The Chisholm and Kooyong Signs Challenges

Update Feb 20: The Federal Court has decided not to report anything to the High Court regarding alleged violations of the Electoral Act.  It has decided that it did not have before it such material as would allow it to decide that "the relevant fault element" was satisfied to be sure that an offence had been committed.  At [28], the Court holds: "We think the better view is that the Court is obliged to comply with s 363 when it is persuaded that there is material from which it can be concluded that a person not only was responsible for the physical elements of a contravention of s 329(1), but also that the relevant fault element was satisfied."

The function of a Disputed Returns hearing is to decide whether a seat needs to be vacated and it is not surprising that the Court did not turn over all the rocks that might have been needed to fully explore the question of whether an offence was committed.  That question was ultimately irrelevant to the outcome of the challenges.  This does however raise the question of what recourse there is over the misleading signs, if any.  Normally these matters would be pursued through the AEC (not necessarily to prosecution, perhaps a warning), but the AEC initially submitted that there was no offence anyway.  It might not be all that interested in trying to now establish whether there was a fault element involved in the production of an illegal sign that it did not consider illegal in the first place.  Especially, it is unclear who would be in a position to demonstrate that fault element in a complaint.   I am not sure therefore whether the matter will go any further.

Again, it is my view that these sorts of signs should be banned.  The above outcome of the case only further highlights why the existing law does not adequately address misleading and deceptive signs that pretend to be official electoral signs.

The only other action of interest today was the costs order.  The AEC is to bear its own costs.  The Court is considering ordering that the Commonwealth bear the costs of the successful defendants Liu and Frydenberg, the unsuccessful challenger (Garbett) in Chisholm (because of public interest) but not the unsuccessful challenger (Yates) in Kooyong (because of a it being a duplicate case that was obviously unrealistic given Frydenberg's margin.) The Court will determine these matters later on further papers from the parties.

Update Dec 24: As expected both petitions have been dismissed. However the Court has asked Simon Frost to show cause why an apparent violation of the Electoral Act by him and possibly others should not be reported, suggesting that it has provisionally found the signs were illegal but has found that the outcome should not be altered as there is insufficient evidence that they changed the results in those seats.  The judgement is available here.

The Guardian has an excerpt:

"“In our view, the corflutes are properly read, not as encouragement to vote 1 Liberal, but as a statement first, that to vote correctly (that is validly), one must vote 1 Liberal and, secondly, that there was an official instruction of the AEC that electors must cast their votes as indicated,” the court said."

The court has found that the signs did have the capacity to mislead electors (albeit naive, gullible or uninformed ones) in relation to the casting of their votes but that at most a handful of voters could have been so misled, nowhere near sufficient to overturn the result.  The court has found the signs only had capacity to mislead where they were placed next to an AEC sign.  Simon Frost has been given a chance to argue that he should not be the subject of a declaration of illegality because he was not represented while giving evidence in court.

Update Dec 21: Judgement in these cases will be handed down at 2:15 pm December 24.

Updates Nov 6-8

The case, before three judges, is now on, and expected to run for three days, after which the court may well reserve its judgement.  Tweeted coverage is being provided by Josh Taylor of the Guardian and I will link here to other reports of interest that I see.  If anything of special interest comes up I may discuss it at length here.

Nov 6 12:30: Of some interest today is discussion about the signs having said something different to what was intended (as touched on below) - Frost says that he provided an intended meaning but the actual signs when translated said something different.  According to Taylor "Frost said he made no inquiries on election day to make sure the corflutes said what he authorised them to say. Frydenberg and Liu didn't contact him to ask about them on election day" and "Frost says he doesn't know if anyone who proof-read the corflutes before election day speak/read Chinese." (It may be significant here that Liu could read the signs for herself.) However later in his evidence Frost said that the Hotham Liberal candidate, George Hua, checked the signs.  According to Taylor, Frost has also admitted that the corflute was intended to convey the impression that it was an AEC sign.

Under Section 329 (5) "it is a defence if the person proves that he or she did not know, and could not reasonably be expected to have known, that the matter or thing was likely to mislead an elector in relation to the casting of a vote." however "Note: A defendant bears a legal burden in relation to the defence in subsection (5) (see section 13.4 of the Criminal Code )." It should be kept in mind that Simon Frost is not on trial for breaching Section 329 at present.



Nov 6 5:10 pm: This verbatim exchange reported by Taylor is notable:

“You intended to convey the impression that this was an AEC corflute, didn’t you?” De Ferrari asked.

Frost, now an adviser to Frydenberg, took a long pause before replying: “It was similar to the AEC colours, yes.”

“So the answer to my question is yes?” De Ferrari pressed.

“Yes,” Frost replied.

During the 2016 campaign Labor's "Mediscare" material resulted in laws being changed to create offences relating to impersonating a Commonwealth body.  Although that is not at stake in this case it will be interesting to see if any complaint might be made under those laws in relation to these disclosures.  I am not familiar with the law in this area though. Graeme Orr on Twitter has noted that the standard of proof required (criminal law) is very high and also that examples given in the explanatory memorandum involve explicit statements of being the same entity (such as use of a false letterhead, etc).

The ABC has a good detailed report.

Nov 7: Very busy today but some quick highlights from Taylor's tweeting.

* Lisa De Ferrari (barrister for the petitioners) has pointed out it is uncontested Gladys Liu knew about the signs.  Her reported argument that Frydenberg knew is unconvincing to me - that people tweeted at him about it and there's no evidence he didn't see the tweets.   Federal parliamentarians, especially major ones, are often tweeted at with all kinds of hostile and ludicrous garbage and would rarely read it all, especially on election day!

* There is mention of the corflute being placed next to AEC material at certain booths (a minority), but not whether any argument is being made about voting patterns in those booths.

* The AEC is arguing that the number of Mandarin/Cantonese-only speakers (as opposed to those with some fluency in those languages and also English) is too small to change the result, and also arguing that voters understand Australia isn't a one-party state and that the AEC wouldn't issue instructions supporting a party.

* Frydenberg's lawyer has pointed out that there are not enough fluent readers in his electorate to change the outcome and has suggested that the case is being brought for other reasons.

Nov 8: Solomon QC for the Liberal Party has, according to Taylor, now said that it was George Hua who prepared the incorrect translation.  Solomon has also said that similar corflutes were used at the Bennelong by-election.

Judgement has now been reserved; there is no indication of how long a decision might take.  I haven't seen anything in the reporting that suggests to me that either result will be overturned but the exact nature of the judges' findings will be of interest.

Meanwhile the Greens have referred the matter of the signs possibly impersonating a Commonwealth body to the Australian Federal Police.

------------------------------------------------------------------------------------------------------------------

Original article

Thought it was time to put up some coverage of the Court of Disputed Returns election signs challenges to the elections of Gladys Liu in Chisholm and Josh Frydenberg in Kooyong.  Frydenberg is facing a distinct Section 44 challenge that has yet to commence.  There has been a fair amount of interest in these challenges so I thought I would post some comments about where I see matters currently.  I am not professionally involved in these cases, which don't seem to have required any advanced voting pattern analysis though in theory the Chisholm one could have done so.

Referral to Federal Court

These challenges were initially heard for directions by a single judge in the High Court but were then referred to the Federal Court.  The High Court has the options under Section 354 of the Electoral Act to hear challenges itself, to refer specific details to the Federal Court or to refer a case to the Federal Court entirely.  It has chosen the latter option.  The case is entirely about interpretation of the Electoral Act and not constitutional issues, and Gordon J rejected the middle option on the grounds that "It is always easier to decide facts in the context of legal questions or other issues and I think it is fraught with danger to split it."

The referral to the Federal Court is inconvenient for observers because transcripts are not automatically available for free online.   There appears to be an application process and a $50 per document fee.  I would like to see changes made so that future CDR cases referred to the Federal Court are automatically treated as public interest cases with documents available online for free as in the High Court. My understanding of the submissions of the parties is limited to what has been reported in the media, which is not necessarily accurate or complete.



What the petitioners are claiming

The petitioners are alleging that Liberal Party signs displayed in the electorates of Kooyong and Chisholm breached section 329 of the Electoral Act by being likely to mislead or deceive electors in relation to the casting of their votes.  As is well known these were signs with Chinese writing that used the same colours as Australian Electoral Commission signs.  They have variously been translated as telling voters that voting Liberal is "The right way to vote", "The correct way to vote", or "correct voting method" (all these were submitted by the petitioners).  The Liberal Party has been reported as agreeing with that translation while saying that the intended translation was “make your vote count”.

The petitioners argue that the signs were likely to mislead electors capable of reading them into thinking that the only a 1 Liberal vote would be accepted as valid, or that the AEC was advising voters to vote 1 Liberal, or both.

Section 329 is not a general truth-in-advertising provision.  An attempt to construe a similar previous wording as a truth-in-advertising provision failed in Evans v Crichton-Browne (1981).  The Federal Court confirmed that the meaning of the current provision is the same in Peebles v The Honourable Tony Burke (2010).  A key passage in the former is as follows:

9. [..] The use of this phrase in s. 161 (e) suggests that the Parliament is concerned with misleading or incorrect statements which are intended or likely to affect an elector when he seeks to record and give effect to the judgment which he has formed as to the candidate for whom he intends to vote, rather than with statements which might affect the formation of that judgment. Certainly par. (d) of s. 161 is concerned only with a particular instance of a misleading or incorrect statement of that kind, namely a statement contained in a document representing, or apparently intended to represent, a ballot-paper. For example, a document in the form of a ballot-paper, which contained directions that a ballot-paper must be marked in a manner different from that provided by the Act, or in a manner that would favour a particular candidate, would, if the other conditions were satisfied, fall within the prohibition of s. 161 (d). It seems reasonable to conclude that s. 161 (e) was intended to deal with misleading or incorrect statements of a similar kind, even though not contained in a representation of a ballot-paper. For example, a statement contained in a newspaper advertisement that a ballot-paper should be marked in a way that would not conform to the requirements of the Act and which would render the vote invalid might mislead or improperly interfere with an elector in the casting of his vote. The same might be true of a statement that a person who wished to support a particular party should vote for a particular candidate, when that candidate in fact belonged to a rival party. An erroneous statement as to the hours or place of polling which had the result that an elector (perhaps in a remote country district) failed to get to a polling booth in time to vote would have misled that elector in relation to the casting of his vote, although it would not have misled him in casting his vote, since in the case imagined no vote was cast. [..]

While this is a fairly narrow formulation then, the exact scope of it is still being fleshed out.  In the 2018 Victorian state election, a Victorian Liberal statewide how-to-vote that advocated a 1 vote for a Liberal candidate who had in fact been disendorsed was found to breach similar Victorian provisions (even though the Liberal Party said it still wanted voters to vote for that candidate) on the grounds that it could mislead the voter into thinking the candidate was still endorsed.  It was also held in the Victorian case (which survived on appeal) that material can be misleading if it is able to fool "any gullible or naïve elector" rather than one of "ordinary sophistication".

So there seems to be plenty of room for the petitioners to argue that although the test is a narrow one, the signs at least could fall partly on the mechanics-of-voting side and not wholly on the formation-of-voting-decision side.  However I would expect that to still be discussed further to try to show that terms like "right way to vote", "correct way to vote" and so on could really be taken to mean that anything else was informal, as opposed to the harmless meaning that voting this way is just a good idea.

What the petitioners need to show

Unfortunately for the petitioners (and this has been missed by a lot of people opining on the matter on social media) the requirements of the cases in terms of having the results voided go much further than just showing that the signs were likely to mislead electors in relation to the casting of their votes.

The obstacles mainly come from Section 362:

(3)  The Court of Disputed Returns shall not declare that any person returned as elected was not duly elected, or declare any election void:

                     (a)  on the ground of any illegal practice committed by any person other than the candidate and without the knowledge or authority of the candidate; or

                     (b)  on the ground of any illegal practice other than bribery or corruption or attempted bribery or corruption;

unless the Court is satisfied that the result of the election was likely to be affected, and that it is just that the candidate should be declared not to be duly elected or that the election should be declared void.

Bribery and corruption are not alleged in this case, so even if the signs did breach the Act, the court must still be satisfied that the result of the election was "likely" to be affected and that it is just that the election in these seats be declared void.  However, "likely" will not necessarily be interpreted as meaning more likely than not; it may just be interpreted as meaning a serious possibility.

Possible outcomes

The following are, in theory, the possible outcomes of each case:

* The case is completely dismissed, either on a technicality or because the signs are found not to mislead the elector in relation to the casting of their vote.

* The court decides that Section 329 was breached but that the conditions of Section 362 were not met.  This would mean the result stands, but it could lead to some Liberal Party figure(s) being prosecuted for breaching Section 329, which would be embarrassing.

* The court upholds the petition in full, which would result in a by-election in the seat.  The unseated incumbent would be able to contest the by-election, but in the case of Chisholm the Liberal Party would have to think about whether to re-endorse Gladys Liu or choose somebody else.

The AEC's role

The AEC is entitled to join proceedings in Disputed Returns cases by leave of the court.  In this instance the AEC is opposing the petitions and arguing that they should be dismissed.  Especially, the AEC is arguing that under Section 355 the petitions could be thrown out for lacking "sufficient particularity to identify the specific matter or matters on which the petitioner relies as justifying the grant of relief".  (Under some conditions laid down in Section 358 the court may choose to still hear petitions that are deficient in this manner).

The extent of the AEC's intervention against the petition has attracted some attention.  The AEC's submissions on the day are implicitly criticised in the submissions but it is interesting that the AEC has sought to go further than just show that its decisions on the day were OK.

The AEC's point goes to the lack of detail in the petitions when it comes to evidence that results were actually affected.  The petitions state translated meanings of the signs and state that they were widely distributed, but mostly just ask the court to infer that as a result a significant number of electors would have changed their mind and that as a result the results would have been affected.  Without having read the AEC's submissions and seen exactly what they have in mind, I did notice that the petitions do not, for instance:

* refer to any evidence that any specific voter reported changing their mind as a result of the signs
* refer to any possible expert evidence that signs of this kind were likely to confuse and mislead Chinese-language voters (given whatever nuances of meaning the signs could be found to contain, and given these voters' various cultural and political backgrounds)

In theory the petitions might also have set out some form of voting-pattern analysis that was compatible with the result being affected by the signs.  However, the Kooyong petition does nothing of this sort at all.  The Chisholm petition does contain something that looks like it might be intended as vaguely relevant, but who knows:

"59. The total numbers of ordinary and absent votes for Chisholm were 76,561 and 4,885 of which 72,984 and 4,583 respectively were formal votes.

60. LIU received 31,445 ordinary votes and 1,687 absent votes.  YANG received 25,735 ordinary votes and 1,349 absent votes."

However that evidence actually shows that the ratio of absent Liu votes to absent Yang votes (1.25) is higher than the ratio of ordinary Liu votes to ordinary Yang votes (1.22), which is the reverse of what would be expected if ordinary voters exposed to the signs were caused to vote for Liu rather than Yang.  The more so since nationally the ratio of ordinary Coalition votes to ordinary Labor votes (1.22) was much higher than the same ratio for absent votes (1.06).  And in any case, the citation of "ordinary vote" figures is not very helpful because ordinary votes include within-electorate prepoll votes, which would also have been cast without access to the signs.

The figures on formal votes show that for Chisholm, the proportion of absent votes that were informal (6.18%) is higher than that for ordinary votes (4.67%).  This gap is wider than nationally (6.35% vs 5.83%) but if that means anything relevant to the case (which it may well not), then it is capable of multiple interpretations. It is consistent with the signs causing intending voters who would have otherwise voted informally to vote formally 1 Liberal, but that could be the case even if all those who were caused to vote formally were intended Liberal voters, in which case the signs would not have deceived anybody.

I've had a quick look for anything else in the voting patterns that might support the idea that the signs significantly affected the outcome in Chisholm, without finding it.  That doesn't necessarily mean it didn't happen - it is just that electoral data are noisy and an event that swings the votes of less than 1% of an electorate is typically not detectable, at least not reliably.

Upcoming hearing

The cases return to court for a hearing starting Nov 6, for which three days have been set aside.  Media reports have said that evidence called will relate to the printing of the signs, the extent to which the candidates were aware of the signs, the language backgrounds of the electorates and the Liberal campaign in relation to the signs.

Outlook

As usual I should add the disclaimer that I am not a lawyer.  I have no view on whether the court will rule that the petitions fail on any technicality.  I have no view on how the court will rule on the basic question of whether or not the signs were misleading under the Act.  I would be astonished if the Kooyong result was overturned and a by-election called, because of the large margins in that seat.  In the case of Chisholm the smaller margin makes things more interesting but still I would be rather surprised if the result was overturned and a by-election called.

The law should be changed

In my submission to JSCEM I have called for a specific change to the laws affecting such signs for the future.  There have been several cases of parties employing signs that look suspiciously like official electoral material in recent elections.  This isn't even the worst such case - for instance in NSW the Labor Party deployed official-electoral-material-like signs encouraging voters to just vote 1 in some electorates, and similar signs encouraging voters to distribute preferences in others.

I believe that whatever the outcome of these cases it should become clearly, specifically and seriously illegal to produce electoral signage that can reasonably be mistaken for official electoral material.  I cannot see any valid reason why such signs should be allowed.