Preferential voting for the Australian Senate has the option of an official donkey vote where the voter puts a 1 in one box above the line to designate the party they prefer and lets the corresponding party effectively fill in the rest of the ballot for them by the preference tickets published before the election.
Because that is obviously easier, the vast majority of voters take that option; 1.26 million out of the 1.35 million who turned up to vote.
For the total votes, AEC publishes total, first-preference party votes from which the WA figures can be extracted.
|Australian Labor Party||281291||26.77%|
|The Greens (WA)||100869||9.60%|
|Australian Fishing and Lifestyle Party||4205||0.40%|
|Animal Justice Party||7308||0.70%|
|Australian Motoring Enthusiast Party||5905||0.56%|
|Shooters and Fishers||10505||1.00%|
|Family First Party||6789||0.65%|
|Help End Marijuana Prohibition (HEMP) Party||10863||1.03%|
|Katter’s Australian Party||2862||0.27%|
|Stop The Greens||1714||0.16%|
|Palmer United Party||54987||5.23%|
|Rise Up Australia Party||3095||0.29%|
|Socialist Equality Party||911||0.09%|
|Secular Party of Australia||1144||0.11%|
|Stable Population Party||1050||0.10%|
|Australian Sports Party||2280||0.22%|
|No Carbon Tax Climate Sceptics||1099||0.10%|
|Australian Voice Party||944||0.09%|
|The Wikileaks Party||7321||0.70%|
I’ve calculated the percentages to help illustrate how a proportional allocation of Senate seats could be done; approximately how it’s currently done. If there are 6 vacancies to fill, then one sixth (16.67%) of the votes is required for each vacancy.
In this case that one sixth is 175,105.5 votes per seat. If we look at the parties, then the Liberal Party automatically has 2 seats and the ALP has one, with others failing to get numbers for even one seat. But both of the parties with seats have “spare” votes floating around (57,599 and 106,185.5 respectively) and that is where preferences begin to flow in strange and wonderful ways.
Well, the result so that are known and were unpopular and unsatisfactory as several people got seats with fewer than 1000 “primary” votes, well ahead of others with more substantial, high preferences.
So back to the ticket votes, as they are called, to examine how a different treatment of “preferences” could be tallied for a more representative result. From the September 2013 polls up in Western Australia I see the above the line votes as: (I could be mistaken.)
|D||Help End Marijuana Prohibition (HEMP) Party||9943|
|E||Socialist Equality Party||790|
|F||Palmer United Party||52768|
|G||Shooters and Fishers||9702|
|H||Australian Voice Party||898|
|J||Secular Party of Australia||771|
|L||The Wikileaks Party||6077|
|M||Katter’s Australian Party||2565|
|N||Family First Party||6409|
|O||No Carbon Tax Climate Sceptics||1029|
|P||Stable Population Party||802|
|Q||Stop The Greens||1605|
|S||The Greens (WA)||89306|
|T||Animal Justice Party||6759|
|V||Australian Fishing and Lifestyle Party||4051|
|W||Australian Motoring Enthusiast Party||5588|
|X||Australian Sports Party||2183|
|Y||Rise Up Australia Party||2786|
|Z||Australian Labor Party||274712|
One needs to refer to the (first) preference tickets (essentially their “how to vote cards”) of the respective parties to determine how the votes would then flow to individual candidates. Fortunately, only the first 6 preferences would be significant.
The following table shows the sum of the top-20 candidates’ above the line and below the line totals after binary-weighting on preferences. The preference weight as discussed previously is also shown as it’s interesting; but only for those who missed out on the top 6 positions.
Disclaimer: There are probably errors in the exact vote figures.
Remember; the weightings of preferences are arbitrary.
The only way to know how the voters value each preference, is to inform them before the election exactly how the vote will be counted, in as simple a manner as possible.
Distribution of preferences on a Senate ballot is evidently very, very
complex; almost chaotic and therefore difficult to grasp even for experts. It’s not sufficient at present to try to map out the preference flows for the first 6 preferences because (IIRC) even the 44th preference has made a difference in the past, getting the last Senator into a seat. The process is near enough to opaque to most voters; not a good look in an open democracy.
Transparency may be achieved by not distribution preferences, but by weighting them in a predictable manner; a simple formula/method that the voter can understand without much difficulty and vote accordingly.