Proposed senate voting reforms would be a Coalition own-goal.

by on 23 September, 2015

The Coalition has again raised the prospect of removing group preference flows from Senate elections.

As detailed in the analysis below of federal elections since 2007, under the proposed method the Liberal and National party would have lost ground to Labor and the Greens at 2 out of the 3 last elections, would not have gained a seat at any and almost certainly have less senators in the long run. Neither method offered either Labor or the Coalition a majority of the Senate seats contested at any of the last three elections.

Assuming voting patterns do not change, or even just assuming that the basic political landscape does not change (ie, the broad left being amalgamated under the Greens, whilst the broad right is spread across a number of libertarian, nationalist, populist and Christian parties), the new method will benefit Labor and the Greens and hurt the Coalition.


This analysis looks at the last 3 federal elections 2013, 2010 and 2007, which happen to conveniently represent three different voting scenarios: A strong showing for the right (2013) a strong showing for the left (2007) and a very close election (2010).

This analysis assumes that people would vote for the same group tickets in a scenario where preferences exhaust immediately. This is essentially the case in the NSW State upper house elections which already use the proposed method.

In order to determine the results under the new method, I took the AEC quota results, and counted off the top six places, whilst exhausting quotas.

Here is an example from the most recent NSW Senate election:

NSW – Senate 2013 Votes % Quota New method Current method
Liberal & Nationals 1,496,752 34.2 2.3942 2 3
Labor 1,381,047 31.56 2.2091 2 2
Liberal Democrats 415,901 9.5 0.6653 1 1
The Greens 340,941 7.79 0.5454 1
Palmer United Party 148,281 3.39 0.2372


The seats obtained by each party under the current method (the actual result returned) is the result of complex preference distributions. Under the new method, each of the two major groups get two seats for their complete quotas and are then left with only the remainders (0.3942 and .2091). The next seat awarded goes to the highest remaining quota, the Liberal Democrats, followed by the Greens. There are no more seats to award, but to illustrate, the theoretical seventh seat would have gone to the Liberals & Nationals as their remainder of .3942 quota was larger than the Palmer United Party’s 0.2372 quota.

As can be seen here, the new method can yield significant and surprising results – in this case, the Coalition lose a seat to the Greens and the minor party seats are unchanged.


2013 Federal Election (and 2014 WA rerun)


Votes % Seats won – current method Seats won – new method
Liberal/National Coalition 5,057,218 37.7 17 17
Australian Labor Party 4,038,591 30.11 12 14
Australian Greens 1,159,588 8.65 4 5
Palmer United Party 658,976 4.91 3 1
Liberal Democrats 523,831 3.91 1 1
Xenophon Group 258,376 1.93 1 2
Family First Party 149,306 1.11 1 0
Democratic Labour Party 112,549 0.84 0 0
Australian Motoring Enthusiast Party 67,560 0.5 1 0
Other 1,384,027 10.32 0 0
Total 13,413,019 40 40


The 2013 election is one in which Labor and the Greens did poorly and the Coalition did well. Under the new method Labor and the Greens benefit significantly at the expense of the minor parties, mostly Palmer United. The Coalition’s overall seat count is unchanged. The change increases the Greens and Labor’s share of seats 5% above their share of votes, and changes the Palmer United Party from being over-represented to under-represented. The Liberal Democrats are still under-represented and the Xenophon group increases their over-representation.

2010 Federal Election



Votes % Current method New method
Liberal/National Coalition 4,871,871 38.3 18 17
Australian Labor Party 4,469,734 35.13 15 17
Australian Greens 1,667,315 13.11 6 6
Family First Party 267,493 2.1 0 0
Democratic Labor Party 134,987 1.06 1 0
Independents 55,786 0.44 0 0
Other 1,255,047 9.86 0 0
Total 12,722,233 40 40


The 2010 election was very close and resulted in a hung parliament in the lower house. In the 2010 election, under the new method Labor gains 2 senate seats, 1 each from the Coalition and the DLP. Because this election was close and the Greens had a strong result, group distribution played a smaller part in the final outcome under the old method, so the new method has a smaller impact. The Coalition would have lost a seat in South Australia and the DLP would have lost their only seat in Victoria (both to Labor).

In terms of representation, The Coalition and Greens are over-represented in both the old and new methods (for the Coalition the new method is closer), whilst the Labor party is much more over-represented in the new method. The minor parties are significantly under-represented under both methods, with the new method being worse.

2007 Federal Election



Votes % Current method New method
Australian Labor Party 5101200 40.3 18 19
Liberal/National Coalition 5055095 39.94 18 18
Australian Greens 1144751 9.04 3 2
Independents (incl. Xenophon) 174458 1.38 1 1
Other 813538 9.34 0 0
Total 12656805 40 40

In the 2007 election, use of the new method would have transferred a single seat from the Greens to Labor. This would have made the final senate less representative of the total vote proportions.


Use of the new method would have had the most effect in the most recent Australian federal election. This is because the new method reduces the influence of groups with a vote share of 7% and less. Over the last 3 elections, using the new method, no candidate would have been elected with less than 7.79% of the vote in their state, (0.54 quotas). By contrast, under the current method, in a fragmented election, parties can be elected with as little as 1% of the total vote in their state.

Overall, use of the new method would have increased the systematic over-representation of the three major parties which is present under the current method.

Because the Australian right is currently more fragmented than the left, the Coalition currently enjoys greater preference flows from parties with a share of less than 7%. A new anti-Islam party, the ALA is launching this year and it will also likely poll in the ~5% range and would be expected to preference the Coalition before Labor or the Greens. Removing these preference flows presents a gift to the ALP and Greens.

In a situation where the Coalition does not expect to use minor and micro-party preference flows, it is still hard to imagine that eliminating the micro-parties will benefit the Coalition. The Coalition has no credible chance of gaining a majority in the Senate with or without these reforms. Surely the Coalition would find it easier to deal with the LDP, Family First or the ALA than Labor or the Greens. Even the presence of the more ideologically suspect DLP, PUP or Xenophon would at least provide an alternative negotiating partner in the Senate.

There is no benefit to the Coalition in the proposed Senate voting rule changes.

Leave a Reply