The case of the surplus votes

SOMETHING extraordinary occurred just after the State election in February 2001.

Several conservative MPs cried foul, claiming the WA Electoral Commission (WAEC) had erroneously transferred crucial Legislative Council preference votes, with one of their down-the-ticket-members losing out. The com-mission refuted the claim.

However, some MPs continued contending they, not the WAEC, were correct, so a ‘Mexican stand-off’ ensued.

Things were tense and one MP even moved to launch a court of disputed returns challenge.

A just-released WAEC report titled Determining the Result: Transferring Surplus Votes in the WA Legislative Council describes this behind-the-scenes tussle.

“The catalyst for the criticism was the outcome of the election in the Mining and Pastoral Region,” it says.

“Within weeks of the 2001 election, concerns were raised about the method used to transfer surplus votes of elected candidates.

“Two political parties claimed in Parliament that the WAEC had failed to correctly apply the process set out in Schedule 1 of the Electoral Act 1907.”

The standoff stemmed from inherent counting complexities, with former Liberal MLC Greg Smith losing out to Greens candidate Robin Chapple.

Most people, including MPs, quickly grasp how WA’s Legislative Assembly single member seat preferential voting operates.

Candidates with least votes are eliminated and their preferences are distributed to remaining candidates until there are two left. The one with the highest number wins.

But counting procedures in the Legislative Council, with its six multi-member regions, became markedly more complicated after WA adopted proportional representation in 1987.

Not to put too fine a point on it, counting is so complex scrutineers without computers become baffled and, therefore, suspicious.

Anyone doubting this should read the WAEC paper by political scientist Dr Narelle Miragliotta, who was commissioned to undertake this challenging project.

Her plainly written report shows she’s one of a tiny group of experts who grasps how surplus vote transference works. Though not bedside reading, it’s a valuable study.

Probably the best approach is to consider a case involving 100 voters and five parties – A, B, C, D, E – each with multiple candidates.

In this case what’s called a ‘quota’ for a candidate to be elected is 16.66 per cent – if five seats are contested, five plus one, divided into 100 – of primary votes plus one vote.

If parties scored: A, 40; B, 38; C, 10; D, 7; and E, 5; clearly A and B have two candidates elected because each scored two quotas (33.4 pc).

But who’s the winning fifth candidate?

Could it be C’s, who scored 10? Or is it B’s, with 4.6 pc, but who may attract most preferences from say D and E? Or might it be A’s third, who gained the remaining 6.6 pc of his party with preferences to come?

It’s a complex issue, as Dr Miragliotta outlines.

“Proportional representation is premised on the idea that the composition of members in parliament should reflect the approximate wishes of the voting public,” she says.

“It attempts to achieve this by lowering the threshold (to 16.66 in five seats and 12.50 in two others) required to win a seat so as to ensure that political groups and candidates who represent minority views and opinions have increased opportunity to gain election to the legislature.”

Significantly, the Legislative Council now has three One Nation and five Green MPs, whereas neither party has even one member in the 57-strong single member electorate Legislative House.

Labor won 37 per cent of the Legislative Assembly’s statewide vote last election but gained 56 per cent of the seats.

Such discrepancies aren’t uncommon in single member chambers – meaning minority party voters become disenfranchised.

Not so with multi-member proportional representation chambers, such as WA’s Legislative Council since 1987.

But the downside is the transference of surplus votes.

Dr Miragliotta assesses all the ways of calculating surplus vote transference: the Random Selection; Gregory; Inclusive Gregory; Weighted Inclusive Gregory; and the Meek systems.

WA opted for Inclusive Gregory, as did SA and Australia’s Senate.

“The research shows that there are significant variations between the systems,” Dr Miragliotta says.

“Not only does each method attribute different values to outgoing or transferred surplus votes, but there are also differences in which ballot papers of an elected candidate are selected for transfer.

“There is also evidence that each of the five methods has the potential to exert a slightly different impact on the outcome of an election.”

Little wonder some MPs were bitter their man lost.

However, Dr Miragliotta doesn’t adjudicate on the Liberal-

Smith/ Green-Chapple imbroglio.

True, this isn’t a clear-cut conclusion to what she shows is a complex issue, meaning it’s time our party leaders convened a special parliamentary assessment committee that reached a consensus on Legislative Council surplus vote transference so electors and MPs don’t again undergo what happened last year.

And Dr Miragliotta’s paper is an excellent launching pad to developing that consensus on this little-publicised but important tabulating question.