, the numbers don’t necessarily support this simple explanation because my primary vote dropped in that by-election too. I’ll cover that in a bit of detail shortly. Ahead of that, let’s just look at Port Fairy as an example of the Napthine factor. In 2014, Napthine received 200 more two-candidate preferred votes than Britnell did in Port Fairy in 2015. That doesn’t sound like a lot, but for one smallish booth, that is significant – 20% of his vote in that booth. And in a close election, 200 votes can make all the difference.

Let’s look at the 2015 by-election result in a bit of detail. Denis Napthine had resigned after two decades as a parliamentary representative for the region, in the seats of Portland and South West Coast , and sometime Opposition Leader and Premier. It was my fifth attempt as a perennial candidate, after 1999, 2002, 2006 and 2014. I wasn’t a candidate in 2010. Looking at the results now, that by-election can be characterised as the election of ‘The Liberals without Napthine and Roy Reekie without Labor’. The ALP decided not to field a candidate for the by-election and I mounted a campaign as an independent. So, how did the numbers fall? As previously noted, Britnell’s primary vote was 7748 down on Napthine’s from a year earlier.  My primary vote was 4495 down on the Labor vote from 2014. At 16.82%, my first preference percentage was down 14% against the ALP average slice of all enrolled votes for the seat of 30.5%. So, where did those 2014 Liberal and Labor votes go? Where were they – or more pertinently, where are they – hiding?

Analysing voting records is an inexact science, closer to astrology than astronomy. That said, my assessment of the fate of those votes is this. Britnell’s lost votes were swallowed by a Nationals vote for Michael Neoh (5581), some Portland Liberals voting for Pete Smith (695) and a 40% slice of Jimmy Doukas’ vote (1586). My vote (and the 1010 lost from the Greens Party’s vote between 2014 and 2015) disappeared to Swampy Marsh (1137), The Animal Justice Party (490), informals being up by 2% (826) – especially in Portland where the informal vote was almost 10% - and the Did-Not-Vote element being up 2.5% (1133).

So, what does this tell us? If you look at my 2014 and 2015 first preference votes against my 2006 first preference result and the ALP’s 2010 first preference results, it could reveal my personal base vote to have been somewhere between 6 and 10% and the ALP’s rusted-on-if-revived-from-slumber core vote in the seat to be about 25%, with a general election average for Labor to be 30.5% of the primary vote. It also might tell us that Britnell’s – or more properly, the Liberals’ – core primary vote is 35-40%, depending on how the wind is blowing. It also suggests that there are – if not swinging, then – wavering voters who moved to personalities like Jimmy Doukas and Swampy Marsh in the by-election and that there is a natural National vote looking for somewhere to go (Gerald Madden – another former Moyne mayor – in 2002, with 11.5% being a case in point).

I promised to come back to Purcell’s 11.63% in 2010. This is where it really gets interesting. Purcell came out of the blue in that election, as much as a sitting mayor can. There was no Nationals candidate. So, did Purcell just pick up Madden’s 11% from 2002? I don’t think so. The Nationals ran in 2006 but only polled at 4.5%, so there was 7% of their vote from 2002 hiding somewhere. The 2006 and 2010 figures suggest that those voters chose Napthine. They appeared to have stayed with him in 2014. In 2015, Neoh scored 14.5% for the Nats. Many of those could have been his personal following in Warrnambool as a high-profile mayor but many others could be Nationals looking for anywhere else than the Liberal Party as soon as they had an excuse and opportunity to jump ship. So, where did Purcell’s 2010 votes come from? He looks like a conservative/Liberal, sounds like a conservative/Liberal, comes from what appears to be a conservative/Liberal background. One would assume therefore that they came out of the Liberal vote. Wrong. A booth-by-booth trend analysis suggests that his 2010 vote largely came from the former ALP vote. His big booths – which all show corresponding drops in the Labor vote – were Warrnambool Temperance Hall, Koroit, Warrnambool North, Warrnambool Tower Square, Warrnambool East and – wait for it – Port Fairy. He also did well proportionately in all three Portland booths, ripping votes off Labor. It gets very difficult to analyse where his votes then went when he didn’t run in the Lower House seat in 2014 because pre-poll voting really became a thing in that election and that means that the booth-by-booth figures are down across the board but most noticeably in the bigger, urban booths in both main population centres. However, it does appear that Purcell’s 11% of votes did not all return to Labor in 2014; some ended up with Napthine. Lessons from this? First, Purcell’s 11.63% result was close to the 5000 he yielded in Upper House first preference votes in 2014, so it’s looking like a very solid voting base for him. Secondly, as suggested in my by-election analysis above, there is a segment of the electorate – some or many of them traditionally National voters, and most of them situated in Moyne – who are attracted to high-profile, noisy candidates and are prepared to go there with their vote.

What then can we conclude from close analysis of historical voting in South West Coast and general gazing into our crystal ball for the coming election?

Port Fairy – and maybe Nationals voters – like shiny things and will go with a high profile candidate.

Labor needs to focus on Portland. There’s a did-not-vote and informal vote there from 2015 to recover and a lot of votes to be gained if the ALP runs a visibly Labor campaign.

James Purcell needs to focus on Port Fairy, Koroit and Warrnambool, except Warrnambool North which – on recent figures – holds up very strongly indeed for the Liberals. It’s becoming a fortress up there.

Pre-polling is very, very important. More people are using that option. Unfortunately for politics tragic like myself, it means that you can’t see the demographics of who is voting which way because voters from everywhere are mixed into one super-booth called pre-poll. Purcell is going to need to spend a lot of time standing outside shopfront polling stations in Warrnambool and Portland over the next four weeks. He needs to dial up his profile there.

There are three missing links in all of this. One, the results of the Seat of Warrnambool election immediately after Adam Kempton’s  single term as a local member after winning the post-Ian Smith by election over Labor and John McGrath. McGrath came back in the general election and beat Kempton using Labor preferences. With a solid base of 5000 votes, a high profile track record, the right conditions for a usurper to win and a good campaign that presses the right buttons, Purcell is a reasonable chance to run second on primaries and then – depending on the level of Labor’s vote and how their preferences flow – come over the top of Britnell unless she can win more than 41% of the primary vote. It could be McGrath all over again.

Secondly, who else is running?

It seems that Michael Neoh is running as an independent, rather than as a Liberal Party candidate, National Party candidate or Natural Law Party candidate. Neoh has a high profile in the Warrnambool half of the electorate and polled strongly there in the by-election. Coincidentally, he did well in places James Purcell polled well in at the 2010 ballot. It’s hard to assess his core vote because he ran under the Nationals banner last time, but it’s safe enough to shave about 4% of his vote to allow for true blue Nationals who might prefer Britnell, Purcell or – the next candidate for examination – Jimmy Doukas, leaving him with a maximum of 10%; enough to run fourth at best. Where his 3 or 4 thousand votes then go is anyone’s guess, but I’m guessing that they’d go to Purcell and then to Britnell, not to the ALP.

Then there is Jimmy Doukas; the dark horse who presents the Doukas Dilemma. Jimmy came off a standing start (albeit as a high profile Moyne councillor) in the 2015 by-election to pull 10% of the vote for the Country Alliance, a relative new player in South West Coast. We can safely take that full 10% as all Jimmy. The dilemma – principally for Purcell – is that Doukas’ best result was in Koroit. He also performed well in Warrnambool North (Neoh/Liberal country) and Port Fairy (Napthine/Purcell/Shiny Objects territory) – all of which are Purcell heartland based on the 2010 results – and in Allansford where Purcell did not score well in 2010. The Doukas Dilemma is doubled when you look booth-by-booth at 14 booths where Doukas and/or Neoh did well in 2015. If you compare their votes against Purcell’s votes in those booths in 2010, it’s clear that there is not room for three independentish, high-profile candidates in those booths. Someone has to bomb badly or that swinging, protest vote will splinter. And for Purcell to win the seat, that vote is going to have hold – and improve – for him while the Doukas and Neoh votes dramatically dive. How likely is this? A lot will depend on how well they all campaigns. Even at that, Doukas and Neoh would have to run dead or do something outrageously stupid to lose big chunks of their vote after only 3 years. As noted earlier, an improved Purcell vote in those booths is most likely to steal votes from Labor.

The other analysis of the Doubled Doukas Dilemma for Purcell is based on a simple sample dip of four bell-weather booths – Koroit, Warrnambool (Temperance Hall), Warrnambool East and Warrnambool North - across the last five general elections to determine the average ALP and Liberal primary votes and therefore the average undecided or swinging vote level. I could bore you with the booth-by-booth and election-by-election breakdown, but the overall averages for those four seats collectively are ALP 33%, Liberal 42% and ‘available’ 20% (with 5% informal). That’s 20% to be divided between Purcell, Doukas, Neoh, the Greens and any other party or candidate that shows up. The problem for Purcell is that this isn’t likely to be enough for him to garner enough primary votes to run into second at the end of primary voting, given that these four booths are four of his strongest based on the available evidence. He needs to blow both Doukas and Neoh out of the water to have any real chance.

If – I repeat, if – he makes it into second, Labor preferences will be vital. They will need to flow solidly his way. Then, depending on how large the Doukas and Neoh vote blocks are, he’ll need at least a third of those as preferences.

Finally, and maybe crucially, how well will the Greens Party do? If Thomas Campbell runs for them and if there’s any sense or justice in politics (which there isn’t), he should – in all fairness – win 50.1% on the primary vote; he’s that good a candidate: poised, accomplished, professional, talented. However, in the real world in this election, he may well hold his first preference percentage or even increase it a little to maybe 10% or more. In that case, where do his preferences go in a two horse Britnell-Purcell second round sprint? This is interesting, because Purcell is closely aligned to the horse-racing industry and has been vocal on the beach training issue. Britnell has largely kept her mouth shut over it. Gaston (ALP) has alternated between tripping over her own feet and placing one or the other in her mouth about it. The Greens have been solid with BCRAG throughout. So, a first preference Greens voter goes where with their preference vote – the horsey Purcell or the hoarse Britnell? If Campbell’s vote is 10%+, his preferences may well save it for Britnell and the Liberals in a run-off with Purcell.

If Purcell fails to make it to second, I can’t see all of his votes – or many of them, for that matter - flowing to Labor, and the Doukas and Neoh  votes are most likely to flow to the conservative side. Greens votes may well flow to Labor, but not in large enough numbers to get Gaston up. In a Britnell/Gaston run-off scenario, Britnell should canter in.

I stuck my neck out at the start of this piece. At the end of this first lap around the course, what’s my prediction now – or more relevantly, what’s yours? With Doukas and Neoh in the race, mine is still a Britnell victory. There’s still a lap to go in November, but from here she has a pre-existing big lead and the advantage because Purcell’s run may well be blocked by Doukas and Neoh. Watch for an advertising blitz from the Libs in the final weeks of the campaign. Still, there is a lap to go and with pre-polling ...? It’ll be interesting to watch. Don’t put down your glasses.