More fearful than feared?

Photo sourced from 

In the Sydney Morning Herald of 15 September 2016, Tony Wright writes about the first speeches of One Nation Party Senators Pauline Hanson and Malcolm Roberts in the Australian Senate.

When Senator Hanson delivered her speech on the afternoon of 14 September she launched a twin attack on welfare recipients and Muslim immigration, saying ‘we are in danger of being swamped by Muslims.’ Wright criticises Senator Hanson’s position, saying that ‘the 2011 Census showed that Muslims made up just 2.2% of the Australian population, or 476,291 people.’

Wright also criticises the first speech of Senator Roberts, in which he denied that global warming could occur.

Wright says that ‘Australia is in danger of being swamped by mindlessness if the likes of Hanson and Roberts are to be treated seriously.’ Noting that the Greens senators walked out of the chamber during Hanson’s speech, Wright says that at that point Senator Hanson smiled as ‘the greens public dudgeon demonstrated they were taking her seriously, thus granting her power she had not earned.’

Waleed Aly also writes about the Greens walking out on Hanson's speech:

Our public discourse, and with it our politics, has evolved to a point where it has never been easier to prosper by pitching your tent as firmly as you can within a subculture. We're not so keen to speak with anyone anymore. We'll speak at, and we'll speak for – we'll call out and we'll represent. But we don't really engage. It's confessional. It's declaratory. But it's not even trying to be particularly persuasive. We're in a phase of mutual amplification: a race to express pre-determined views as loudly as possible for the thrill of those who share them. Neither Hanson nor the Greens were hoping to win over anyone in this episode.

Wright suggests that it is better to wear down those with different views by argument and facts. He says that as a matter of fact ‘One Nation got 593,000 votes at the election. That’s almost 120,000 more people than Australia’s entire population of Muslims.’ This leads to his final flourish ‘so who should fear whom?’

Representation by numbers

Presenting the national primary votes for One Nation in the Senate allows for an interesting juxtaposition of the numbers of the feared and the fearful. The figure of 593,000 primary votes is 4.285 percent of the voting population in 2016 (as opposed to 2.74 percent of the national population in 2011).

The primary votes cast in the Senate election are an indicator of support but because voting is compulsory in Australia, electors can use the allocation of their first vote as a form of protest, without necessarily believing that the person voted for will be elected and knowing that the value of that vote will be preserved by being allocated to their next preferred candidate if the first candidate is eliminated. The primary vote for a candidate or party is an imperfect measure of support given that the voting system is preferential. Electors also have equally valid votes for six candidates in a half Senate election and twelve in a double dissolution election. In the 2016 election One Nation candidates stood in each of the states where a total of 12 seats were being contested. All except Senator Hanson struggled to get across the line. 

At the other end of the vote count, the value of all votes are distributed according to the order in which they have been prioritised by each elector. In order of priority set by the voter, if a vote is given to a candidate who already has a quota, or to a candidate eliminated as having the least number of votes in each count, that vote then goes to the next candidate as numbered by the voter. As the candidates with the least number of votes are progressively excluded the votes that are transferred from them become more and more remote from each voter's initial preferences. This is particularly so in a double dissolution election where the voter has 12 votes to distribute rather than six. 

In this context, we should be wary of arguments that pre-suppose that our system of democracy in Australia is somehow perfect. We see this when a government sitting on a majority of one seat in the House of Representatives claims to have a mandate for all its policies. Clearly, the true picture is that the electorate is more or less evenly split and both sides represent closely matched numbers of electorates. In addition, the electoral system for the House promotes majoritarian rule and fails to give voice to those outside the two major parties. 

Importantly, because the Senate is elected according to a system of proportional representation it gives voice to a much wider range of views then the single member electorate system used for the House of Representatives. See for example, the analysis provided by Tim Trudgian who says (referrring to the six Senators elected at half Senate elections):

For the House of Representatives, each electorate chooses one person to represent it; for the Senate, each state chooses six people to represent it. This leads to a “fairer” representation in the Senate, at least with respect to the proportion of votes cast and seats won. Having even more senators would make this even fairer.

Trudgian points to the Constitutional requirement for a referendum to increase the number of Senate seats. We query whether further entrenching the preferencing of the views of those in the less populous states is democratic. We would add that the House of Representatives is much more easily reformed to make it more representative - particularly by halving the number of seats and requiring a male and female to be elected for each seat.

And yet, there is a danger of overstating how well individual senators represent the people given the strong element of chance in the allocation of the last seats to be filled. We are at risk of overestimating community support for the views of those who attract sensationalist attention while at the same time feeding into the rhetoric of alienation and marginalisation from politics that will ultimately strengthen support for those who enter politics in this way. This is amplified when those who represent extreme views end up with the balance of power.   

Voters in the smaller states by population are overrepresented

The Australian Constitution is a federalist compromise that gives equal Senate representation to each state irrespective of the population of that state. This means that in the less populous states far fewer votes are required to secure a Senate seat as shown in table 1. Table 1 shows the population of electors for each state and territory as used by the Australian Electoral Commission in allocating Senate quotas for the 2016 federal election and the number of votes required in each state and territory to achieve a quota for a Senate seat.

It can be seen that the quota for election for the double dissolution election held in 2016 is roughly half the number of voters that would have been required if a normal half Senate election had been conducted.

If Senate seats were evenly allocated according to the 2016 national population of electors (without reference to state boundaries) the quotas would be 182,090 in a double dissolution election and 364,181 in a half Senate election.

Table 1 Voter population and quotas for Senate calculations

Source: AEC website – distribution of preferences by State and Territory

How the One Nation Senators got elected

In the double dissolution election of 2016, four One Nation Senators were elected. In Queensland, Senator Hanson achieved a quota on primary votes. Had the election been a normal half Senate election and the flow of preferences the same, Senator Hanson would have likely been elected on preferences.

The other three senators, at best, achieved just over half a quota on primary votes and finally were elected at the end of the distribution process. It appears highly unlikely that these senators would have been elected in a half Senate election. It also appears that the exhaustion of votes that occurred because of the changes to the Senate voting system introduced in 2016 is likely to have had an important role in the final distribution of votes to these three One Nation Senators.

For example, see commentary by Adrian Beaumont who says:

Relative to national vote share, the major parties and the Greens are somewhat overrepresented in the Senate. This is because these parties can get high fractions of a quota, which puts them well in the hunt to win more seats than their overall vote share suggests. More popular parties also did better on preferences than micro parties.
Others are underrepresented in proportion to the overall Others vote because these parties could not individually reach a sizable fraction of a quota, and without the old group voting system, their preferences did not combine. At this election, One Nation was the only non-established party that was able to win seats in multiple states, with the Nick Xenophon Team (NXT) struggling outside SA.

Table 2 The role of primary votes and preference distribution in the election of One Nation Senators

Source: AEC website – distribution of preferences by State and Territory

* Elected on the final distribution of 209,430 votes in the final count when all other votes were exhausted and the votes allocated to Senator Roberts totalled 209,430, which was 45 less than the quota.

  • In Queensland, the quota for election as a senator in the double dissolution election was 209,475. If the election had been a normal half Senate election the quota would have been 389,023.

Senator Hanson received 249,983 primary votes which was sufficient to gain a quota in the double dissolution election but which would have been 139,000 votes short of a quota in a normal half Senate election. Senator Roberts received 77 primary votes and the 40,508 votes for Hanson in excess of a quota were allocated to Roberts. It was not until the 841st count that Senator Roberts was finally elected after the distribution of all other available votes. At that point, Senator Roberts had been allocated 209,430 votes which was 45 votes less than the formal quota. Interestingly, the number of votes that had been exhausted by the final count was 208,964 which almost matches the number of votes in the final distribution to Senator Roberts.

  • In New South Wales, the quota for election as a senator in the double dissolution election was 345,554. If the election had been a normal half Senate election the quota would have been 641,742.

Senator Burston received a primary vote of 182,849. This was just over half the quota required for election. It was not until the 1054th count that Senator Burston was elected. In the final count the number of exhausted votes in New South Wales was 414,656 which was almost 70,000 more than the formal quota.

  • In Western Australia, the quota for election as a senator in the double dissolution election was 105,091. If the election had been a normal half Senate election the quota would have been 195,168.

Senator Culleton received a primary vote of 54,712. This was just over half the quota required for election. It was not until the 539th count that Senator Culleton was elected. In the final count the number of exhausted votes in Western Australia was 85,766 which was approximately 20,000 votes short of a formal quota.

Table 3 shows how primary votes for one nation party candidates in the states of Victoria, South Australia and Tasmania fell well short of a quota. In the Australian Capital Territory and the Northern Territory no One Nation candidates stood for election.

Table 3 Primary votes for One Nation party candidates in other jurisdictions


NB: Election tally figures presented by the ABC and used here are party aggregations and differ from the primary vote counts from the AEC used in Table 2.