In yesterday’s blog post, I looked at how many numbers voters tend to fill out on their ballot paper. Today I’m going to look at how those preferences flow to the major parties.
The term “two-party-preferred” (2PP) refers to preferences distributed between Labor and the Coalition. This can be the same as the two-candidate-preferred, but the 2CP could instead include an independent or minor party candidate. At the 2019 election, the 2PP and the 2CP were the same in 74 out of 93 seats (the “classic” seats).
The main statistic in this piece I’m looking at is the flow of preferences. For those votes cast for candidates other than Labor or Coalition, we calculate how many of them flowed to Labor, the Coalition or neither at the point of the 2PP.
The AEC has been publishing two-party-preferred preference flows per candidate and per party at federal elections for quite a long time, but it’s only been possible to do in NSW since 2015 (when Antony Green did a blog post on the topic). I don’t believe this is available for most other state elections.
Before we get to the individual parties, I wanted to modify a chart from last week’s blog post. This chart shows the proportion of preferences that flowed to Labor, the Coalition or exhausted at each election since 1988.
The previous chart showed the flows as a share of total formal vote, whereas this chart shows them as a share of the preference pool, and thus add up to 100%. There's also a dotted line for 2007-2019 which shows the same statistic if you include the 2PP for the non-classic seats.
The uptick in preference flows to Labor in 2019 appears to largely be a consequence of the total pool of preferences increasing, as the number of exhausted votes went up. As a proportion of the pool, the increase is much smaller. And it's even less if you include non-classic seats. Labor tended to gain slightly stronger preference flows in non-classic seats in 2011 and 2015, but did much worse in those areas in 2007.
Overall preference flows to Labor dropped to their worst ever level in 2011, but have since rebounded in 2015 and 2019 to the highest levels since 1995.
Now that we've seen the overall flows, next up I'm going to break the 2019 data down by the party of first preference.
The Greens stand out clearly, with a majority of their votes flowing to the ALP. But Keep Sydney Open and Animal Justice also have a sizeable share of their votes flowing to the ALP. It's also worth bearing in mind that the Greens make up a large part of this preference pool - about 38% of the pool in 2019. If you remove the Greens preferences, the remaining preferences only slightly favour Labor - 21% to 17.9%.
Generally preference flows from the right wing minor parties to the Coalition are weaker than the left-wing minor parties, although the late Christian Democratic Party had a similar flow to the AJP and KSO. One Nation, had much weaker flows, and Shooters preferences were almost even between the major parties.
These two charts make it clear that Labor does have an advantage against the Coalition when it comes to receiving preferences, but how much is that advantage muted due to optional preferential voting? After all, there's that big grey area in the middle that doesn't go to anyone, while primary votes continue to have the same weight.
For this next chart, I've compared preference flows overall and for select parties at the last two NSW state elections and the last three federal elections.
The ratio of Labor to Coalition preferences for each party is usually similar for each party, but the exhaust rate usually reduces the total volume of preferences to about half of what it would be at a federal election, although for some parties the discount rate is worse - One Nation voters are more likely to exhaust.
This exhaust rate effectively reduces the value of preferences and makes primary votes more valuable in winning a seat, compared to the system used for the House of Representatives. But that might be a topic for another day.
Thanks to Antony Green for sharing the tables he had previously calculated using this data from the 2015 election which helped inspire what tables to extract from the dataset.
If you're interested in analysing this data yourself, I have added spreadsheets with the 2PP and 2CP preference flows per candidate at the seat and booth level to the 2015 and 2019 data repository folders. The 2019 folder is free for everyone to access while 2015 is for Patreon donors.
I found that particularly insightful, especially for the One Nation and Christian Democrat parties where I assumed the vast majority of preferences would flow to the Coalition. It seems that given the choice it is a case a pox on both Labor and Liberals as far as One Nation and Christian Democrat parties go.
Imagine if OPV was extended to the Federal elections. Both major parties will be hard pressed to state that they were popularly elected.
One of the things that could change the numbers in this Election is the vote for Independents, you have the seats of Orange and Barwon, the numbers show they should retain those seats, this may lead to more Independents being Voted in, this preferences will be very important.
Yes Neil. If you are a labor candidate then in a tight seat you always want a one nation candidate in there.
@roger the problem with that is that one nation votes often split at 60-40
If the exhaustion rate continues to increase, might there be an argument to just switch from OPV to FPTP?
With plurality seeming to become the most likely outcome in the next decade. Along with the complete drop in primary vote for the Major parties.
Would seem like a better test of who can stand on their feet electorally rather than the golf handicap via preferences, we give to candidates from major parties to make it seem like they actually performed well
Strathman scoop, I think the Senate or ACT/Tasmania method of using some form of STV and multi member districts with optional preferences could work. It would encourage the major parties to broaden their base, as with Tasmania and ACT you could see members being defeated by another candidate from the same party.
Yoh An- I’m not a fan of MMP:Multi-Member districts because I think it allows for politics to become to centralised and the fact that voters wouldn’t have a direct line to their local MP.
It becomes easier for politicians to blame each other for local issues, knowing they won’t face any electoral consequence or indeed remain unaccountable during their period of governance. Also with MMP and shared electorates the geography increases, which makes local issues and concerns even harder to address.
ACT is probably not the best example to as it is both a city council and a gelded state government.
From my understanding of Tasmanian parliament they have single electorates just in reverse (with these being in the Upper House).
MMP/Multi-Member may work in smaller areas like councils or tiny states/territories like Tasmania and ACT but it would be basket case if was implemented in a state like NSW.
One thing I am sure of onp will overwhelmingly take votes off the coalition but they cannot direct them back solidly. They are the nutters of the political landscape full of the “paulines” and the disaffected who believe in conspiracies abound.. Labor and the unions are a pink woke plot. Their other side are the “turncoats” who have the Agenda of power…… pity anyone who.gets in their way
Yoh An,
I tend to think that any form of PR tends to see parties shrink, not broaden, their base. Not the parties fault, but once you have PR it is far easier for bits to break off which don’t necessarily come back. The Greens (however erroneous I think this actually is) breaking from the ALP here is an example, as I think is the rise of the Teals, who are Libs mostly animated by opposition to the Nats.
OPV is nothing like first past the post. This blog post demonstrates that some voters choose to take advantage, and others don’t. In other words, it’s optional. I don’t understand how some people choosing to not use preferences mean we should take them away from everyone.
But yes PR would be superior. Then voters could be directly represented in parliament and those representatives can then do the coalition-building we currently leave up to the voter when numbering their preferences.
Probably agree Labor voter, although I’m not sure what the best solution is. Trying to keep parties large but having multiple factions (as occurs not just in Australia but also with the 2 major US parties and to some extent with UK conservatives) can be problematic as you then face difficulty trying to unite warring factions.
Either way, I think having a large party split up its base into smaller parties may be beneficial because it then encourages the parties to engage in dialogue/collaboration which may not work out as well under a factional based system.
The problem of party system and electoral system interaction has largely been answered. The number of seats in the assembly multiplied by the average magnitude of each district produces a seat product – the bigger the seat product, the more parties tend to get elected to parliament. So yes PR would likely lead to more parties being represented in parliament and less dominant parties but there are many types of PR. Low magnitude PR would likely be much less fragmented then pure PR as in the Netherlands or Israel.
@mick we think the same about the greens. just a bunch of communists
@ mostly labor actually if you look at the votes they come from both sides and vary from whihc party but they predominately take votes from labor greens and minor parties on the left
I guess the question then becomes under PR more parties will be represented but this does not mean local issues addressed, as they would be more wedded to what come them elected; ideology.
Whereas the current system people can dissent in major parties, for the most part and it is in the interest of major parties to run electorate by electorate campaigns to win votes and listen to community concerns.
A PR Parliament risks politicians speaking in generalities, ambiguous policy making vs what we have now which is targeted spending and regional connective infrastructure planning (like what has happened in NSW).
I don’t get the fascination with PR, as essentially you’d be duplicating the Upper House/Senate.
Are people people who are advocating for PR also advocating for abolishing the Upper House/Senate?
Yes, I would say under a PR system you wouldn’t need an upper house. Strathman scoop, I will issue a partial rebuttal to your claim about local issues being addressed. With State/Federal governments, most of their priorities are strategic in nature and local issues are generally confined to the council level.
I think where PR systems work best in many European countries, they distinguish better between regional/provincial levels of government and the national level. I think Australia could well move to that tier system, abolish the large states and have the country still at 3 levels (National, regional/provincial and local). I think in some countries like Italy they have 4 tiers of government (local councils, provincial/regional government, state government and the National/federal government).
I think we have it right here Yoh An, AV in the lower house and PR in the house of review.
Having smaller parties negotiate bills through the Senate is a better option than having those same parties extract their pound of flesh to prop up a Government. In the first instance, you do get genuine negotiation to get bills through. In the latter, the Greens or Nats might try to get a policy they have taken to the election as their price at the expense of a policy the main party took to the election. That is ripe for losing trust in the process and, eventually, MAGA style populists. For a good example see the Carbon Tax’. Whatever the merits of the policy or the disingenuousness of the opposition, it had a huge impact on the trust the public had in the Government.
I would like to see OPV at Federal elections and everywhere. As I get older and more cynical I want the choice to deprive the undeserving of my preferences. We have it in upper houses – why not lower houses as well.
To be honest these complaints about PR are nonsense. Almost every PR system has local representation. The level of locality is always arbitrary but it still exists, and it ensures that different concerns have local representation, whereas SMD tends to produce sweeps of a region by one party.
I’m also hearing a lot of complaints about PR that apply a lot more to non-PR systems! The MAGA movement didn’t emerge under PR! And there’s plenty of disgruntlement that comes from lack of representation under the current system – possibly having something to do with a government that only represents 33% of voters!
I don’t support abolishing upper houses, I think they have a distinct role they could maintain if you designed the voting systems well. I’d go for a proportional system with a relatively low magnitude. You’d probably only have 4-5 parties in the lower house and tend to have governments of two parties, maybe 3 at the most. Then you have a super-proportional upper house which allows in smaller parties. The overall political balance of the upper house would ideally be in line with the lower house (ie if there’s a Labor-Greens government in the lower house, there’s an upper house majority who they can cooperate with), but outside of exceptional circumstances wouldn’t be controlled by the lower house.
Through this balance you can have the separation of powers that is the best part of a presidential system without concentrating executive power in one elected official. Steffen Ganghof has a book on this theory which he calls ‘semi-parliamentarism’ and I’m going to organise a podcast on the topic after the NSW election.
But the other reason I don’t argue for upper house abolition is because I don’t think it’s a practical possibility and I don’t think “well the upper house has PR” is a good reason to not use a better voting system for the more powerful and more important house.
FPTP is shit for one very simple reason: it punishes voters for having choices. As usual, a great example for this is the Swan by-election of 1918, which precipitated preferential voting at the federal level.
I would be happy to see savings provisions for federal “full preferential” voting that amounted to optional preferential in practice, provided that it remained illegal to advocate filling out an incomplete ballot. Just Vote 1 campaigns can get in the bin. (That’s assuming single-winner districts though; under a proportional system preferences matter less.)
More and more evidence accrues that Labor owes its electoral success to Greens voters more than the Coalition ever will to minor party voters on the right. Yet despite the increasing reliance on those preference flows, the ALP continues to drift rightwards on policy and coordinate a deliberate campaign of escalating antagonism against the Greens and the very voters it relies on to stay in power. Obviously the hope on the ALP is to cut the balls off the Greens and signal to its voters of the futility of left-wing dissent. But it’s even more obviously an incredibly risky strategy going forth. The ALP can’t afford for Greens voters to arrive at the (correct) perception of a decreasing amount of policy difference between the major parties.
@redistributed if you dont want them getting i t dont bother voting
Those independents were all voted in on the sff ticket though. Independents only do well wen they are well backed and have high profile and the current party is in disfavor. Look what happened in vic the nats clean sweeper the sitting independats
Fugitive,
I suspect the exact opposite is true. The ALP need to distance themselves from the Greens rather than get closer. I posit that for every vote they lose to the Greens by going ‘right’ they save one going to ON (or equivalent). And vice versa of course. But if ON are sending prefs 60-40 to the Coalition but the Greens send prefs 80-20 to the ALP, then it makes obvious sense to hold the right and let the left go.
The other point is quite obvious as an old Labor man. From a socioeconomic point of view, the Greens are the exact class of people the ALP was formed to oppose, so getting closer to them risks gutting the true voter base of the party.
To put it more bluntly – if the ALP get closer to the Greens they gain no primary votes and no preferences, at the risk of losing votes and preferences to the ‘right’. Once that happens, the party is in a death spiral it can’t get out of. Holding voters on the right of the party holds no such risk, unless the Greens start preferencing the Libs, but in that case they are likely to pick up some of the Libs voters, some of the ON/UAP prefs, and/or the Nats become more neutral.
Agree Labor voter, that is what the Liberals also need to do despite forming a close bond with their junior Coalition partner. Having distinct brands enables better separation and allows the party to enhance their true base.
@labor the problem with that is they lose votes to the greens and over time seats. labor going right is the same as liberal going left it will only cost them more and more votes from their base. and that hasnt exactly worked for the liberals has it
I am no supporter of the greens but they are the lesser of the evils. Onp are nuts but they are getting support from the “pissed off “. Labors may forward is to recognise economic factors will shift energy use away from oil coal gas. Ensure the damage done by such technologies is reduced. Including any troubles with both lithium and and fracking. Then realise economic activity and jobs are part of the fossil fuel industry and manage the inevitable transition along the necessary time frame
@Ben & @mick – I think the opposite.
Ben, the ALP getting closer to the Greens costs them votes. I doubt there are many more votes to lose to the Greens, but a lot at the other end. And those lost to the Greens come back via preferences, those at the other end no not. That is the key.
Mick, that is the problem the ALP will have. The people who are going to hurt if we stuff up the energy transition (and I think we are) are the exact people the ALP are supposed to support.
Comments are closed.