The Rest Are Preferences

In advocating for a new voting system in the UK, Alastair Campbell and Rory Stewart may be overlooking Australia's preferential voting system

A sample ballot for the Australian House of Representatives

Regular listeners to the The Rest is Politics podcast with Alastair Campbell and Rory Stewart will be aware that barely a week goes by without the pair advocating for a change in the United Kingdom’s voting system. This is understandable, as First-Past-The Post (FPTP) is a terrible system. One that consistently produces seat-counts that are misaligned to the percentage of votes received. 

At the 2019 General Election the Conservative Party won 56.15% of the seats in the House of Commons from 43.6% of the vote. Which isn’t great, but not as extreme as FPTP can be. In India, at their last election the BJP won 55.8% of the seats in the Lok Sabha with only 37.36% of the vote. The sheer number of parties in India means the bar to achieve plurality of the vote in each seat (rather than a majority) is very low.

However, in their advocacy Campbell and Stewart have been promoting Proportional Representation (PR) – a system that allocates seats based precisely on the percentage of the vote a party receives. On the surface, this seems like a good idea – PR is technically the most democratic system – yet there are broader issues to consider, in particular how PR would work with Westminster parliamentary system. 

Australia is often looked towards for its system of compulsory voting. This is something that both Campbell and Stewart have also become advocates for. This doesn’t mean we don’t have bad politicians, we have plenty of those. Nor does it mean extremism doesn’t exist – as it does around the fringes, often able to gain seats in our parliaments (federal and state). But it does mean that our largest parties are incentivised to speak to the broader public who are mostly concerned with kitchen table issues, rather than those with high ideological motivations. It also instils a sense of civic responsibility in the public. 

However, alongside compulsory voting, Australia’s system of preferential voting (what is also called Ranked Choice Voting or the Alternative Vote) also plays a stabilising role. Preferential voting also happens to be more democratic than FPTP, and aligns far better with a Westminster parliamentary system than PR.  It seems odd to me that Campbell and Stewart are not advocating for it instead of PR. 

Part of the reason for this is probably that in 2011 the UK held a referendum on changing the voting system from FPTP to what they called the Alternative Vote. Campbell and Stewart may think that there’s no point trying to advance this system again.

This referendum was part of a coalition agreement between the Conservative Party and the Liberal Democrats. The referendum failed easily, as the Conservatives campaigned against it, the Labour Party took no position. Both parties do very well out of FPTP. 

However, you can see why the Liberal Democrats pushed for voting system reform as part of any coalition agreement. FPTP has been a system that has been most starkly to their detriment. Prior to the 2015 election – at which the party was punished for forming the coalition with the Conservatives – the Liberal Democrats were a major party in terms of votes, just not in terms of seats.  

The 1983 UK General Election was the most egregious example of this. Although this was prior to the Liberal Party and the Social Democratic Party formally merging, their alliance at the time effectively made them a single force during the election. The alliance was able to secure 25.4% of the vote, but for just 23 seats.  Labour secured only 2.2% more votes, but won 209 seats. 

From the mid-1990s up until the 2010 election, the Liberal Democrats got much better at targeting seats that they could potentially win, and were able to increase the number of MPs they had. But even with better targeting of seats – winning 62 at the 2005 election – the fickle nature of FPTP still created a large disconnect between the party’s share of the vote and its number of seats. At the 2010 General Election, the party increased its vote share by 1% – up to 23% – but lost five seats.

However, political parties are highly self-interested entities, so changing the voting system to help the Lib-Dems is not going to be an argument to convince the Tories or Labour. The Lib-Dems are merely the most obvious example in the UK of how the FPTP system produces less democratic outcomes.

So what can preferential voting offer the country as a whole? And why is it a better choice for the UK to PR?

The democratic advantage of preferential voting is that it builds a consensus of each individual seat. By ranking each candidate a conversation of ideas takes place in the counting. No-one can be declared the winner of the seat until a majority of people decide that they prefer one candidate to all others. At the 2019 UK General Election, 229 seats were won by candidates securing less than 50% of the vote. These MPs lack a majority consensus.

A look to Scotland demonstrates why consensus is important. In 2019, the separatist Scottish National Party won 38 of its 48 seats with less than 50% of the vote. In East Dunbartonshire the party won with only 37.1% of the vote. Even with the addition of the Scottish Greens (also separatist) the vote for separatist parties was only 38.8%. A preferential voting system would likely see the parties in favour of a united UK preference each other ahead of the SNP.

Separatism is obviously an extreme example, but it’s easy to see how other big issues could be supported by one candidate and opposed by all others, and how FPTP can create victories for issues without majority support. Consensus is what mature politics should be about. Therefore building it into the electoral system is a great positive.

Unlike in FPTP, in a preferential system there are no wasted votes. Even if the voter’s preferred candidate is unlikely to win, their ranking of their ballot contributes to the overall result. Alongside this, the way preferences flow creates a more substantive picture of what matters to the electorate. It demonstrates politics as an amalgamation of different ideas, values and interests, rather than a zero-sum contest.

In our current era of ideological flux, following preference flows can also provide a much more sophisticated understanding of how voters think. Most voters don’t categorise parties, candidates or issues into “left” and “right” the way the political class does. Dare I say that people without Arts degrees are often able to see issues more clearly as they’re not working with preconceived boxes.

You only need to see how parties that claim to speak for the working class are increasingly held in suspicion by the working class to recognise that something is happening that the political class doesn’t quite understand. However, political parties are desperate for this kind of data to construct voter profiles. This is data that FPTP cannot provide.

Preferential voting also allows you to send strong signals. At present in the UK there is the discussion about “tactical voting” between Labour, the Lib-Dems and Greens voters as a way to defeat Conservative candidates. This means voters are often voting negatively. The voter gives up on the party or candidate they actually think is the best.

Preferential voting doesn’t involve these kind of trade-offs. In Australia, in most seats people who vote for the Greens know that the party won’t win, but their vote sends a strong signal about issues that are important to them, while their preferences demonstrate they would prefer the Labor candidate to win (at least to the Liberal or National candidate).¹ A winning Labor candidate is incentivised to take into consideration the size of the Green vote. A tactical vote in the UK lacks this kind of voter pressure on the winning candidate – knowing who voted strategically is impossible. Labor, however, knows exactly how many Greens preferences it received.

Currently in Australia, Labor’s primary vote (the number of first preferences it receives) is lower than the Liberal and National parties (that exist in a permanent coalition and mostly don’t compete against each other)². Labor wins elections on Greens preferences. Of course, the Greens are far too immature to take advantage of this, but a more serious party would work its leverage more productively. The point being that the voting system has these opportunities for influencing larger parties built into it.

One thing that Campbell and Stewart should be wary of is that PR significantly lowers the bar to entry for extremist parties. Were the UK change to PR it will not simply produce similar results but with a more equitable distribution of seats. As we are seeing in countries like Germany and Sweden, once radical parties clear the electoral threshold (usually 4 or 5%) they are able to gain considerable traction.

The consensus model of preferential voting is able to restrain these forces. Australia has a radical reactionary party in One Nation – not to be confused with One Nation Tories in the UK. The system offer them oxygen – as liberal democracies should – but the party’s inability to gain preferences means they fail the test of consensus. The cordon sanitaire exists at the ballot box by the voters, rather than inside the parliament by parties like in Europe. Across Australia’s federal, state and territory parliaments, One Nation only has 1 lower house seat (in Queensland, of course). However, in Australia’s upper houses, where PR is used, the party has regularly been able to secure seats. Although, fortunately, never many.³

This, however, doesn’t mean that preferential voting is immune to new political forces entering parliaments. Voting systems that prevent new movements emerging are not healthy. But, rather than sweeping new political forces, preferential voting offers pathways on a seat-by-seat basis.

Preferential voting still has strong advantages for well-established parties. These parties can rely on the public’s awareness of them to maintain large primary votes and have preferences trickle back towards them. However, when the public senses an opportunity to remove a major party MP from office, preferences can be used highly effectively.

Campbell and Stewart have frequently mentioned the “Teals” when discussing Australian politics. These are a loose collection of women who have been able to win 7 urban seats from the Liberal Party.⁴ However, the model they used to do so – the Community Independents Project – was actually developed in the rural seat of Indi in 2013, and has won four elections there since (with two different candidates).⁵

This movement has cleverly used preferential voting – knowing that in order to win seats they need to siphon off enough Liberal Party voters to bring its primary vote down into the low-40s, and then gain enough primary votes themselves be in the final two run-off. This enables them to scoop up the preferences from other parties. Positive grassroots campaigns that build community trust has enabled them to become the consensus candidates (something One Nation can’t do).

This movement has also understood something very important about modern politics – people have become suspicious of political parties.⁶ Creating a movement, but without a party, is central to their success. Alongside their grassroots efforts that focus on local issues, they’ve also found three big issues to gain national attention – climate change, women’s equality, and establishing parliamentary integrity commission (all weakness of the Liberal Party).

Preferential voting encourages new movements to go seat-by-seat. In doing so it protects the core component of Westminster systems – the seat itself, and the need for each seat’s unique interests to be represented. Although parties have come to dominate these systems, the principle remains that it is the candidate who is elected, rather than the party. This is why by-elections are held when a candidate resigns, or why MPs can leave or switch parties. It’s also why the ability for independents to get elected is an essential feature of the system.

PR places greater power in the hands of parties. The party, rather than the candidate, becomes the central component of the system. In Europe, consensus-forming comes in post-election coalition talks by the parties, not from the voter’s individual ballot. PR also makes it incredibly difficult for independent candidates to reach the threshold as PR usually creates larger constituencies.

By focusing on the individual seat, preferential voting protects the political system from radical shocks. This is something that Stewart especially should be sympathetic towards, as this is a stabilising mechanism. It allows the public to test-run new political movements. For them to gain some traction, demonstrate what they have to offer, and then have the public choose whether to institutionalise them.⁷

It is clear that the Australian public is currently trying to construct a new party system. Looking towards the United States, Australians are very aware of what a strict two-party system creates both politically and socially, and see this as a major risk. Preferential voting allows them to weaken the major parties, but do so in increments. This may not be as exciting as radical change, but it probably a lot healthier. Allowing new movements time to build institutional knowledge.

In a preferential voting system concentrations of votes is still the key – the individual seat remains central so there is no perfect seat to country-wide vote ratio. So in the UK, were the Liberal Democrats to return to the vote percentage of their pre-coalition era it is not guaranteed that their seats won would accurately match their percentage of the overall vote. However, in seats where they are competitive their chances of winning would be increased.

At the 2019 election, the Liberal Democrats finished second in 91 seats. Most of these seats are Lib-Dem/Tory contests, and Labour voters are far more likely to preference the Lib-Dems over the Tories. The opportunities to become the consensus candidate in these seats are greater. But it should also be recognised that changing voting systems changes voter behaviour – a voter freer from the zero-sum choice of FPTP would also shift the calculations they make. And PR would shift voter behaviour the most.

Preferential voting offers the voter far more opportunities to make political calculations themselves. It encourages the voter to think more, to weigh options, make priorities, and to develop a greater political awareness. Numbering boxes may not seem like a great leap from simply ticking a box, but within that numbering there is an acknowledgement of the complex web of politics - how the differing ideas, interests, values and beliefs of the public are part of society as a whole. Each ballot is a conversation – or a way to disagree agreeably, as Campbell and Stewart would say.

1. There are a handful of seats where the competition is between Labor and the Greens and this calculation isn’t the case.

2. This gets a bit messy as in seats not held by either the Liberal or National parties in NSW and Victoria the parties do compete against each other, so the primary vote of the Coalition is boosted by this. They also compete against in each in WA and SA, as these branches of the National Party are not subject to the Coalition agreement. These votes count to the Coalition’s primary vote, even though they technically shouldn’t. Also, in Queensland the Liberal Party and the National Party are fully merged into a single party – the Liberal National Party.

3. Queensland has a unicameral parliament. Upper Houses in Australia can take different formats, and most use PR (with a preferential element). The federal Senate is divided by states with 12 elected senators and 2 senators each for the NT and ACT. Victoria divides itself into 8 districts of 5 members, while New South Wales has a single state-wide electorate delivering 42 members via PR.

NSW makes it easiest to elect parties like One Nation to their upper house, and 3 were elected at the last election. But given the chaotic nature of the party 2 of these now sit as independents.

4. The name “Teal” comes from these candidates being a mixture of economically liberal (the blue of the Liberal Party) with a focus on climate change (green). The movement is effectively a split in the Liberal Party – a liberal revolt from an increasingly conservative party. But the party is splitting at the voter-level, rather than the party-level. Under a European-style PR system the Liberal Party would probably split into 3 distinct parties. They’re incentivised to stay a “catch all” party by the voting system. But they’re no longer catching wealthy, highly-educated, female voters.

5. This episode of Australian Story on Cathy McGowan, who first won the seat of Indi, and is the godmother of the movement, is worth watching to understand the movement.

6. There has been a significant decline in the percentage of primary votes Labor and the Coalition have been receiving. In the mid-1970s their combined primary votes were just under 96% – that is, almost everyone gave their first preference to a major party. By 2007 this had steadily dropped to 85%. However, then has fallen dramatically – at the 2022 the percentage was 68.5%

7. The Greens have successful used this pathway. Although they are able to gain a larger presence in the Senate due to its use of PR, the party now has 4 seats in the House of Representatives.