Written evidence submitted by Dr Graeme
Orr and Professor K D Ewing (PVSCB 10)
SUMMARY
This submission:
(i) briefly explains AV and its various forms;
(ii) recommends the Committee considers specifying
in advance which type of AV is being proposed, and suggests that
optional preferential voting is the better option;
(iii) highlights issues in referendum law which
may require attention prior to the referendum, notably the application
of expenditure limits to the media; and
(iv) alerts the Committee to Australian experience
of how shifting from first-past-the-post to AV may affect electors
and campaign practices. These regulatory questions are not pressing
prior to the referendum but would need to be faced if AV were
to be legislated.
INTRODUCTION
1. The Alternative Vote (AV) is a well established
system of voting, most closely associated in modern times with
Australia. It has been used for Australian national elections
since 1919, and is employed in six out of eight State and Territory
lower houses. Whilst it is not widely used elsewhere:
— AV is not an Australian invention. It
was developed by an American in the 19th century as a variant
of the "single transferable vote" (STV) which was developed
in the UK and championed by the likes of John Stuart Mill.
— AV has been on-and-off the UK policy
agenda for a century. It was recommended by a Royal Commission
on Electoral Reform in 1910, and proposals for AV were contained
in a (Labour) government bill introduced in 1930.
— AV is similar to, but cheaper than, runoff
or second ballot elections, a la France (and for this reason was
recommended by the Royal Commission in 1910 in preference to the
second ballot). In the USA, reformers who support AV call it "instant
runoff voting".
2. The essence of either AV or STV is that
the ballot enables the elector to rank candidates in order of
preference. STV is employed in multi-member electorates: if a
large number of members are to elected, it generates a form of
proportional representation. AV is preferential voting in single-member
constituencies. A sample Australian AV ballot can be found at:
http://www.aec.gov.au/Voting/How_to_vote/Voting_HOR.htm In AV,
if no candidate gains 50% of the "primary votes" (aka
"first preferences") the ballots of the least favoured
candidates are then scrutinised, and allocated according to their
supporter's second preferences. And so on until a majority winner
emerges.
3. AV is thus majoritarian, not proportional.
This is borne out by Australia's long history of stable government,
with a two-party system dominating its lower houses. AV's main
benefits over first-past-the-post are:
(a) Increasing electoral choice. By being able
to rank all candidates, electors have more choices. Supporters
of minor parties are less likely to feel their ballot is wasted.
Tactical voting and "vote-swapping" should all but disappear.
Party decisions to stand candidates will be less driven by tactical
fears of "vote-splitting". If electors feel their votes
have more saliency, turnout may improve somewhat.
(b) Increasing a sense of legitimacy in electoral
outcomes. (Though some will feel AV emphasises an electoral choice
between the least disfavoured rather than the most favoured).
4. AV will not disturb the UK's three-party
system, although depending on how electors' preferences fluctuate
and are expressed, it will probably enable the Liberal Democrats
to improve their vote-to-seat share. In some ways AV may actually
reinforce the "cartel" nature of the major party system.
We note ways in which AV emphasises the power of party machines,
in paragraphs 17-20 below.
WHAT TYPE
OF AV?
5. A central issue for Parliament in framing
the referendum is "which type of AV will be proposed?"
By this we mean, how many preferences will electors be required
to express to cast a valid or formal ballot?
(i) A minimum of one. This is known as optional-preferential
voting (OPV).
(ii) A full set of preferences—electors
must rank all candidates on offer.
(iii) Two. This is known as the supplementary
or contingent vote.
For reasons of principle and culture we recommend
OPV. In any case it is desirable to concretise the debate so that
the referendum not be conducted in the abstract, but the type
of AV be specified. We will briefly "rank" the three
types.
6. OPV gives electors maximum choice. They
do not need to choose between parties or candidates about whom
they may have no preference or information. They do not have to
make invidious choices between two extremist parties, or between
parties they may see as undifferentiated. OPV probably best suits
the UK's system of voluntary voting. It is used today in lower
house elections in New South Wales and Queensland. The downside
of OPV is that its outcomes may seem less majoritarian than if
full preferential voting were mandated. In a worst-case scenario,
it can fall back into first-past-the-post (eg if most electors
just plumped with "1" on their ballot). There is room
for a party which believes it would benefit through vote-splitting
to encourage this by appealing to the electorate to "just
vote `1'".
7. Full preferential voting forces electors,
especially minor party supporters, ultimately to choose between
the major parties. It is employed in Australian national elections
and three States. It may be justified by reasoning that since
someone must represent and govern for us, it is reasonable for
Parliament to require us to choose between the major parties.
However that approach fits better with the Australian system of
compulsory voting than more liberal UK traditions. Parties may
also prefer full preferential voting since it means their preference
recommendations are more likely to be followed, by supporters
unable or unwilling to differentiate all the options themselves.
Mandating full preferential voting is likely to increase the number
of wasted ballots, as some will make mistakes in numbering and
others protest at being forced to rank candidates they cannot
choose between. (Informal voting is likely to increase anyway,
for reasons discussed in paragraphs 15-16 below.)
8. Finally, although toyed with by the Plant
Report of 1993, the two-choice supplementary vote is especially
unappealing. It has been described as a "half-baked compromise".
It will not eliminate tactical voting: indeed an elector, given
only one preference other than her primary one, is effectively
asked to second guess which candidates will remain in the count.
Unlike OPV it does not liberate or maximise electoral choice,
but artificially restrains it. And unlike full preferential voting
it does not ensure majoritarian outcomes. A form of optional two-choice
or partial AV was included in the schedule to the 1930 Bill referred
to above. The supplementary vote was experimented with in Queensland
from 1892-1942, where it was known as the contingent vote.
REFERENDUM LAW
ISSUES
9. A couple of issues involving the conduct
of the referendum under the Political Parties, Elections and Referendums
Act 2000 (PPERA) need to be considered. One is that the Electoral
Commission is required to publish its views on the intelligibility
of the referendum question, yet the government does not appear
to be required to accept that advice. It is vital that the referendum
question be as simple as is reasonable whilst remaining descriptive,
but above all that it not be tendentious. This may be an issue
given that within the government—indeed within each of the
major parties—there will be political divisions over the
AV proposal.
10. A second issue concerns public financing
of referendum campaigns. The Electoral Commission has a key role
in designating two groups to spend public money, in effect on
a "yes" and "no" campaign. This is an especially
significant duty. In many ways the referendum is not really a
straightforward battle between supporters of first-past-the-post
and supporters of AV, since there are other voting systems that
could have been proposed, notably proportional representation.
When multiple choice is limited to binary alternatives, the status
quo can have an undue and artificial advantage: for example
supporters of first-past-the-post may make odd bedfellows with
supporters of proportional representation, to defeat the "compromise"
option of AV. (Similarly, in the 1999 Australian Republican referendum
monarchists and those seeking a directly elected President defeated
the compromise option of a parliamentary appointed head of state
in a two-way vote, even though a majority of Australians were
republicans. Given that the UK referendum is really only an indicative
vote, in principle it ought ask citizens to rank first-past-the-post,
AV and proportional representation: but this of course would increase
the complexity of the debate).
11. A third issue concerns the policing
of expenditure limits. The official or designated campaigns can
spend up to £5 million each; the parties have a similar limit,
though due to vote shares in 2010, only the Conservatives are
entitled to the full limit. Third parties are limited to £500,000.
Curiously, unlike in the spending limits that apply to elections,
the definition of referendum expenses does not expressly exclude
"the publication of any matter relating to an election, other
than an advertisement, in| a newspaper or periodical". On
the contrary, the list of referendum expenses refers to "any
material to which section 125 applies": this includes any
material which provides general information about the referendum,
or puts any argument for or against any particular answer to the
referendum question. The Act would thus appear to limit newspapers
to "spending" at most £500,000 each in providing
information or in advocating one position or another during the
referendum campaign period. Moreover, newspaper companies will
be able to do so only if they comply with the registration or
notification procedures necessary to be a permitted participant;
otherwise they will only be able to spend up to £10,000.
12. It is not entirely clear whether the
drafters of PPERA intended this result; although given the concentrated
power of newspaper proprietors there is no reason in democratic
theory why they should have unlimited rein to campaign when parties
and other civic groups do not. Once this apparent oversight becomes
known, there will no doubt be great pressure from the newspaper
companies to change the law to their advantage. An alternative
approach might be to amend the law the other way to apply the
existing referendum rules in PPERA so that they apply also to
general elections.
IMPACTS OF
AV
13. In this final section, we sketch the
key implications of AV from the Australian experience. We do this
to inform the Committee and debate. However regulatory responses
to these matters need not be faced until after the referendum,
assuming AV is to be legislated in practice. The implications
can be grouped into two sets of effects:
Electors—complexity and votes wasted to
informality
14. The key effect of AV on electors is
in widening choice. However this benefit is not cost free. The
downside of increased choice is the potential for confusion. Informal
voting in the UK under first-past-the-post is almost negligible,
ranking 146th out of 146th in a recent international survey: under
AV, Australia ranks 46th, with informal voting of over 3% of turnout
being common at national elections. Informality, in the Australian
experience, tends to disproportionately disenfranchise electors
in constituencies with lower levels of education or higher levels
of non-English speakers, where rates of informality tend to be
higher than the 3% average. Compulsory voting in Australia probably
accounts for about 1% of that 3%: a proportion unlikely to be
encountered under voluntary voting in the UK.
15. Informality in the UK will be exacerbated
anyway by the introduction of a new voting system on top of the
plethora of other voting systems used in European and sub-national
systems, and the possibility of another system being adopted for
the upper house at Westminster. Again, Australian experience is
instructive: when voting systems change, or when electors are
faced with several different systems, informality increases.
16. We do not wish to overstate the informality
problem. Compared to the sense of votes being wasted under AV,
informality is a regulatory problem rather than an argument against
AV per se. Education and ballot design is crucial in minimising
such wasted votes; although it is far from a perfect panacea.
The need to minimise disenfranchisement through informality however
is another argument for OPV over full preferential voting—since
requiring electors to rank all candidates on offer increases the
likelihood of mistakes, especially if, as is likely under AV,
more candidates take part.
Parties—campaign practices and preference
deals
17. The least appreciated aspect of AV in
the UK is its likely effect on campaign culture. The Australian
experience is salutary. The key marginal seats, which decide the
fates of governments, hinge on preference flows. Whilst AV expands
electors' choices, it also—somewhat paradoxically—also
empowers the party machines. There is a long history of learning
and interchange of staff between Australian and British parties
(especially between the Labour Party and the Australian Labor
Party, and the Conservative Party and the Liberal Party of Australia):
it seems inevitable that Australian practices under AV may at
least be trialled in the UK.
18. In Australia, preferences have become
a crucial form of political "currency" and are the subject
of deal-making between the parties. A typical example, unfolding
as we write, involves a deal between Australian Labor and the
Australian Greens: http://www.smh.com.au/federal-election/greenslabor-preference-deal-20100718-10g33.html?autostart=1
Note that in Australia, a minor party typically looks to barter
its lower house preference recommendations for a flow of preferences
to it in the upper house, where it is more likely to be in the
hunt for seats. This would be replicated if the UK moved to AV
for the Commons and proportional representation through STV for
the House of Lords. But even with just AV in the Commons, deals
will surely arise since the UK is not a simple two-party system
(like the US): including the nationalist parties and now the Greens,
there are at least six parties seriously vying for seats across
Britain alone, and parties will seek to trade their preference
recommendations in seats where they are weak for preference recommendations
in seats where they are competitive.
19. This deal-making can have upsides. Indeed
one of the benefits of AV is that it forces the major parties,
and especially the parties hoping to win government, to take more
seriously the views of those inclined to support other parties.
Preference deals can bring minor parties in from the cold: for
instance a party like the Greens or UKIP may generate policy leverage
if it has sufficient support in key marginals to affect the outcome.
Such developments may challenge traditional UK political sensibilities
and look too much like log-rolling. Outside nationalist politics
the focus in the UK until now has been very much on the major
parties alone.
20. The classic method of communicating
a party's preference recommendations is via "how-to-vote"
material: impressions of a ballot paper, ranked as the party suggests,
distributed to electors through advertisements but most commonly
as flyers distributed outside each polling station.
DEAL-MAKING
AND HOW-TO-VOTE
MATERIAL RAISES
SEVERAL REGULATORY
ISSUES:
(i) Electoral education. It is important
that electors realise that the ranking of candidates is entirely
up to them. Party how-to-vote material is a guide only. Yet the
media typically talks as if parties "directed" or "allocated"
preferences. Parties do little to correct this misperception:
it generally suits them to have a perception that their preference
recommendations are decisive.
(ii) Ethics and electoral bribery. Some
deals, such as purchasing preference support, are corrupting.
Policing such arrangements is difficult when they occur in private;
simpatico parties also share resources innocently. The current
Representation of the People Act would need amendment to extend
the notion of vote-buying to preference deals.
(iii) Misleading campaign practices. A routine
and lamentable feature of Australian campaigns in recent decades
involves major parties issuing "second preference" material.
This is not guidance to the major party's supporters, but appeals
to supporters of specific minor parties. Often, these appeals
try to pass themselves off as if they came from the minor party:
eg use of green paper and headings such as "Thinking of Voting
Green?" Attempts to use election day or post-election litigation
and petitions to respond to such material is costly and untimely.
How-to-vote material needs to be registered in advance and electoral
authorities empowered to ban material that is not clearly authorised
by its true party source. Even then, there have been cases of
major party activists misleading voters orally and through their
dress.
(iv) Waste versus expression. Many
Australians think the millions of how-to-vote flyers distributed
at polling stations are a huge waste; yet to appear impartial,
millions still politely accept flyers from every party. The freedom
of political expression guaranteed under European law may militate
against banning such flyers in the UK. The major parties in Australia
will not ban them as they can man each polling station and it
allows their activists to participate on polling day (UK party
activists may be stretched, given they will also need to remain
focused, on polling day, on last minute canvassing and encouraging
turnout). Regulatory options in Australia have included: limiting
the distribution of how-to-vote material close to polling stations
with a cordon sanitaire, and displaying official posters in each
polling station, showing each party's preference recommendations.
18 July 2010
| 
