The declining major party vote


I’ve been focused recently on the longer historical trends which play out beyond the individual election cycle, or even beyond a single decade of political history. So much of electoral analysis is focused on the swings between one election and the next, and looking at the back and forth that take place over the term of a government.

Most of the time, you can’t perform long-term analysis at a scale more granular than the state level, since electoral boundaries regularly change. I did a little bit of historical analysis at the state level after the 2019 election for my book chapter.

My new dataset comparing election results at the local level back to 2004 makes it possible to get closer to the ground.

In this post I wanted to examine the historical trend which has seen an increasing proportion of the electorate voting for someone other than Labor or the Coalition in federal elections. While there is a short-term back and forth between the two major parties, and the major party vote does sometimes increase, the inexorable long-term trend continues onwards, with a record low major party vote in the House in 2019 and in the Senate in 2016.

I’ll also use my electorate-level dataset to look at where the drop in the vote has been more dramatic.

First, let’s zoom out to the national level, and look at the House. Thanks to David Barry’s excellent historic results tools for this first chart.

Almost 98% of formal votes in the House were cast for Labor, Liberal or Country Party in 1951. The decline started in 1955 with the split within Labor that eventually led to the formation of the Democratic Labor Party, but the major party vote during the era of the 1949-1972 Coalition government never dropped below 89% (with 89.4% in 1958).

If you stopped your analysis in 1987, you could argue that a slight drop in the mid-50s had then stabilised, with the major party vote well over 90% at most elections. Over the next decade, there were individual elections with big drops in the major party vote, but no consistent trend.

This vote dropped to 82.9% in 1990, an election famous for a strong Democrats result and the emergence of state-based Greens parties. After the major party vote mostly recovered in 1993 and 1996, the 1998 election saw a new high for independents and minor parties that wouldn't be beaten until 2013.

The major party vote steadily climbed from 2001 to 2007. One Nation and the Democrats collapsed, and the Greens only filled part of that space.

Since 2010, the decline has been steady, from 85.5% in 2007 to just 74.8% in 2019. In that time we've seen the Greens vote peak in 2010, but not drop that much since. Meanwhile we've seen a growing right-wing minor party vote, not just concentrated in one party, and the emergence of a non-Greens left wing minor party vote, along with the growing independent movement that I expect to contribute to this trend continuing in 2022, at least in the House.

The major party vote in the Senate has always been lower than the House, but the trends are similar. The gap between the House and the Senate has been getting bigger in recent decades.

The trend was particularly different at the 1970 Senate-only election, which saw almost 20% vote for minor parties and independents.

The collapse in the major party vote between 2007 and 2016 was more dramatic, falling by over 15%, compared to less than 9% in the House. This coincided with the flourishing of preference harvesting leading up to the 2013 election, and the double dissolution in 2016 being a unique opportunity for smaller parties.

The major party vote slightly ticked up in the Senate in 2019, but it's still lower than in the House and I see no reason why this is a long term trend, as opposed to adjustments from a return to a half-Senate election without the performance enhancing drug of group voting tickets.

So now, let's zoom into the seat level. Unfortunately my Senate dataset only dates back to 2013, but I can compare the House primary vote back to 2004. This first map shows the change in the House vote by seat since 2004.

There's been quite a large drop in the House major party vote since 2004, so unsurprisingly there are very few seats where the House vote has gone up.

The biggest increases are in two seats that were represented by independents in 2004, but have since returned to the Nationals: New England and Calare. The major party vote dropped as low as 38.1% in 2007, but was 67.6% in 2019. Calare dropped to 53.2% in 2004, climbed back to 82.9% in 2013 but has since dropped back to 66.8%.

Two other seats have seen the major party vote increase by over 6%: Bennelong and Cunningham. They were both seats with significant campaigns from the Greens in 2004, with an incumbent Greens MP running in Cunningham and Andrew Wilkie focusing on John Howard in Bennelong, three years before Labor picked up the ball and won the seat. Neither seat ranks quite so highly on Greens target lists in the 2020s. The major party vote has also gone up in Sydney, a seat that has a relatively high Greens vote but was much more of a Greens priority in 2004 than it is today, despite two state Greens MPs representing the are.

Interestingly all ten seats with an increased major party vote are in NSW. The other five are all in Western Sydney.

At the other end of the spectrum, the four seats with the biggest drop are all now held by crossbenchers:

  • Clark (dropped 44.5%)
  • Warringah (34.7%)
  • Melbourne (33.6%)
  • Indi (32.1%)

Kennedy doesn't rank highly, since Bob Katter has represented the seat as a crossbencher since 2001. Interestingly, Mayo doesn't rank quite so highly, since it has a long history of being a strong seat for parties like the Greens.

Most of the other high-ranking seats tend to be in regional areas. Mallee, Blair, Farrer, Cowper, Capricornia, Wright and Hunter all have seen the major party vote drop from the 80s to the 60s.

The decline in urban seats is less dramatic in Sydney, Perth and Adelaide. The only Sydney seats to have seen major party declines of more than 10% have been Warringah and Wentworth, where strong independents ran in 2019.

There is an inner ring of seats around Adam Bandt's seat of Melbourne have seen the major party vote decline by 12-20% in seats like Cooper, Wills, Macnamara, Higgins and Kooyong. This reflects the Victorian Greens' relative improvement compared to their interstate parties, and the concentration of Greens efforts on inner city electorates. Most seats in south-east Queensland have seen bigger drops than in Sydney, Perth or Adelaide.

Now for the Senate. It would be good to be able to compare the same datapoints as we have for the House. The House trends are distorted by focused local electoral campaigns by the Greens and independents. Big votes for lower house independents is an important part of the story, but it doesn't give a good sense of the broad underlying trends.

Unfortunately I only have data since 2013 for the Senate, but it is interesting in another way. The 2013 and 2019 elections produced a similar major party vote in the Senate, but in quite different places.

The most obvious thing that jumps out is South Australia. The Nick Xenophon Group polled almost 25% in 2013. By 2019 the successor Centre Alliance polled an unimpressive 2.6%.

There were big drops in the major party vote across regional Queensland, with drops of over 10% north of Bundaberg or west of Brisbane. Even in the south-east there were no seats where the major party vote went up. One Nation polled strongly in Queensland in 2016 and 2019, but had not yet re-emerged in 2013.

Drops in the major party vote were consistent across most of Victoria and Tasmania, with the drops biggest in regional seats.

The Jacqui Lambie Network contributed to a particularly large drop in Braddon, her heartland. Lambie was elected in 2013 as a largely unknown Palmer candidate with 6.6%, compared to almost 9% in 2019.

It's interesting how the picture is much better for the major parties in New South Wales and Western Australia, with numerous urban seats producing stronger major party Senate votes in 2019, while the drops in rural areas were more modest than in Queensland, Victoria or Tasmania. I think this is mainly explained in New South Wales by the huge 9.5% vote for the Liberal Democrats, who benefited from first position on a very large ballot paper, leading to a lot of confused voters. Indeed the increased major party vote in Sydney was mostly concentrated in Liberal electorates in northern Sydney and the Sutherland Shire.

In Western Australia, I think it's partly explained by my dataset merging all Nationals candidates in with the Coalition. The WA Nationals, who run on their own ticket and have a very loose relationship with the federal Coalition, polled over 5% in 2013 and only 1.4% in 2019. They should probably not be included in the major party vote, but they are here.

This has been a very long blog post. I think there's even more interesting information if you dive into these maps. You could see other historical quirks showing up if you made maps comparing other years within this time period, but that's enough for me.

As for 2022? The Greens and One Nation don't appear to be set for any great advances, but they are also not set for a collapse. We don't know yet whether any big-budget Palmer campaign will net more votes in the aftermath of two years of COVID-19 responses. And the total number of minor parties is down after law changes raised barriers to entry, forcing some parties to merge and numerous others were deregistered. But I don't know if that will actually reduce the minor party vote, or just concentrate it amongst fewer players.

In the House, though, I expect to see a big increase in the independent vote, probably enough to produce a record low major party vote, worse than the 75% polled in 2019. Even if there's only a handful of independents who have a chance of winning, I think many of them are capable of polling in the high single digits, and there are a lot running.

Why does this all matter? I've spent over 1500 words analysing the trend without explaining why it should matter.

It is interesting because it says something about how polarised Australians are, and how attached they are to the parties that have monopolised government for over a century.

An increasing minor party and independent vote will put much more pressure on Australia's outdated electoral system. A single-member voting system doesn't do an amazing job of accurately translating votes into seats when most votes are cast for two major parties, but it gets even worse when that concentration breaks down.

Seventeen electorates were non-classic races (where the final count is not Labor vs Coalition) in 2016, and that number dropped slightly to 15 in 2019. A dropping major party vote will lead to more complex preference counts and more races that aren't a straight Labor vs Coalition fight. It may also lead to more marginal seats, if multiple fronts are opened up (hello to the Liberals fighting off climate independents in inner city seats).

At the 2016 election, more than two thirds of MPs were elected with less than 50% of the primary vote, for the first time ever, and that number went down even further in 2019.

There's nothing wrong with using preferences to elect MPs, it's the best method of electing a single representative. But electing your second-best choice is not the same as being represented by your first choice, and that awkward compromise will become less tenable if the major party primary vote were to drop even further.

That's it for now. Thanks for reading this long blog post, and enjoy the maps.

Liked it? Take a second to support the Tally Room on Patreon!


  1. Interesting … I have a database of polling and results back to 1983 which, among other things shows that the sum of the two majors Primaries has declined pretty steadily by an average of 1.8% per year and is now at 71% (was a tad lower for a while in 2017)

  2. Great post, but I totally disagree with your second last paragraph. Being represented by your second, third or fourth order candidate is not a “bug” with our system of democracy. It’s a “feature”.

    In fact it’s such a GOOD feature that it should somehow be enshrined in the constitution to prevent our cynical major party operators reverting back to first past the post, two party system (witness how well that works in the USA 😟)

  3. Being represented by a lower preference representative is a minor consolation, far from ideal. It wouldn’t be necessary if we didn’t have a single-member system. The fact that there are worse systems does not mean it is good.

  4. Well is it any surprise the Coalition votes are declining. Being in a Liberal Blue ribbon electorate, the current Liberal advertising is an embarrassment.
    Not one mention of policy, totally undermining the opposition. The parties need to understand that an election is not all about the election, it is also about growing their membership.
    This shocking advertising approach is not doing the Libs any favors.
    My request is remove the current ad !!!

  5. There is nothing worse than living in an electorate where the Labor party take the safe Labor seat for granted and the LNP have decided not to contest the seat.
    This added insult will now play a part when the NSW state election comes around next year and the LNP expectedly will grovel for a vote in a now marginal seat and will find wanting. Is it any wonder the major parties are losing the trust and votes of the people.

  6. As commented by Paulus, excellent post but disagree with the 2nd last paragraph.
    The person elected is the least ‘dis-liked’, which seems to me to be the fairest outcome in a this situation.
    Our compulsory preference voting system over-seen by an independent body is not perfect, but so much better than what I have have mostly seen reported elsewhere.
    If you want it simple it may not be fair, if you want it fair it may not be simple.

  7. “The person elected is the least ‘dis-liked’, which seems to me to be the fairest outcome in a this situation.”

    Eh, not quite. If we wanted that, then you just elect whoever has the fewest last preferences!

    We elect the candidate who would win a sequence of run-off elections, assuming no voter changed their preference order between run-offs and everyone who voted in one voted in all, etc.

    (In real run-offs, some do change order, e.g. there a small group of people who voted for Macron on the 10th and for Le Pen on the 24th and vice-versa).

  8. Technically, we vote to elect a candidate in our electorate’s not to elect parties.


Please enter your comment!
Please enter your name here