Thursday, August 10, 2017

Electoral Process, But Not As We Know It: Postal Plebsicite V2

An article I wrote about the serious defects of a postal plebiscite (back on the annual day reserved for silly jokes) has for some unfathomable reason more than doubled its hit tally in the last 24 hours.  Now that a postal plebiscite (but run by the Australian Bureau of Statistics, not the Australian Electoral Commission) has been announced by the government, it's time to update certain aspects of my commentary.

What it appears we will have (unless it is disallowed by the courts) is something so bizarre that it was not anticipated in any of the many polls about a plebiscite.  Effectively, it is a national vote on whether the government will allow a conscience vote to be brought on in the parliament.  (If the plebiscite proceeds and the "no" side wins, then the government will block a conscience vote, presumably ending any prospects for same-sex marriage for so long as the Coalition stays in power.  This rather heavy-handed approach appears to be an attempt to prevent a mass boycott from working.)

Is it constitutional?

I don't know, but we'll probably find out soon enough.  At least two sets of campaigners against the proposed plebiscite are filing for injunctions against it.  Section 83 of the Constitution requires that appropriations must be supported by law, and no law has been passed for this plebiscite.  However there are various standing general-purpose appropriations that governments have flexibility to use for the ordinary running of government, and also in emergencies.  The question will be whether an appropriation for this purpose is valid.

Issues that might be raised include

* whether the collection of statistics about opinions rather than facts is outside the ABS's purpose (see Anne Twomey's comments)
* whether the matter is urgent and unforseen enough to justify the appropriation, especially given that the Government has already presented bills on the matter to the Senate and knew that if it presented the voluntary postal plebiscite to the Senate it too would be rejected.
* whether the plebiscite is really a quasi-vote in the guise of a statistics collection (as a transparent attempt to avoid fatal legal problems with using the AEC) and hence not what it is being claimed to be

(See the Wilkie, Marlowe and PFLAG application for an injunction here).

I may well add others to this list.  It's also possible that the scheme will be found to be legal, but in a way that makes it impossible to really function without fresh regulations (which the Senate might then disallow).

The government claims a high level of confidence that its position is legally sound, while opponents seem to be making less bullish noises about their prospects than if the plebiscite was to be a fee-for-service plebiscite conducted by the AEC.  However, that does not necessarily mean anything.

The time frame for the Court to resolve the matter is not yet known, but I may edit in comments on this as they become available.

Process Issues

Many of the process issues I canvassed in the earlier article also apply to the current model.  Absent of at least regulations, there would seem to be no legal authorisation requirements, no laws against misleading electors about the electoral process, no obvious law against vote-buying (to which postal voting is more readily prone than booth voting), and no legal process to challenge the accuracy of the result.  A section of the Census and Statistics Act 1905 may allow for the creation of regulations to enforce things but they carry only weak penalties (and could be disallowed, though I doubt that the Senate would do that).  Possibly, other sources of regulations may emerge, and information on regulations is apparently coming on Monday. [EDIT: It looks like these issues may be addressed by special legislation, which the Senate would look pretty silly if it rejected it.  But the Senate might instead amend it - more on this later.]

The lack of an authorisation requirement is a problem because while circulating gay hate material may not have been illegal in a compulsory referendum (this would actually vary by state), without an authorisation requirement no-one is forced to put their name to hate fliers.  (Note: normal AEC rules on misleading material would only have policed false and misleading statements about the mechanics of the voting process.  They are not truth-in-advertising laws.)  (Update: I have now posted detailed coverage on the issue of hate speech laws.)

Issues regarding the extent of AEC cooperation remain poorly fleshed out at this stage, as do issues regarding exactly what information the ABS can obtain and the level of AEC discretion in providing it.  The ABS is permitted by the Electoral and Referendum Regulation 2016 to access the roll for the purposes of "collecting, compiling, analysing and disseminating statistics and related information".

If It's So Ridiculous, Should We Just Boycott It?

In the event that the postal plebiscite goes ahead, my strong advice to supporters of same-sex marriage is do not boycott.  (My very strong advice to opponents, of course, is to get out of other people's lives.)

Much as this farce deserves boycotting, it is already clear to me that a boycott would be highly unlikely to be a successful mass political movement.  To be such it would need, at least, enthusiastic support from Labor and the Greens, but it is clear that politicians from these parties are already gearing up to fight for the yes vote.  This makes sense given that recent polling by Essential finds that while the current option has only 9% support as the most preferred option, it is nonetheless one of many options that more voters than not consider acceptable (43% approve to 38% disapprove, with remarkably little partisan spread).  It is basically those who care deeply about the issue who are most fussed about how it is resolved.  Those with less deeply felt views about it are likely to just want the issue resolved and out of the way soon and not to be too fussed as to how.

There would be large political risks for the ALP in supporting a mass boycott.  It would look like the Coalition was keen to get on with resolving the issue while Labor was wrecking one opportunity too many to do so in the hope of getting all the credit for itself.  It is not so long ago that large numbers of Labor MPs voted against allowing same-sex marriage, and the party's formal opposition to it (including a 2019 sunset clause for conscience votes against) is a relatively recent evolution.  Moreover, Labor cannot seriously espouse the sort of  purist commitment to quality of electoral process that a boycott would entail after its dismally clueless performance in last year's Senate reform debate.

Absent of Labor support, I believe that a small-scale boycott, even if strongly supported by leading LGBTIQ groups, would only have the effect of watering down the yes vote without affecting turnout enough to do the result's credibility any further harm.  It doesn't help here that nobody has any idea really what sort of turnout there should be, so the impact of a boycott would be very hard to gauge, unless it was on such a massive scale that it was obvious most same-sex marriage supporters were refusing to vote.

It's notable also that a recent large LGBTIQ survey found that at least 56% of respondents wanted to fight in some way to win the plebiscite (despite almost all disapproving of it) compared to 15% who wanted a boycott.  This survey was not necessarily representative of all LGBTIQ people, being conducted through opt-in methods, but is very likely to have, if anything, leaned towards including those who feel most strongly affected by the issue.  (Rodney Croome has stated a follow-up is intended given the specific details of the form of the plebiscite that have now been released.)

Supporters should however be very understanding that some people affected by the plebiscite situation may feel - initially - that they simply cannot vote.  People are right to feel deeply disgusted about having their personal lives effectively held up for public judgement in this matter.  They may be especially bemused by the words of George Brandis, who argues that marriage is an intimate relationship and politicians cannot know the secrets of the human heart.  But instead of reaching the logical conclusion (that same-sex marriage is therefore no-one's business but that of the two people who want to marry), he instead concludes bizarrely that such "intimate" details should be outsourced to the judgement of millions of strangers. Er, what? Such is the tortured and pitiful logic to which the Coalition's desperation has driven it.

The Coalition's Motives

And on that, I think a little subtlety (but not too much) in discussion of the government's motives here shouldn't go astray.

The government has got itself into the sort of mess that governments get into when they make stupid promises for short-term political gain and then find that they can't admit their error.  The plebiscite has its origins in Tony Abbott's doomed attempt to save his leadership, but then Malcolm Turnbull seems to have been held to it as a condition of his own Prime Ministership.  The policy, while always unsound in any form (and not even considered important enough to be released as a formalised election policy), was pretty successful in neutralising the issue at the last election, because at least the Coalition were offering some kind of a way forward, which was seen as an improvement on endless blocking.

If Turnbull actually stood for anything much apart from being PM for the sake of it, I suspect he could have got same-sex marriage through the parliament at various points, at a very high risk of being brought down as leader and genuinely splitting the party.  Anyway, that didn't happen, and anyone wanting to get all partisan about this should remember that Julia Gillard was at least as bad if not worse.  Now, the Coalition under a moderate leader in a very grumbly political time finds itself at enormous risk of losing votes to One Nation, Australian Conservatives and other right-wing parties, at a time when it is already polling badly.  Such votes wouldn't even reliably return as preferences.

So this is one reason the Coalition is keener to keep this ridiculous "election promise" (if it even qualifies as such) than all the better ones it's ever broken.  The less savoury one is that the Coalition contains plenty of religious extremists who would rather destroy the party than let same-sex marriage pass without giving a platform to every anti-gay bigot in the country.  I don't think it is helpful to stereotype the Coalition's motives as a whole as anti-gay.  It is rather that the moderates are in a bind and unwilling to blow up the farm to get marriage equality passed.

The endgame of this isn't, realistically, stopping same-sex marriage forever.  The main aim is to keep the religious and redneck right in the fold (or retrieve those who have strayed) by pointing to the lengths the Coalition was willing to go to to tell those people that their concerns were taken seriously.  It is not that different to the Howard playbook of roping back in Coalition voters who defected to One Nation in 1998 by throwing them symbolic red meat on issues that weren't central to anything his government cared about from time to time.

No doubt I will have more to say about this vote should it actually go ahead.  A quick note that it sometimes takes me a while to approve comments on this site because of work commitments.

A Few Postscripts

Bernard Keane's Purist Nonsense

In today's Crikey email, Bernard Keane has mounted an argument that participating in the plebiscite is legitimising the idea that the government owns human rights and that votes of this kind are a valid process.

This kind of argument was crushed decades ago by Gough Whitlam with his famous comment that "only the impotent are pure".  It is frequently necessary to use mechanisms you disapprove of in order to get a result, and it doesn't "legitimise" them at all.  A similar example is the old Group Voting Ticket system for Senate elections.  Parties had to use Group Tickets or they could not poll a significant vote, but this did not mean all parties using that system agreed with it or legitimised it by filing a Group Ticket. Indeed some fought for many years to get rid of the system, and eventually succeeded.

It is entirely acceptable to use systems you don't agree with in seeking to overthrow injustice.  The most important thing is that in such a case one should criticise that system, and given the chance seek to prevent or discourage it from being used in the future.

Anti-SSM Tactic: Distract, Distract, Distract

It is notable that the anti-SSM crowd, particularly the Australian Christian Lobby and Tony Abbott, are trying to make the debate about all kinds of distractions - free speech, Safe Schools, gender, political correctness.  They seem to have grasped that if the debate is simply about same-sex marriage then their position is obviously stupid and mean, so there is a need to make the debate look like a wedge to something else.  These tactics were also seen aplenty in the Irish referendum, where they failed.  Hopefully they will also fail here.

Anyone claiming to support free speech while opposing the right of same-sex couples to meaningfully say their wedding vows is being very selective, at best.

Identification Issues

A further issue now being raised is whether ballots will be identifiable.  I think this issue may be being misconstrued, as the ABS site says only that there will be no personal identifiers on the survey form.  However, a privacy-protected postal system is still possible via the same method used in Tasmanian local government elections.  The reply-paid envelope is user-identified using a flap with user details which the voter signs.  The signed flaps are checked against the electoral roll, then they are detached and then the envelopes are opened and votes counted.  In theory someone might steal a vote and vote on someone else's behalf, but that is a mail theft crime, and also the original voter could then complain to the ABS and have their stolen vote stopped.  There would, however, be nothing to stop someone who was given (or sold) their vote by another person from voting multiple times.


  1. Comment from Michael Maley (also posted on Antony Green's site and Facebook) - Part 1


    The Census and Statistics (Statistical Information) Direction 2017, given to the Australian Statistician by the Treasurer yesterday to initiate this process, has a number of interesting features.

    Most notably, it does not require the ABS to acquire information by a postal process, nor does it require that all electors be approached for input: it simply requires the collection of “statistical information about the proportion of electors who wish to express a view about whether the law should be changed to allow same-sex couples to marry”, along with information about the proportion (of those electors who wish to express such a view) who are in favour or against. This information is to be compiled by State and Territory, and by electoral division.

    It would therefore be entirely open to the ABS to obtain this information using a well-constructed sample survey with a sufficient sample size in each electoral division to ensure reasonably tight confidence intervals.

    And in fact, that would be a better, cheaper and more accurate option than what the government seems to have in mind. Right now, presumably for legal reasons, what is being attempted is a plebiscite dressed up as a statistics gathering exercise, but entirely lacking the features which would validate it either as a plebiscite or as a survey. Plebiscites, like elections, depend for their validity on certain basic “free and fair” criteria, which I spelt out some years ago in a Senate Occasional Lecture, several of which look unlikely to be satisfied.

    “The process is administered impartially, and opportunities exist for complaints about the process to be lodged and dealt with in an even-handed and transparent way.” This is problematical as things stand now, with the use of the postal service potentially discriminating against voters less likely or able to use it, and with no rights of “voters” or scrutineers defined by law (and, in particular, no right to a transparent process for handling complaints). I might add here that the behaviour of the ABS in the aftermath of last year’s census night was hardly a model of transparency.

    “People qualified to vote, and only people so qualified, are able to do so.” This is problematical with a postal process with uneven coverage, especially in relation to voters overseas, and as things stand there will be no legal ban on impersonation, let alone proof of identity requirements.

  2. Comment from Michael Maley - Part 2

    “Votes are not bought and sold.” At present there seems to be nothing to prevent someone from setting up a market in which those uninterested in the issue can sell their “votes”.

    “Voters can cast a secret ballot, without fear of any adverse consequences.” The provisions in the Commonwealth Electoral Act which require even postal votes to be marked in private will not apply. People could be invited to mark their “votes” ceremonially in churches in front of the whole congregation.

    “Everyone votes only once.” The provisions in the Commonwealth Electoral Act making it an offence to vote more than once will not apply.

    These are major defects, of the type which would often lead international observers to deem an election not free and fair.

    Surveys, on the other hand, depend for their validity on a properly designed probability-based sample, along with appropriate weighting for non-response bias. ABS runs these sorts of surveys all the time. What is proposed here, however, isn’t such an exercise, but rather a “survey” with a self-selected sample, with likely inherent biases associated with the postal process, for which no corrections are proposed. As a survey intended to give an accurate snapshot of opinion, it would be close to worthless.

    So the ABS really need to be held to account on this. They need to be asked why they will comply with the words of the Direction from the Treasurer by adopting an expensive and methodologically defective process, rather than a cheaper and more accurate one. The extent to which they are really undertaking this process in an independent, professional and transparent way should be urgently tested through an immediate Senate inquiry.

    1. And another comment by Michael Maley!

      In all of this, there’s quite some doubt about the role to be played by the AEC. In a comment on his blog today, Antony Green observed that “…clearly the ABS will be out-sourcing large parts of the conduct of the ballot to the AEC.” Senator Cormann’s media release MC 46/17 of 8 August 2017 states that “The ABS will make arrangements for the secondment of officers from the Australian Electoral Commission to assist the Australian Statistician with this process as required”, while his media release MC 47/17 of 9 August goes somewhat further, stating that “the ABS, supported by AEC officers as appropriate, will make relevant announcements about timetables and practical arrangements …”. The ABS, in a statement published today, has said that it “has formed a Taskforce to deliver the survey and will publish the results by 15 November. The survey Taskforce includes staff from the Australian Electoral Commission seconded under ABS legislation.”

      There are two problems with this. The first is that by themselves AEC officers cannot do all that much to help any more: the postal voting process has long been automated, so what the ABS would really need would be access to the AEC’s offices, procedures, IT systems, management structures and contracts. (They would also need access overseas to staff from DFAT and Defence who run postal voting outside Australia, including for deployed forces.)

      The second problem is that it is arguable that any such support from the AEC - indeed, any AEC support other than secondment of staff in circumstances which had them strictly on the ABS’s books and off the AEC’s - would exceed the AEC’s powers and functions under section 7 of the Commonwealth Electoral Act 1918. In particular, it is clear that the AEC’s power under paragraph 7(1)(d) of the Act “to provide information and advice on electoral matters to the Parliament, the Government, Departments and authorities of the Commonwealth” isn’t applicable in this case, as section 5 of the Act clearly defines “electoral matters” for the purposes of section 7 as “matters relating to Parliamentary elections, elections, ballots under the Fair Work Act 2009 or the Fair Work (Registered Organisations) Act 2009 and referendums”; and this doesn’t include what purports to be a statistical exercise conducted by ABS.

      In fact, it is clear from the Supplementary Explanatory Memorandum to the Electoral and Referendum Act 1992 (which inserted in the Electoral Act sections 7A and 7B, permitting services to be provided by the AEC on a fee for service basis) that that amendment was intended to cover a lack of power to provide such services under section 7. (“This clause inserts a new section 7A of the Principal Act to empower the Australian Electoral Commission, in a manner not inconsistent with the performance of its primary functions, to provide goods and services to other organisations or individuals (for example, providing a "scanning” service to State electoral authorities, or assisting in the conduct of an election other than a federal election).“ See

      Support under section 7A would be permitted, but it seems that the government has already decided that trying to make use of section 7A would be too legally risky. Not being a constitutional lawyer, I have no view if it would make any difference if such an arrangement were between the ABS and the AEC, rather than between the government and the AEC.

      This makes the conduct of the process by ABS even less plausible. It will be interesting to see if the litigants planning to challenge the process will seek to have the courts restrain the AEC from taking any actions which would exceed its statutory powers.

  3. That's a fascinating thought Michael. I wonder if things may play out, say, should the Court find various gaps in statutory powers of the ABS, or with statutory usage of the electoral roll, or with potential conflicts with privacy laws which government agencies must adhere to arising from the practicalities of the proposed "voluntary survey" (the ABS' terminology in their press release today), that at some point the ABS finds that it must urgently resort to good old statistical methods to answer the ministers' Directive, and just do a quality 10,000 respondent poll instead.

    Come to think of it, I wonder what the professional statisticians at the ABS think of the very idea of a "voluntary survey" as a means of gathering statistics? Since the Senate is apparently launching an inquiry, the leading staff at ABS could be asked their professional opinion on the matter.

    Another question is what professional commentary the ABS ultimately adds to it's report on the Directive should the raw survey results represent only 40% turnout, geographically and demographically clumpy, etc. If the ABS adds caveats that the nominal/raw survey result is unreliable, or better still attempts to scale and weight the raw numbers so as to meaningfully return an answer to the Directive, things would become even more absurd.

    I fear that peak stupid may still be some way off with this whole sorry business.

  4. Due to the excellent commentary here, I have decided to call the federal government's bluff and participate by voting 'yes' to extending the rights and obligations of the Commonwealth Marriage Act to same-sex couples.

  5. Another Michael Maley comment:


    Regarding voting by 16 and 17 year olds in the plebisurvey, the AEC has today issued the following statement:

    “There has been speculation that 16 and 17 year olds will be allowed to participate in the marriage survey. This is not correct. The survey instruments will be sent to those who are on the roll; 16 and 17 year olds are not on the roll.

    The normal enrolment processes apply – 16 and 17 year olds are not added to the Commonwealth electoral roll until they turn 18. They are provisionally’ enrolled until they turn 18. Under section 90B of the Commonwealth Electoral Act the AEC only provides the details of electors that are enrolled, not provisionally enrolled. Section 90B is the AEC’s authority to provide the roll to the ABS, therefore 16 and 17 year olds will not be included in the Australian Marriage Law Postal Survey.”

    There are real problems with this AEC statement.

    First, the Commonwealth Electoral Act 1918 recognises no concept of "provisional enrolment" for 16-17 year olds. Subsection 98(1) of the Act states that "Names may be added to Rolls pursuant to ... claims for age 16 enrolment".

    Subsection 93(3) states that a person on a roll who is less than 18 cannot vote at an election. The existence of this provision clearly implies that 16 and 17 year olds can be on the roll, since if that were not the case, subsection 93(3) would be unnecessary.

    Subsection 93(4) says that in relation to an election, a person is deemed not to be enrolled unless 18 by polling day. This deeming provision is limited in scope, as it operates only “for the purposes of this Act in its application in relation to an election”. It has no explicit relevance to a plebisurvey.

    If a claim for age 16 enrolment is “in order”, s 102(1)(b) of the Act requires the claimant to be added to the roll "without delay". Subsection 100(2) of the Act makes it clear that this requirement applies to a claim for age 16 enrolment “as if the person making the claim were 18 years of age”.

    Finally, an AEC officer who fails to enrol a person when required to do so commits an offence against section 103 of the Act.

    So it would appear that if the AEC hasn't been adding 16 and 17 year olds to the roll, something has gone badly wrong.

  6. As to your question "It's ridiculous, so should we boycott it?", one of the birds in the latest First Dog has the answer. It's holding up a placard that says


    So yes it's ridiculous, but assuming the High Court doesn't stop it (and my guess is they won't), then everyone should vote! If our elected representatives are so weak that they want our guidance, then we should give it to them - as many of us as possible. And if we think they should have made up their own minds, that's something we can bear in mind next time we vote.

  7. First Dog cartoon may be seen at

  8. Oh, and an insight into the conservative mind (you know, those with brains that are fearful of change and can't process complex ideas - maybe fearful of change _because_ they can't process complex ideas). On the phone to an old frenemy of mine last night, he said "So a Yes vote will be a victory for Malcolm eh?" I would have thought that even calling the "plebisurvey" was an admission of hopelessness by Malcolm, but if his supporters think that way, there's a good source of Yes votes.

  9. I worry a little that if a consensus emerges that a No vote will be a crucial defeat for Turnbull, then some people may vote No just in the hope of bringing him down.

  10. Ahhh Kevin - but the answer to that is that a No vote is a vote for the return of Tony Abbott. He as good as said so. The Yes people should use that line: "Vote Yes if you don't want Tony to come back as the leader of the Libs"


The comment system is unreliable. If you cannot submit comments you can email me a comment (via email link in profile) - email must be entitled: Comment for publication, followed by the name of the article you wish to comment on. Comments are accepted in full or not at all. Comments will be published under the name the email is sent from unless an alias is clearly requested and stated. If you submit a comment which is not accepted within a few days you can also email me and I will check if it has been received.