‘Yet each man kills the thing he loves.’ The new Special Minister of State, Mal Brough, would appear at this point to have confirmed Oscar Wilde’s observation by his early musings on the need for reform of the Senate electoral system, which led to a predictably fierce response from several cross-bench senators, and some rapid back-pedalling from the Prime Minister, who clearly believes that now is not the time for making new enemies. Paradoxically, however, these exchanges have served to highlight the power that the cross-benchers enjoy, and to provide a further justification for careful thinking about the electoral sources of that power.

It would be most unfortunate if the Minister’s intervention proves fatal to Senate electoral reform, as the current system continues to face a crisis of legitimacy. This post outlines some of the different ways illegitimacy undermines the current system before looking at whether the proposed reform would address these. I will also discuss briefly the related but often overlooked concerns arising from the current system of filling Senate casual vacancies.

Genesis of the current reform proposals

Major Senate electoral system reforms have been seen only rarely: in 1948, when ‘single transferable vote’ proportional representation (PR) was introduced (replacing the majoritarian system which had been used in various forms since federation), and in 1984, when the ‘ticket voting’ system still in use was tacked onto the underlying PR system.

In 2014, the Parliament’s Joint Standing Committee on Electoral Matters (JSCEM), in its report on Senate voting practices, unanimously recommended that there be another major change to the system – replacing ticket voting with the ‘above the line preferential’ system now used for New South Wales Legislative Council elections. The recommendation highlights the extent to which recent events have been perceived as putting the current model under strain. Of particular significance was the 2013 election, which as I have noted elsewhere, was remarkable in a number of ways.

  • The 40 vacancies were contested by a record number of candidates, 529.
  • The percentage of votes polled by parties already represented in the Parliament dropped significantly from 2010.
  • In five out of the six States, a candidate was elected from a party which had never previously been represented in the federal Parliament.
  • For the first time ever, the seats in one State, South Australia, were divided between five different parties.
  • In Victoria, a minor party candidate was elected after having polled only 0.5 per cent of the first preference votes cast in the State.
  • In Western Australia, the Senate election was declared void by the High Court sitting as the Court of Disputed Returns (Australian Electoral Commission v Johnston) and had to be re-run in early 2014, after it had been found at a recount of the votes cast in 2013 that a number of them had been lost, on a scale sufficient to have had the potential to affect the result.
  • Allegations were reported in the media that some of the parties listed on the ballots had only a formal existence, and had been created for the purposes of exploiting opportunities for strategic manipulation of the ticket voting system.

For some commentators, this confluence of unusual developments sufficed to prove that a crisis was at hand; but a consensus on that point does not necessarily imply that causes and appropriate solutions will be obvious. To its credit, the JSCEM put considerable effort into analysing these, and came up with proposals which, if implemented, will produce the best electoral system ever used at Senate elections.

A deeply rooted problem

To see how this ‘perfect storm’ developed, it is necessary to go right back to the introduction of PR for the Senate in 1948. The single critical decision taken then, which looks like a point of minor detail but in fact has fundamentally influenced all later developments, was that voters would be required to indicate preferences for all the candidates on a Senate ballot paper in order to cast a formal vote.

Prior to the introduction of PR, it was virtually impossible for a candidate to be elected to the Senate from outside the ranks of the major parties. With PR, however, the Senate was transformed into a feasible battleground for minor parties as well, as became apparent with the ALP split in the mid-50s, and the rise of the DLP. Successful minor party forays encouraged further candidacies from outside the mainstream, the number of candidates per vacancy trended upwards, ballot papers became larger, and the task faced by voters in numbering them all became more onerous. With the passage of time, voters came to be ever more dependent on ‘how-to-vote’ cards issued by the parties to get their numbering right (especially since party affiliations of candidates were not printed on ballot papers until 1984), but the informal vote percentage nevertheless grew inexorably, topping 9 per cent nationwide at every Senate election from 1970 to 1983. The phenomenon reached its nadir at the 1974 Senate election in New South Wales, where 73 candidates stood for ten seats, and every voter had to number every candidate on the ballot paper: the informal vote reached 12.31 per cent.

The Hawke government, when it came to power in 1983, perceived the paramount need to address the problem of informal voting at Senate elections, but was constrained (by lacking a majority in the Senate) from being able to introduce its proclaimed policy of optional preferential voting, which would have relieved voters of the obligation to number every candidate. It accordingly opted for the current ticket voting scheme, which in effect enables a voter, by the marking of a single square on the ballot paper, to adopt in total the how-to-vote card of his or her party (as formally lodged with the Australian Electoral Commission).

A further quaint feature of the system, included to meet the particular political needs of the Australian Democrats, is that a group of candidates may choose to lodge two or three tickets in lieu of one, in which case the votes cast with the intention of adopting the group’s ticket are divided evenly between the two or three options, so if there are three tickets lodged by a group and 30,000 votes are cast for that group, the law deems 10,000 votes to follow each of the tickets. Finally, voters retain the option of indicating their own preferences for candidates, but those who do so are still instructed by the ballot paper to number every square. As there are two fundamentally different modalities for marking a vote, the ballot paper is divided in two by a large horizontal line, with squares for ticket voting appearing ‘above the line’, and individual candidates listed ‘below the line’.

Because it was apparent from previous election statistics that the vast bulk of voters had been following how-to-vote cards anyway, the change was not seen as being a particularly momentous one. But one important point was overlooked, and this was the issue which came to a head in 2013. Up until 1984, the only parties which were able to issue how-to-vote cards were those which had the membership base, field structure and resources to enable them to distribute cards physically at polling places. With ticket voting, on the other hand, every group of candidates on the ballot paper could provide voters with, in effect, a ‘virtual’ how-to-vote card. This ultimately led directly to the phenomenon of ‘preference harvesting’, which enables a host of ‘micro-parties’ to exchange preferences with each other for their mutual benefit.

Preference harvesting as a technique was first refined at New South Wales Legislative Council elections in the 1990s. But as the percentage of the vote required to elect a candidate there (the ‘quota’) was much lower than at Senate elections, it took longer for the micro-party vote to build to levels at Senate elections at which preference harvesting would be feasible. The 2013 Senate election was the first at which such a strategy demonstrably succeeded.

While preference harvesting may ensure that at least one candidate from the parties sharing preferences will get elected, which candidate from which party will ultimately be successful is very much a matter of luck, being greatly influenced by the order of exclusion of candidates during the count, which in turn is heavily dependent on the (all relatively small) numbers of first preference votes they poll. It follows that a slight change in the first preference tally of one candidate can have major flow-on effects. Elsewhere I have described this as follows: ‘in effect, … [micro-party] candidates buy a lottery ticket, the price of which is the cost of the deposit, with first prize being six years in the Senate’.

Legitimacy Problem Mark I – True and false preferences

All of this having been said, it is by no means clear that the election of candidates with small first preference vote percentages on the strengths of preference flows to them constitutes a problem: indeed, the possibility of such a thing happening is an essential feature of single transferable vote PR. The problem rather lies in the fact that it is most unlikely that the ‘preferences’ processed by the system are in fact held by the voters.

This is so because ballot papers these days are largely clogged up with obscure candidates, sometimes running for parties whose beliefs, as professed in their names, cannot necessarily be taken at face value. Many of these candidates scarcely campaign, leaving voters with no basis for assessing their relative merits. Faced with such ballot papers, voters who wish to specify their own preferences rather than adopting a ticket as their vote have no option but to lie, writing random or otherwise meaningless numbers against candidates once they have run out of genuine preferences to express. For voters who wish to use the ticket voting option, the proposition that they have consciously adopted as their own personal preferences for each candidate those contained in a ticket is little more than a legal fiction.

It follows that it is entirely possible that a majority of the preferences which the Australian Electoral Commission enters into its computer system for the counting of Senate votes are spurious, in the sense that they do not correspond to a genuinely held belief on the part of the voters. This, more than any other factor, is the source of the crisis in the legitimacy of Senate elections.

Legitimacy Problem Mark II – Discrimination against some voters

The second fundamental problem with the current system is that it discriminates against some voters, by making the act of voting comparatively difficult for them. This is so, at least, for any voter who wishes to cast a first preference vote for a candidate who does not appear in the first position on a ticket; such voters are required to number every square on the ballot paper.

One can give as an example here the case of a voter who wishes to vote for all the female candidates before all the male ones. This issue of discrimination was litigated early in the life of ticket voting (McKenzie v Commonwealth). While a single judge of the High Court declined on that occasion to issue an injunction restraining the conduct of the 1984 Queensland Senate election, it might be noted that there were only 29 candidates for that election, in contrast to the 110 who contested the 2013 Senate election in New South Wales.

It would seem at least arguable that there must come a point at which the number of preferences required to be expressed by a non-ticket voter would be such as to compromise the genuineness of the ‘choice’ which Senate voters make under section 7 of the Constitution.

Legitimacy Problem Mark III – Unmanageable ballot papers

At the 2013 election, the Senate ballot paper in New South Wales met a particular benchmark of absurdity: magnifying sheets had to be provided at polling places to enable voters to read the print, which had been reduced in size to enable all the candidates and groups to be shown. This highlights that mechanisms for imposing reasonable controls on ballot access – which are a feature of virtually all democratic elections – are fundamentally failing.

Antony Green has noted that:

The tactic of the micro-party alliance is to nominate candidates in every state to create giant ballot papers that encourage people to just vote one, to increase the pool of votes in the ‘other’ category, and to then toss them around from party to party using group ticket votes. The current electoral system encourages parties to enter the lottery for the final seat and encourages the fractionalisation of parties that might otherwise coalesce.

The JSCEM’s proposed system change

Under the ‘above the line preferential’ system proposed by the JSCEM, the voter can place not just a mark in one square above the line, but further preference marks as well in other above the line squares. Where a person so votes ‘above the line’, he or she is deemed to have expressed preferences for candidates as follows.

  • In relation to the group in whose square the number ‘1’ is placed, the voter is deemed to have expressed a first preference for the first candidate in the group, and subsequent preferences for the other candidates in the group in the order in which they appear on the ballot paper.
  • Numbers written in other squares above the line are deemed to imply subsequent preferences for candidates of additional groups, again in the order which the candidates appeared within the group on the ballot paper.

The outstanding feature of the JSCEM’s main proposal for reform is that it addresses (to a greater or lesser extent) every one of these fundamental sources of illegitimacy. The Committee’s proposed optional preferential voting scheme would eliminate tickets to be adopted in toto by voters, and instead would require preferences to be ones determined and expressed by the voters themselves. With the requirement for full preferential numbering abolished, voters would no longer have to fill their ballots with meaningless numbers purely to ensure that their genuine first preference would be counted; they would, in other words, be able to vote truthfully. While there would still be some disparity between the ease with which different voters could vote, depending on the preferences which they wished to express, it would be vastly less than is currently the case. Finally, the possibility of preference harvesting strategies would be eliminated, and this could be expected to lead to a reduction in the number of groups and candidates, particularly if more rigorous ballot access mechanisms (such as increased deposits) or a changed ballot paper design were introduced.

In addition, while there are some points of detail yet to be resolved, it is highly likely that any marking of a ballot paper which is currently formal would continue to be formal under the reformed system, which would substantially reduce the risk of voter disenfranchisement.

Of course, in any major electoral reform, there will be winners and losers, and few will be surprised that governments tend to avoid system changes under which they will be losers. This, by itself, is not an argument against modifications which can be justified on legitimate democratic grounds. Certainly the JSCEM’s proposal will make it harder for micro-parties to win seats through preference harvesting, but except from the perspective of those micro-parties’ candidates, that would arguably be a good thing. It has also been suggested that this change might tend to prevent new minor parties from emerging and consolidating. This argument is rather more problematic: the major element of luck involved in preference harvesting could well have the effect of seeing candidates from different micro-parties elected at each poll, which would limit the extent to which each such party might be able to consolidate its position.

It should be noted also that the JSCEM’s recommendations flowed from consideration of a wide range of reform options, some of which are discussed in detail in its report. For example, it considered, but decided against recommending, the deceptively simple proposal that a party should have to receive a specified minimum percentage of first preference votes in order to be eligible to win seats. (My own submission to the Committee dealt mainly with that suggestion.)

The elephant in the room – Senate casual vacancies

Finally, it is worth mentioning that for all the questioning since 2013 of the election of certain senators through preference harvesting, at least those senators appeared on a ballot paper and received some votes. Increasingly, however, the Senate has been graced by the presence of senators who, having been appointed by State Parliaments to fill casual vacancies, have not necessarily even appeared on a ballot paper.

Antony Green has explored this issue in some detail, and cites the case of Senator Santo Santoro from Queensland, who, having been appointed to fill a casual vacancy, ‘served as a Commonwealth Minister without ever facing election’. Green documents how the 1977 amendments to the Constitution have contributed to the ‘revolving door’ of Senate casual vacancy appointments, and one might add to his examples the case of Senator Bob Carr, who was elected in 2013 from the first place on the ALP ticket for a term due to begin on 1 July 2014, but resigned his seat on 24 October 2013, barely weeks after the declaration of the poll. He was replaced by (now) Senator Deborah O’Neill, who had not been a candidate for the Senate at all, but was rather a defeated ALP House of Representatives candidate, and whose term in the Upper House is now due to expire on 30 June 2020. One could hardly hope for a better example of ‘bait and switch’ in electoral processes.

While the filling of casual vacancies is governed by section 15 of the Constitution, one step within the power of the Parliament would be to insert a provision in the Commonwealth Electoral Act 1918 requiring the Australian Electoral Commission, when a casual vacancy occurs, to conduct a recount similar to those used to fill casual vacancies in the Tasmanian House of Assembly and the ACT Legislative Assembly. The outcome of the recount would then be advised to the relevant State Parliament.

Since all of the ballot details are captured electronically as part of the current Senate election process, the costs (with the possible exception of those associated with writing the computer programs needed to do the recount) would be trivial.

While the result could not legally bind the State Parliament in the exercise of its constitutional discretion, and could be subverted by a party if it so chose, at least the Parliament or the party would be put under some pressure to justify a decision to ignore the recount result. (There might be cases in which the winner of the recount was not from the party of the departed senator, in which case someone else would have to be appointed; but they would probably be relatively rare.)

There is a precedent for a process along these lines in section 282 of the Commonwealth Electoral Act 1918, which requires the Australian Electoral Commission to conduct a recount after a double dissolution election to obtain information which is provided to the Senate to assist it in the exercise of its discretion under the Constitution to identify which senators will have long and short terms.

The need for action

Some 16 months have now passed since the JSCEM made its recommendations for change, but we appear to be no closer to seeing any serious attempt on the part of the government to deal with the systemic problems which have been clearly identified. However, the matter should be treated as urgent for two reasons.

First, it needs to be borne in mind that the unprecedented candidate numbers and ballot paper sizes seen in 2013 arose even before the results of that election proved that preference harvesting could work at Senate elections. Now that the genie is out of the bottle, there is simply no knowing how many more candidates might decide at the next election to ‘buy a lottery ticket’ in the hope of winning a life-changing prize.

Secondly, any system changes need to be able to be implemented, and since Senate counting has been computerised since the 1990s, this means that the Australian Electoral Commission will need time after any reform legislation has been passed to reprogram and thoroughly test the software which it uses. (The alternative, of trying to revert to the sort of hand counting which applied previously, would give rise to a host of other difficulties too complex to discuss in detail here.)  That having been said, the system changes associated with the JSCEM recommendations do not appear to be overwhelmingly daunting, and a computerised counting system which embodies the proposed system already exists for use at New South Wales Legislative Council elections.

Finally, it needs to be emphasised that both of these challenges will be greatly magnified if the next Senate election follows a double dissolution. In such a case there would be twice as many vacancies in the States as at a normal half-Senate election, and a lower quota for election, both of which factors would tend to encourage more candidates to run. And the newly elected senators would have to be ready to take their seats not later than thirty days after the day appointed for the return of the writs, rather than on 1 July 2017.

Michael Maley is one of the ACT Co-Convenors of the Electoral Regulation Research Network, and a member of the Editorial Board of the Election Law Journal.

Suggested citation:  Michael Maley, ‘Senate Electoral Reform’ on AUSPUBLAW (29 September 2015) <>.