How Not To Do Hare-Clark Public Funding Threshholds In Tasmania

Updates added during debate in November (at bottom)

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

In late 2021 the Tasmanian Government released draft Bills concerning electoral funding and disclosure and miscellaneous electoral matters, and called for submissions.  I sent in a submission, which was duly published, but my advice concerning a major flaw in the electoral funding model was then completely ignored in the version of the Bill introduced in parliament and the accompanying second reading speech.  (The Bills were actually introduced in May - unnoticed by me at the time - but the Electoral Disclosure and Funding Bill has just reached the second reading debate stage).

Not to put too fine a point on it, I am getting really jaded with writing submissions.  I keep having to spend my spare time (which does not exist) writing submissions for nothing to try to ensure that governments do not make serious errors, only to find that in this case it does not even stop the mistake being made.

If the submission process does not result in the correction of howlers that are pointed out in submissions then what is the point of calling for submissions and what is the point of writing them?  The Tasmanian Government, for some reason that has never been explained, even sets the deadlines for public submissions at 5 pm, which makes meeting submission deadlines just a little bit harder for people who have to juggle countless other things while trying to write them.  Is someone seriously going to start working on them at 5 pm?  Would it really matter one iota if the submission was received at 10 pm or even 7 am the next day instead?  (Federal JSCEM submissions have a midnight deadline.)

Public Funding Threshhold

Anyway, my main concern here is with the way eligibility for public funding (which will be new for Tasmania) will be determined. The current Electoral Funding And Disclosure Bill has the following as Clause 132 (was 124 in draft):

132. Registered parties eligible for public funding of Assembly election campaigns 

A registered party is, subject to and in accordance with this Act, eligible for payments from the Election Campaigns Fund in respect of an Assembly general election, or an Assembly by-election, if – 

(a) it is a registered party on the polling day for the Assembly general election or the Assembly by-election; and 

(b) there was, in relation to the election, an Assembly candidate who was, immediately before the polling day, an Assembly candidate endorsed by the registered party; and 

(c) either (i) the Assembly candidate is elected at the Assembly general election or the Assembly by-election; or (ii) the total number of formal first preference votes received by the Assembly candidate is at least 4% of the total number of formal first preference votes in the division in which the candidate was duly nominated for election.

Clause 133 then goes on to pay the party $6 per vote received by candidates who qualify under Clause 132, up to a maximum of full reimbursement of expenses for the party and all of its candidates (ie it is not possible to make a profit).  Clauses 134-135 are the same as Clauses 132-133 except that they apply to independents.

This mirrors the way in which parties are funded for House of Representatives, but applying the same model to Tasmanian House of Assembly elections is completely inappropriate.  Why?  Because Tasmania has Hare-Clark, and in Hare-Clark elections parties run multiple candidates per division.  Especially in the case of the major parties, the candidates compete against not only against other parties but against each other.  This free competition between candidates within a party is much valued by Tasmanians as it gives them a chance to choose to replace MPs within a party without needing to change the party that they vote for.  An extreme example was the 1986 Tasmanian election, at which there was virtually no change in the party makeup of the Assembly but 13 of 33 recontesting MPs were replaced by new MPs from the same party.

The proposed model, however, means that whenever a party runs a candidate who polls less than 4% of the vote and is not elected, that party will receive no public funding for those votes.  This especially affects the Greens, because in some divisions it is currently touch and go whether their lead candidate will receive 4%, but also because a higher proportion of their votes go to candidates who will not receive 4%.  But it could cause all kinds of undesirable distortions in the strategies of major parties and within-party tensions over electoral strategy as well. 

The following table shows which votes, for parties polling 4% or more within a division, would have been funded in 2021 if this system had existed and the same votes had been polled:

All the main three parties had significant support (over 4%) in every division but there are great discrepancies in the share of vote that is eligible for funding by party.  Leading independents get funding for all of their votes, the Liberal Party gets funding for over 90% of its votes, but the Greens get funding for just over 60% of theirs.  But in one division (Clark), Labor is even more disadvantaged than the Greens, getting $47988 for its 22.06% of the vote while the Greens get $56814 for 20.03% - just because the Greens' vote was more concentrated in their lead candidate.  

The unfair advantage to the Liberals over Labor is because they were the best performing party on primary votes - this means their candidates are individually more likely to either break 4% or win, and means a higher proportion of their votes are funded (as if being eligible for more funding just by virtue of getting more votes would not be advantage enough).  If I applied the same model to 2002 and 2006 I expect I would find Labor advantaged over the Liberals in the same unfair manner, but under all circumstances the model will unfairly advantage independents who can get 4% over parties, and will unfairly advantage both major parties over the Greens.  

The following are some possible negative strategic impacts of the change, most of which have been identified (with one miswording corrected) in my submission:

* Because small parties like the Shooters will not get any funding if their vote is split across multiple candidates (eg Lyons 2018) they will be encouraged to only run one candidate per division.  But when minor parties run a less than full slate of candidates the informal vote by intending voters for those parties rises (in part thanks to disinterest in savings provisions, but that's another story.)

* Green voters will be discouraged from voting for minor Greens candidates (assuming the party even runs them) since these candidates will not reach 4% and every vote that is for them and not for the lead candidate is likely to cost the party $6.  But it could be a lot more than that: in Lyons 2021 the Greens' lead candidate cleared the 4% bar by 42 votes.  Under this model, had 42 voters for the Greens voted for minor candidates instead, those 42 votes would have cost the Greens $17298.

* Since major parties get no funding for candidates who poll below 4% without being elected, there is an increased incentive to ensure "filler" candidates get as few votes as possible and the vote is concentrated in the major candidates for the party.  This increases the risk that where those candidates are later elected on recounts, voters will know nothing about them. It also affects preselections by discouraging parties from preselecting candidates who they fear will poll substantially without reaching 4%.

* There is a possible incentive for spoiler candidates such as fake independents to be run to try to drive small parties' lead candidates (eg at present the Greens in the northern divisions) below 4% and thereby deprive those parties of public funding.  

The solution is extremely simple: 

For Hare-Clark funding to parties should be based on party votes not candidate votes. 

There is no excuse for a by-candidate model and no justification has been given for using funding by candidate for Tasmania.  The obvious thing for anyone wanting to draft a funding model for Tasmania to do would be to look at how funding is handled in the ACT which is the other jurisdiction with Hare-Clark.  The ACT issues funding by party: a party is funded if its candidates together poll 4% in a division.  This should obviously be adopted for Tasmania, and if desired the funding could be amended from $6 per vote to $5 per vote to compensate for the increased number of votes eligible for funding.  

I would prefer not to have to write to all members of the Legislative Council for the first time in a decade to explain that a Bill that has passed the Assembly is dangerously unfit for purpose.  I hope that the Government will now amend its own Bill to fix this incorrect funding model which is entirely contrary to the spirit of Hare-Clark elections.  

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

Update (10 November)

The Bill was debated today with Ella Haddad (ALP), Cassy O'Connor (Greens) and Kristie Johnston (IND) all referencing this article and/or my submission and with the following further comments:

* Haddad said Labor would move an amendment on this point.  

* O'Connor said the Greens would move an amendment to simply abolish the 4% threshhold (so that all votes received would be eligible for public funding). (This in my view would be far better than the current Bill but I think it would encourage too many uncompetitive candidates to run.)

* Johnston stated that she opposes public funding of candidates and parties, unless accompanied by a complete ban on donations, and also stated that if the Bill was to go ahead with public funding it should be amended to address the issue.  

* Elise Archer in defending the Bill made the following comments which I have transcribed more or less in full:

"In response to the criticism of Dr Bonham, I would respectfully offer the following - so while I acknowledge the incredible and extensive knowledge and understanding of Dr Bonham in this area I don't agree that the Bill will have the dire consequences that have been foreshadowed.  The Bill was drafted so as to use the 4% per candidate model to reflect the fact that under our system we do not have above the line voting for the House.  Although the ACT shares our Hare-Clark system they do also have an option for above the line voting  This fact makes it impossible to ascertain which specific candidates received a first preference vote from each elector if that elector chose to vote above the line.  

It was therefore decided that the 4% per candidate model was more appropriate.  In relation to Dr Bonham's belief that the 4% per candidate model would resist, sorry would result, in parties actively seeking to have voters not vote for their own members, I can't see that this would be so.  

So I would encourage Ms Johnston before she comes in here to actually do the research, it reminds me of [recounting of Johnston's apology to Matt Groom snipped - KB]. It's really important to get facts straight particularly in such a detailed debate."

Attorney-General Archer's statements about the ACT system are false and have (I assume inadvertently) misled the House.  The ACT does not have above the line voting, and has not had above the line voting since it was abolished in 1994.  The ACT has never had Hare-Clark and above the line voting at the same time.   An attempt by the Follett Labor government in 1993 to graft above the line voting onto Hare-Clark fortunately failed, with crossbenchers Michael Moore and Helen Szuty apparently refusing to negotiate further (and expressing a lack of confidence in the Chief Minister) until it was dropped.  

There is some truth in the origin story because in 1989-1992 the ACT had public funding alongside the much-ridiculed Modified D'Hondt system, which did have above the line voting.  However it has retained the by-party share model for funding that was initiated under that system ever since.  It has held eight elections in 28 years with Hare-Clark, without above the line voting and with public funding paid by party vote share.  All the votes are credited to individual candidates.  

But even if the ACT did still have above the line voting and was therefore constrained to choose by-party funding over by-candidate funding, that would still not make by-candidate funding in any way a good idea.  You don't choose to do something differently just because you can, and no argument for using a per-candidate threshhold rather than a per-party threshhold has been advanced.  (The money doesn't even go to the party candidates; it still goes to their parties).  We just have unsubstantiated faith that the system won't have any of the impacts I've foreshadowed (why on earth risk it?) when I have proven one of those impacts above by reference to the funding that would have been paid based on actual results.  I guess this isn't the first time politicians have thought they were above the laws of mathematics.  

Archer also subtly, and again I assume unintentionally, strawmans my position as being that the system would result in parties seeking to have voters not vote for their own members.  What I have said is that it will result in parties seeking to engineer which of their own members voters vote for, and in perverse impacts on candidate choice.  Would anyone for a moment believe that under such a system the Greens would have risked costing themselves over $17,000 (probably even over $20,000, since he would have taken votes off their lead candidate) in public funding by running Tim Morris as a support candidate in Lyons 2021 for example?  No, they would have run a nobody instead, or perhaps not even run five candidates.  

Following such an error-riddled contribution, to attack another MP for not doing research and not paying attention to facts was, well, just embarrassing.  Or should be.

Update (22 November): The error, which was a result of incorrect advice, has now been corrected on the record at the start of the continuation of debate on the Bill.