Content uploaded by Graeme D. Orr
Author content
All content in this area was uploaded by Graeme D. Orr on Aug 29, 2017
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=cajp20
Download by: [Professor Graeme Orr] Date: 18 June 2016, At: 18:29
Australian Journal of Political Science
ISSN: 1036-1146 (Print) 1363-030X (Online) Journal homepage: http://www.tandfonline.com/loi/cajp20
Rethinking voter identification: its rationale and
impact
Graeme Orr & Tracey Arklay
To cite this article: Graeme Orr & Tracey Arklay (2016): Rethinking voter
identification: its rationale and impact, Australian Journal of Political Science, DOI:
10.1080/10361146.2016.1197181
To link to this article: http://dx.doi.org/10.1080/10361146.2016.1197181
Published online: 18 Jun 2016.
Submit your article to this journal
View related articles
View Crossmark data
ARTICLE
Rethinking voter identification: its rationale and impact
Graeme Orr
a
and Tracey Arklay
b
a
School of Law, The University of Queensland, Brisbane, Australia;
b
School of Government and International
Relations, Griffith University, Brisbane, Australia
ABSTRACT
Voter ID is a contentious issue in electoral democracies worldwide.
This article surveys arguments for and against voter ID in the
Australian context, presenting data from the first election in the
country to require it. The data demonstrate a differential impact
on regional electorates and on electorates with concentrations of
Indigenous voters. While the law in question (from the State of
Queensland) was moderate in its overall impact, confusion
created by it may have suppressed turnout. The law has since
been repealed, but voter ID now has the support of a
conservative majority on the Commonwealth Parliament’s
electoral matters committee. We conclude that voter ID is not a
solution to eliminating fraud, but an additional bureaucratic layer
upon the ritual of casting a ballot and a hurdle with unintended
consequences.
ARTICLE HISTORY
Accepted 14 May 2016
KEYWORDS
Elections; voter identification;
electoral law; Australian
elections; voter suppression
Australia is one of a relatively few liberal democracies that does not require electors to
produce identification (ID) to vote. Until recently it seemed likely to remain so. Australia
lacks any form of national ID card or papers, which might make voter ID relatively easy to
administer. It also has a long, progressive tradition that every vote is sacred, a tradition
traceable through nigh on a century of compulsory voting and compulsory enrolment
(Sawer 2003). However, in 2014 a conservative Queensland government enacted voter
ID in that State.
1
Although that initiative was quickly repealed by the Labor Party on
its return to office in 2015, the Coalition majority of the Commonwealth Parliament’s elec-
toral matters committee has recommended adopting the experiment for national elections
(JSCEM 2015: 112–20).
2
Voter ID is common across many developed democracies. True –and notably from the
Westminster and common law perspective shared by Australia –neither New Zealand nor
the UK (outside Northern Ireland) employ voter ID. But even in the UK, a recent report
has recommended its introduction (The Electoral Commission 2015: 26). In other
respects, democracies that do not require voter ID are a motley minority, like Georgia,
Samoa, Sudan and Tuvalu (Smith 2014: 48).
If the numbers of nations requiring voter ID is taken in isolation, it may appear as if
voter ID is inevitable, even natural. We argue, however, that international precedent is
© 2016 Australian Political Studies Association
CONTACT Graeme Orr g.orr@law.uq.edu.au
AUSTRALIAN JOURNAL OF POLITICAL SCIENCE, 2016
http://dx.doi.org/10.1080/10361146.2016.1197181
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
not itself a justification for the adoption of voter ID in Australia. On the contrary, voter ID
runs counter to Australia’s history of compelling registration and turnout. We also report
and analyse, for the first time, data from the Queensland experiment with voter ID. These
statistics shed light on voter ID’s disparate impact in the Australian context, particularly
on Indigenous and regional electors. We conclude by arguing that while voter ID might
provide a kind of reassurance to those concerned about electoral integrity, it would add
an unnecessary level of bureaucracy and potential confusion to the voting experience
and, if applied strictly, would be a form of electoral discrimination.
The politics and principles of electoral integrity and participation
Australia’s electoral system is often held up as a reasonable model, both in terms of its
rules and administration (Mercurio and Williams 2004: 3). As Farrell and McAllister
(2006) indicate, a majority of Australian voters are content with the transparency and fair-
ness of their electoral system. McAlister (2011:80–86) attributes Australians’relatively
high level of satisfaction with electoral democracy to compulsory voting and the under-
lying belief that their vote matters. There has been a general, especially expert, consensus
that any electoral fraud –through false enrolments or multiple voting –is isolated and not
at a level to affect outcomes (Hughes 1998; Hughes and Costar 2006).
Yet there have also long been those questioning the integrity of Australian electoral
practice and culture (e.g., McGrath 1996,2003 on behalf of the HS Chapman Society).
Scepticism received a fillip after the debacle of the 2013 Western Australian Senate elec-
tion, which had to be re-run after the loss of 1300 ballots by the Australian Electoral Com-
mission (AEC) (see Douglas 2014).
Tension between contentment and suspicion in electoral systems is a familiar theme,
repeated across the world. It is not confined to emerging democracies. Canada and the
UK, for instance, have recently proposed tightening electoral rules, ostensibly for integrity
purposes. The Canadian Electoral Bill C-23, known also as the Fair Elections Act, proposed
abandoning a system by which Canadian electors could have a friend or relative vouch for
their identity at the polls. This Act had a controversial genesis, with critics suggesting it
provided partisan advantage to the government through the disenfranchisement of
many, otherwise, eligible voters (Mas 2014; Shepherd, Stoney and Turnbull 2014: 18).
Similar arguments greeted the ‘Electoral Fraud in the UK’report, which suggested that
voter ID was necessary to remove ‘actual and perceived weakness in the system’, despite
finding that voter fraud was rare (The Electoral Commission 2015:1–2). Voter ID laws
remain a particularly ‘hot-button’issue in the US (Mycoff, Wagner and Wilson 2009).
A recent New York Times editorial (2014) suggested that as ‘there is virtually no in-
person voter fraud; the purpose of these laws is to supress voting’. Undeterred by such cri-
ticisms, conservative governments continue to press arguments that mandatory voter ID
enhances integrity through at least the appearance of a more secure system.
3
Disputes between those championing electoral integrity and those championing elec-
toral participation are almost emblematic of the divide between conservatives and social
democrats (Orr 2010: 68; see also Lee 2012;The Economist 2012). In ideal terms, of
course, integrity and participation should be a single goal for fair and inclusive elections.
After all a system with 100 per cent accurate rolls and 100 per cent turnout would by defi-
nition maximise each of these values. Debates, then, are conducted at the level of values
2G. ORR AND T. ARKLAY
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
and principles. It is an inescapable truth, however, that whatever rules are set, electoral
systems can be structured by incumbents so as to be ‘the most specific manipulative instru-
ment of politics’(Sartori 1968: 273, cited in Grofman and Liphart 1986).
4
Voter ID in international perspective
It is worth putting the potential adoption of voter ID in Australia in international context.
Despite differences between Australian and other, especially voluntary, voting systems,
which reduce the ability to compare cause and effect, the normative and even political
arguments for and against voter ID in Australia tend to follow a similar trajectory to
those overseas. Those supporting the measure claim a desire to eliminate voter fraud,
or at least the perception of voter fraud. Whereas those against argue that voter ID
raises barriers to participation and may disenfranchise certain groups of citizens.
While acknowledging that the political system in say the US is very different from Aus-
tralia, we pause to briefly examine the debate over voter ID in North America, before pro-
viding evidence as to why voter ID is both unnecessary and unhelpful in the Australian
context. This analysis adds to the international literature examining the current trend
by conservative political parties to engage in voter suppression policies (Bentele and
O’Brien 2013; Shepherd et al. 2014).
In May 2014 the Canadian Parliament passed the Fair Elections Act (mentioned above)
to amend the Canada Elections Act. Pierre Poilievre, the Conservative Party Minister of
State who sponsored the bill asserted it ‘made it easier for law-abiding Canadians to
vote and harder to break the law …[i.e.] more difficult to vote illegally or fraudulently’
(Poilievre 2014). In response, an expert group of 19 international academics claimed
‘Canada’s international reputation as one of the world’s guardians of democracy and
human rights is threatened’by the reforms (The Globe and Mail 2014). It is important
to note that –as in the UK, Australia and even the US –evidence suggests there is
minimal voter fraud in Canada (Neufeld 2013: 6; Shepherd et al. 2014: 19).
One provision of the Canadian reforms removed ‘vouching’provisions. Vouching is an
alternative to ID that allows one elector to vouch for another’s identity at a polling
booth. Removing vouching altogether, it was estimated, could have disenfranchised
over 120,000 electors, particularly seniors, students, the poor and Indigenous citizens.
On that issue, the government relented and a revised bill allowed for an elector carrying
physical ID to vouch, in writing, for the name and address of a voter who lacked ID at
the polls (Bryden 2014). Nonetheless, Burgmann (cited in Smith 2014: 72) has argued
that the move to mandate physical ID in Canada reduced Indigenous voting there by
10 per cent.
Over 30 States in the US mandate some form of voter ID.
5
There is a substantial body of
literature that shows voter fraud is rare in the US and that, when it does happen, it is tends
to be the result of ‘election officials taking steps to change election results or …involves
absentee ballots’(Hasen, quoted in Lee 2012). Voter ID will not prevent either of these
types of fraud. Ongoing litigation in US federal courts has led to findings of racial discrimi-
nation in Texas’s strict photo-ID requirements (Veasey v Abbott).
6
While such litigation,
and see-sawing Republican and Democratic majorities in state legislatures, means some
US States are now reviewing their harsher ID restrictions, the most relevant Supreme
Court rulings (Crawford v Marion County Election Board;Shelby County v Holder)
7
still
AUSTRALIAN JOURNAL OF POLITICAL SCIENCE 3
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
preserve considerable leeway for individual States to set their own rules (Mann and
Wakeman 2013).
US scholars claim that the harsher laws are mostly likely to impact people who are unli-
kely to vote conservatively (Shepherd et al. 2014: 19). In the Texas case, for instance,
besides obvious sources of photo ID such as driving licences or passports, gun licences
and military ID were accepted but, for example, student cards were not. As noted
earlier, conservative politicians and activists are more likely to frame the issue in terms
of electoral integrity and liberals or social democrats in terms of making participation
easier.
8
In 2001, the then Liberal Party Director gave evidence that members of his
party had ‘mixed views’on voter ID (Crosby, cited in Holmes 2014). More recently,
however, Smith (2014: 74) reported an observable partisan divide across Canada, the
US and Australia in regards to those who favour voter ID laws.
These philosophical instincts may also map onto perceptions of partisan gain. After
examining various variables, Bentele and O’Brien (2013: 1103) conclude that, in the US,
‘a straightforward picture emerges’whereby ‘the emergence and passage of restrictive
voter access legislation is unambiguously a highly partisan affair, influenced by the inten-
sity of electoral competition’and that ‘demobilization efforts are not a blunt practice’.
Marginalised groups in society are most likely to have difficulty obtaining voter ID, and
US studies confirm ID adds a layer of complexity to their participation in the political
process (Alvarez, Bailey and Katz 2008; Hershey 2009: 87 citing Vercellotti and Anderson
2006). As Robert Peel put it in the 19th century, ‘[t]hat party is strongest which has the
existing [voter] registration in its favour’;
9
a sentiment echoed in Harry Truman’s
alleged observation, a century later, that ‘decisions are made by those who show up’.
Yet explanations as to why conservative parties might favour voter ID as a form of voter
suppression in the US and Canada do not readily apply to the Australian case. Reasons for
this include the widespread use across many States in the US of differing and sometimes
draconian requirements for voter ID; compulsory as opposed to voluntary voting laws in
Australia; and demographic, cultural and socioeconomic differences between the
countries. The compulsory nature of voting in Australia also helps ensure governments
are more likely to consider the entire Australian population in their policy formulations
for fear of reprisal at the ballot box.
Voter ID law in Australia: another first for Queensland
Voter ID laws can, at one level, be seen as but a continuation of a pioneering tradition in
Queensland. After all, Queensland was the first Australian jurisdiction to enact compul-
sory voting, in 1915 (the Commonwealth followed, for national elections, in 1924). It
also employed ‘contingent voting’, a form of optional preferential voting, as early as
1892. Queensland has been something of an outlier in perpetuating rules and structures
that have been problematic for electoral democracy and accountability. Most notable
were its longstanding system of zonal malapportionment, which skewed rural and agricul-
tural interests over urban ones (Orr and Levy 2009; Wanna and Arklay 2010: 616), and its
lack of an elected upper house to check majoritarian rule (Aroney et al. 2008; see Murphy
1980 on its abolition). Each, in its own way, was a form of political manipulation designed
to enhance the ruling party and increase executive governmental power (Wear 2005: 87).
In electoral reforms, then, Queensland has been both forerunner and recalcitrant.
4G. ORR AND T. ARKLAY
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
In Australia, generally, voter fraud has been rare. Significantly, no election result has
been affected by fraudulent voting practices (Hughes 1998). Independent electoral com-
missions maintain the electoral roll, which is updated using government data to further
safeguard their accuracy. Proof of identity is required at the time of enrolment to
protect the integrity of the rolls. The Queensland government’s own discussion paper
(2013: 29), which floated the voter ID initiative, admitted that voter fraud was not a an
issue at previous Queensland elections and that voter ID might be considered a ‘dispropor-
tionate response to the risk’.
Besides concerns about disproportionate impact on vulnerable groups, opposition to
voter ID in Queensland centred around two propositions. One was the mismatch
between compelling people to turnout at the polls on pain of a fine, yet potentially
turning them away for forgetting or being unable to produce ID. The other was the fact
that postal voting was simultaneously made a right,
10
yet postal voters (who skew
towards incumbents and tend to be older) were not required to forward ID. Only those
voting in person on polling day or, increasingly, at early voting centres were required to
submit ID. If the purpose of voter ID is to erect a bureaucratic barrier to fraud, this is a
curious omission. Postal voting is notoriously more subject to fraud than in-person
voting.
11
These two objections to voter ID were united by the apparent contradiction of
simultaneously making the ‘when’and ‘where’of voting more convenient, yet making
the ‘how’of voting more of an administrative chore.
There are, admittedly, thousands of apparent cases of multiple voting at each elec-
tion in Australia. However, these are overwhelmingly the result of administrative
error (people with similar names being marked off the voting lists) or innocent elec-
toral error. According to the AEC at least 80 per cent of multiple voting cases not
attributed to administrative error involve elderly people or people with poor language
skills (AEC 2009). There are also institutional mechanisms in place to investigate
such problems (AEC 2010). In any event, multiple voting is not obviously deterred
by voter ID: multiple voting might better be tackled with on-line marking off of elec-
toral lists. Personation, rather than multiple voting, is the primary, practical target of
voter ID.
To its credit, and unlike many US States, the Queensland Government’s initial exper-
iment with voter ID did not mandate photo-only ID. In two respects, it was relatively gen-
erous. First, a variety of ID was ministerially regulated as acceptable (and in the case of
accounts/statements, which are often e-mailed to people today, electronic copies held
on a mobile device were accepted):
.a current driver licence;
.a current Australian passport;
.a voter information letter issued by the commission or similar document evidencing
electoral enrolment;
.an identification card issued by the Commonwealth or State evidencing the person’s
entitlement to a financial benefit (e.g., a Commonwealth seniors health card, health
care card, Medicare card, pensioner concession card);
.an adult proof of age card issued by the State;
.a recent account or notice issued by a government or a public utility provider including
a telecommunications company (e.g., an Australian Taxation Office notice of
AUSTRALIAN JOURNAL OF POLITICAL SCIENCE 5
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
assessment, council rates notice, electricity, water or gas account, or telephone or inter-
net bill, containing the elector’s name and address);
.a recent account statement, current account card or current credit card issued by a
financial institution.
Second, electors who did not produce acceptable ID were to be offered a ‘declaration
vote’. That is, to complete a declaration as to their identity and address, with their accom-
panying ballot paper set aside for special scrutiny, and possible inclusion in the count, at a
later stage.
While the types of acceptable ID were broad, there were still concerns about voter ID
adding another layer of complexity to polling day. Around 10,000 casual workers are
employed by the Electoral Commission Queensland (ECQ) during a general election. Train-
ing would have to be thorough and well resourced; yet even then inconsistencies might arise
in interpreting and applying the list of acceptable ID or in offering declaration votes. The
Indigenous and Torres Strait Islander Social Justice Commissioner argued that many
within his remit lack a birth certificate, which creates a ‘vicious cycle’in terms of acquiring
formal ID (Gooda 2014). On AEC data, 1.5 per cent of new enrolments are by people who do
not have a driver’s licence or birth certificate (Howie 2013). Other concerns included whether
voter ID would further bureaucratise the electoral experience and potentially lengthen
queues at polling stations. Most of all, fears were expressed that the initial voter ID rules
could be the thin end of a wedge, with ID requirements being tightened and the declaration
vote fall-back removed, by a future Attorney-General and Parliament, respectively.
The experience of Queensland’s 2015 election
The ECQ captured and published online data relating to ‘uncertain ID’voting at the 2015
general election, across Queensland’s 89 electoral districts. (‘Uncertain ID’voters were
those electors those who presented on polling day, but had to apply to vote by declaration
ballot because they did not present acceptable ID). We were able to analyse the data by
comparing them with ABS census statistics about various socioeconomic factors. These
were: income, median age, per cent non-English speaking background, per cent Indigen-
ous (i.e., Aboriginal and Torres Strait Islander) residents. Population density was also
employed as a direct proxy for a rural–urban spectrum. The data are presented below.
It should be noted that the manner of the ECQ’s collection of data posed two issues. First,
returning officers only collected voter ID data on a seat-by-seat level, not by polling stations.
Hence, the demographic variables also had to be measured at that level. This limited the
granularity of the analysis, since electoral districts vary less by, for example, average age
or income, than do electoral sub-districts. Nonetheless, the seat-by-seat level demographics
still exhibit significant differentiation, especially as regards Indigeneity and population
density. A smaller problem was that due to an administrative misunderstanding, uncertain
voter ID was not collected in one seat, Toowoomba South. However, that seat has a demo-
graphically almost identical neighbour, Toowoomba North, so extrapolations were made
from the number of uncertain ID voters in the latter seat.
Before the presentation and discussion of the results, by way of background Table 1
provides a summary of the data, showing the state-wide averages –and the variation
between electoral districts –for the key variables under study.
6G. ORR AND T. ARKLAY
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
Bivariate correlations were conducted, examining the association between the number
of uncertain ID voters and the demographic variables. Table 2 summarises these results.
The results demonstrate a significant relationship between ID-less voters and Indigen-
ous population (r= .650, p< .001), such that the percentage of uncertain ID voters
increases with an increase in a seat’s Indigenous population. The relationship between
population density and uncertain ID voters was also significant (r=−.353, p= .001),
with fewer ID-less voters in regions with higher population density. Relationships with
income or median age were non-significant. This may not disprove a relationship
between age, but merely be an artefact of the tendency for seats, as opposed to sub-dis-
tricts, to not vary significantly by average age.
In relation to non-English speakers, especially newer immigrants who might otherwise
be assumed to have greater problems with either producing ID or knowing of its require-
ment, any such effect may have been outweighed by their relative urbanity. The values (p
= .07 and .09) are close to significant, indicating some relationship between NESB status
and ID problems, but only a more fine grained analysis within electoral sub-districts could
clarify that.
A regression analysis was also employed to determine if the demographic character-
istics of a region significantly predicted the proportion of uncertain ID voters. Results
are presented in Table 3. Together, these variables account for 49.4 per cent of the variance
Table 1. Summary data.
Variable Mean Standard deviation Minimum Maximum
Uncertain ID votes as percentage of turnout 0.622 0.298 0.23 2.36
Total turnout as percentage of enrolment 89.849 2.153 83.59 92.67
Income pp pa ($) 31,328 5714 20,540 45,448
Median age 36.741 3.900 29 45
NESB (%) 8.110 9.236 1.16 57.26
Indigenous (%) 5.514 7.955 0.61 58.15
Density (voters/km
2
) 410.939 449.127 0.03 1816.65
Table 2. Pair-wise correlations, with significance values.
Variable Correlation with ‘uncertain ID’voters Significance value
Income −0.0948 0.376
Median age 0.0196 0.885
NESB % −0.1807 0.090
Indigenous % 0.6502 0.000
Density −0.3531 0.001
Table 3. Regression model.
Variable Coefficient (B) Std Error (B) Significance (p)
Income .000 .000 .379
Median age .007 .007 .338
NESB % −.005 .003 .057
Indigenous % .025 .003 .000
Density −.001 .001 .412
Constant .126 .414 .761
R
2
0.4948
Adjusted R
2
0.4644
AUSTRALIAN JOURNAL OF POLITICAL SCIENCE 7
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
(R2
adj = .464, F(5, 83) = 16.26, p< .001). While the percentage of Indigenous residents sig-
nificantly predicted uncertain ID votes (B= .025, p< .001), other variables were non-sig-
nificant. Results of the post-estimation analyses of the residuals and the variance inflation
factor were satisfactory.
Discussion of the Queensland data
The data presented above establish a not unexpected correlation between higher levels of
Indigeneity and electors experiencing problems with ID. It also demonstrates a correlation
between rural–urbanity and such problems, so that electors in more urbanised seats
experienced less problems with ID. The top 10 seats, by percentage of voters without
acceptable ID (between 2 and 0.9 per cent of total turnout), include a roll-call for the
larger, remote and more Indigenous electorates: Mt Isa, Cairns, Warrego, Cook, Keppel,
Townsville, Thuringowa and Barron River.
Explanations for these findings are several and revolve around access to information
and to ID. In relation to ID, there are the twin issues of owning ID, and one’spropensity
to either routinely carry it (or be able to shuttle from the polling station to home and
back again, possibly at considerable distance, to obtain it). There may also be cultural
factorsatwork.Forinstancepeopleinmoreremoteregionsmaylivefreerlives,
less concerned with bureaucratic niceties such as regulations about keeping or present-
ing ID.
There were also anecdotal reports of some polling booths providing different infor-
mation than others (Robertson 2015). If true, and without being critical of the ECQ,
this would be unsurprising. The 2015 poll was a snap election, employing an entirely
novel voter ID process. There was limited time and resources for public education
about voter ID, and a need to distribute thousands of polling stations across a geographi-
cally large jurisdiction. Perversely, the broader and hence more complex are the list of
acceptable ID and the rules permitting declaration votes, the more likely that casual
polling staff might interpret or explain those rules differently, compared to rules that
are narrow, crude but ‘bright-line’(such as a ‘no drivers licence, no vote’rule).
Of course, these findings are of correlations only and, as earlier noted, of correlations at
the level of electoral districts. And in some electorates, indigeneity and remoteness
overlap. Beside anecdotes, there are no data at the level of individual electors. Electoral
commissions retain copies of applications for declaration votes. But due to privacy laws,
these are not available to researchers. Even if they were, they would shed little light: elec-
tors do not enrol as ‘Indigenous’,‘NESB’or ‘poor’. Nonetheless, the correlations are both
strong and not unexpected. The parliament was presented with evidence indicating that
Indigenous people had less access to ID. For example, only 38 per cent of Indigenous
Queenslanders hold a driver’s licence (Human Rights Law Centre 2014: 5), and that
around 13 per cent of Indigenous births in Australia are not registered (Castan Centre
for Human Rights Law 2014: 2).
In absolute terms, the numbers of Queensland voters who lodged uncertain ID declara-
tion votes was not huge: 16,852 out of 2,435,432 who voted in person.
12
This represents
0.69 per cent of in person turnout (a further 244,442 voted by post and hence were not
subject to the ID law). Of those who cast declaration votes for ‘uncertain ID’, only 402
were not admitted to scrutiny. This either suggests that almost no-one sought to cast
8G. ORR AND T. ARKLAY
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
fraudulent votes, or that the ECQ simply checked the declared name and address against
their records.
There appear to be two overlapping, but distinctive, reasons for the fact that only about
0.7 per cent of those who cast ballots in person did not produce appropriate ID. The first is
that overall turnout dropped by almost 1.1 per cent. This drop was surprising, given the
2015 election was expected to be (and was) the first close election in Queensland since
1998. The election featured a high level of resentment and backlash against the govern-
ment, which lost office on a swing of 14 per cent. Pre-poll enrolments were in the
order of 60,000, up from just 23,000 at the previous election, an ‘unprecedented’spike
which suggests a high level of interest and motivation (Branco 2015). Close elections ordi-
narily motivate higher turnout; predictable ones the reverse.
13
In the nine previous elec-
tions in the 26 years of the modern (or post-Bjelke-Petersen) era, turnout in
Queensland only dropped by a non-negligible amount in elections that were predictable
landslides for an incumbent government (2001, 2004, 2006). In addition, the ability to
postal vote and options for early voting were both expanded for this election.
Turnout of course is subject to multiple factors, so there can be various reasons for a
decline in turnout –e.g., the 2015 Queensland election was called early, with a polling
day of 31 January. It is reasonable, however, to speculate that a considerable number of
Queensland electors in 2015 were deterred from turning out because of a misapprehension
that ‘voter ID’meant what it implied. Certainly, one known–unknown from the voter ID
experience was the number of electors who did not turn up at the polls either because they
lacked ID, or were out and about on election day, intending to vote late, but realised they
were not carrying ID.
Such confusion is not fatal to voter ID per se: one response would be greater time and
resources to advertise the voter ID requirements and the declaration vote fall-back.
However, such education will always be imperfect. Numerous journalists, including pol-
itical and news reporters and commentators, were unaware that ‘voter ID’really meant
that electors were entitled to lodge a declaration vote.
14
To give a flavour of this, one
online news story commenced: ‘No ID. No Vote. No Thank You’(Brisbane Times
2014). It is hard to imagine all citizens grasping the jargon of ‘declaration voting’.
A second, more optimistic, explanation for the relatively modest percentage of uncer-
tain ID declaration votes is that the typical modern elector, especially in Queensland’s
cities and provincial towns, probably does carry a wallet or purse containing, for
example, a drivers licence or Medicare card. This is a claim that supporters of voter ID
use when arguing to reassure mainstream audiences (this reassurance, of course, means
little to vulnerable sub-communities that may lack that ID).
Conclusion: voter ID a solution in search of a problem
The now-repealed Queensland experiment with voter ID was notable for three things.
First, at least compared to the US, the law was relatively moderate. Second, at the polls
it impacted on barely one per cent of electors, but confusion about what voter ID
‘meant’seems likely to have depressed turnout. Third, voter ID impacts more heavily
on remote and Indigenous communities than urban and non-Indigenous ones.
In our view, both the normative arguments and the Queensland data given above
demonstrate that there is no real case for voter ID in Australia. This is not to dismiss
AUSTRALIAN JOURNAL OF POLITICAL SCIENCE 9
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
the concerns of all of those who advocate voter ID. The onus, however, must be on those
wishing to impose hurdles to the ballot to justify those hurdles.
15
This is doubly so given
Australia compels both enrolment and turnout. It is not sufficient to simply point to the
prevalence of voter ID in other countries. From a pragmatic and moral perspective, com-
pulsion in Australia is in itself an integrity measure: a roll that is as broad as possible, and
turnout as high as possible, militates against electoral fraud.
Voter ID, in a moderate form, may well have symbolic and expressive value.
16
Social
democrats argue that regulation of political finance may be justified to increase trust in pol-
itical integrity (the appearance of corruption being a concern as much as its actuality).
17
Conservatives can similarly point to a concern that electoral laws appear to guard against
fraud, as a desirable end in itself. Indeed, the real dispute about voter ID in Australia has
so far been joined primarily at this level –it has been less about partisan benefit or detriment
as has been the case in the US. As long as voter ID laws are accompanied by a well admi-
nistered and easily accessible declaration ballot –and decent public education –then its
impact is less at the level of partisan effect and more at other levels. (At the level of electoral
effect, voter ID’s impact on Labor-leaning Indigenous communities might be balanced by
its impact on conservative rural communities. Similarly, lack of ID may impact on left-
leaning youth as well as conservative-leaning senior-citizens).
The other levels of effect, even of a moderate voter ID law, must be accommodated in
any argument about symbolic and expressive effects. First, voter ID does impinge on the
actuality and experience of electoral democracy. An elector who attends a polling station
and is denied an ordinary ballot, has no way of knowing if their ballot is admitted to scru-
tiny. Declaration votes are separated and put, in effect, into a ‘black-box’. The declaration
voter leaves the polling station not knowing if their ballot will actually be scrutinised. At a
minimum, declaration voters should be contacted by the Electoral Commission after
polling day and told if their ballot was accepted into scrutiny, and if not, why not. The
ritual of the polling station, too, is transformed when electoral officials have to solicit
voter ID. Some may feel the act of casting a vote is made more solemn; others may experi-
ence it as a more bureaucratic and less trusting encounter (Orr 2015:25–26).
Ultimately, whether there is a community perception that voter ID is a common sense
means to address any appearance that elections are subject to fraud, the reality is that voter
ID can do little to prevent those cases of deliberate multiple voting, which currently exer-
cise concern (cf. JSCEM 2015). As we noted at the outset of this article, scholars such as
Colin Hughes (himself a former head of the AEC) and Brian Costar have argued that such
concerns are inflated. Nothing about voter ID per se stops someone presenting themselves
at different polling stations to vote more than once in their own name: at best it may deter,
at the margins, anyone impersonating another elector multiple times or require them to
forge or copy ID.
Hence, whatever the intentions or motivations for its introduction, as Rayner (2014)
observed, if governments were serious about eliminating fraud then a universal national
ID might need to be introduced. (Voter ID is less of a practical and cultural burden,
and more fairly and equitably enforceable, where such ID papers are the norm).
Further, electronic, interactive versions of the roll would need to be provided at as
many polling booths as is feasible (allowing for reliable and secure internet connections)
if multiple voting is to be eliminated. Each of those reforms would involve significant costs,
not to mention privacy and security concerns.
18
In the meantime, further bureaucratising
10 G. ORR AND T. ARKLAY
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
the voting experience and causing disparate impact on indigenous and remote electors,
would be, to quote Green (2014), to legislate an ‘answer trying to find a problem’.
Notes
1. Electoral Reform Amendment Act 2014.
2. The minority MPs (Labor and The Greens) on the committee rejected voter ID (JSCEM 2015:
162–3). The same committee had also rejected voter ID as unwarranted some 14 years earlier
(JSCEM 2001: 45).
3. In a government press release mooting proposed amendments, then Queensland Attorney
General Jarrod Bleijie (2013) stated that ‘The Newman Government wants to ensure the
state has an electoral system that meets high standards of integrity and accountability’,
and that ‘Fair and effective electoral laws are central to the promotion of participation in
our democracy.’This echoes comments made by the Canadian Minister, Pierre Poilievre:
‘We will have a requirement for physical ID every time someone votes …It is fair, it is
reasonable and, as of today, it will be the law’(Maher 2014). As in Queensland, Canadians
were permitted to bring electronic ID (e.g., documents held on a smartphone).
4. Grofman and Lijphart (1986:2–3) cite 13 elements of political laws that when manipulated
may change political outcomes. These include registration requirements, ease of voter and
candidate access to the balloting process, the structure of political competition (including
ballot format, districting procedures, financing rules, campaigning and, of course, vote hand-
ling and counting procedures. Voter ID relates to ease of voter access to the process.
5. Election law in the US, even for national elections, is primarily governed by State law.
6. Veasey v Abbott (US Court of Appeals, Fifth Circuit, 5 August 2015).
7. Crawford v Marion County Election Board Shelby County v Holder 553 US 181 (2008)
(upholding Indianan photo ID-only law); Shelby v Holder 570 US _ (2013) (watering
down the Voting Rights Act of 1965).
8. Curiously, the position is reversed when it comes to money in politics.
9. Peel in a letter to Arbuthnot, 1838, cited in Seymour (1915: 125).
10. Previously, an elector needed a good reason related to distance or mobility.
11. In a recent UK case illustrating this, Simmons v Khan [2008] EWHC B4, the elections judge
concluded that all-postal local elections had rendered ‘wholesale electoral fraud both easy and
profitable’.
12. The raw figure for ID-less ballots was 16,591. We have incorporated a proportionate figure to
account for the missing Toowoomba electorate (see discussion above).
13. In contrast, turnout rose appreciably in each election that was expected to be close (1988 and
2009) and was static in elections which led to a change in the status of the government (1989,
1995, 2012).
14. The authors, in giving media commentary before the election, including on voter ID, encoun-
tered both public and commercial journalistswho assumed that ‘voter ID’meant ‘no ID, no vote’.
15. As Justice Isaacs put it in Kean v Kerby (1920) 27 CLR 449 at 459: ‘[T]he ballot, being a means
of protecting the franchise, should not be made an instrument to defeat it …’
16. On law as a form of governmental or social expression, see Anderson and Pildes (2000).
17. A position accepted even under libertarian first amendment (free-speech) doctrine in the US:
see Buckley v Valeo 424 US 1 (1976).
18. In 1986 and again in 1987 the Australia Card Bill was defeated after lengthy parliamentary
debates and vociferous opposition from sectors of the population (Clarke 1987).
Acknowledgements
The authors wish to acknowledge Gregory Dale and Kathryn Bernier for their research assistance,
and the ECQ for fielding our queries.
AUSTRALIAN JOURNAL OF POLITICAL SCIENCE 11
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes on contributors
Graeme Orr is a professor in the Law School at the University of Queensland.
Tracey Arklay is senior lecturer in the School of Government and International Relations at Griffith
University.
References
AEC [Australian Electoral Commission]. 2009. Submission 169.19 to the Joint Standing Committee
on Electoral Matters: Report into non-voting and multiple voting at the 2007 federal
election. URL: <http://www.aphref.aph.gov.au-house-committee-em-elect07-subs-sub169_19.
pdf>. Consulted 21 October 2015.
AEC [Australian Electoral Commission]. 2010.Electoral backgrounder: Electoral fraud and multiple
voting.URL:<http://www.aec.gov.au/about_aec/Publications/Backgrounders/files/2010-eb-fraud-
and-multiple-voting.pdf>. Consulted 26 May 2014.
Alvarez, R., Bailey, D. and Katz, J. 2008. The effect of voter identification laws on turnout. Social
Sciences Working Paper 1267R. Pasadena: California Institute of Technology. URL: <http://
jkatz.caltech.edu/research/files/wp1267R.pdf>. Consulted 21 October 2015.
Anderson, E. and Pildes, R. 2000. Expressive theories of law: A general restatement. University of
Pennsylvania Law Review 148: 1503–75.
Aroney, N., Prasser, S. and Nethercote, J.E. 2008.Restraining elective dictatorship: The upper house
solution. Perth: UWA Press.
Bentele, K.G. and O’Brien, E.E. 2013. Jim Crow 2.0? Why states considered adopting restrictive
voter access policies. Perspectives on Politics 11: 1088–116.
Bleijie, J. 2013. Public to have a say on electoral reform. URL: <http://jarrodbleijie.com.au/news/
public-have-say-electoral-reform>. Consulted 21 September 2014.
Branco, J. 2015. Queensland election 2015: ‘Unprecedented’voter enrolment. Brisbane Times
(online) 13 January. URL: <http://www.brisbanetimes.com.au/queensland/queensland-state-
election-2015/queensland-election-2015-unprecedented-voter-enrolment-20150113-12n2id.
html>. Consulted 16 October 2015.
Brisbane Times.2014. ID checks at poll booths ‘unnecessary’. 23 January. URL: <http://www.
brisbanetimes.com.au/queensland/id-checks-at-poll-booths-unnecessary-20140122-3199p.html>.
Consulted 16 October 2015.
Bryden, J. 2014. Controversial Fair Elections Act pushed through parliament by Conservative gov-
ernment. National Post, 13 May. URL: <http://news.nationalpost.com/news/canada/canadian-
politics/controversial-fair-elections-act-pushed-through-parliament-by-conservative-government>.
Consulted 16 October 2015.
Castan Centre for Human Rights Law. 2014. Submission to the Queensland Legal Affairs
and Community Safety Committee on the Electoral Reform Amendment Bill 2013. URL:
<https://www.parliament.qld.gov.au/documents/committees/LACSC/2013/ElectoralReform2013/
submissions/010.pdf>. Consulted 14 February 2016.
Clarke, R. 1987. Just another piece of plastic for your wallet: The ‘Australia Card’scheme.
Prometheus 5(1): 29–45.
Douglas, M. 2014. Ballot bungles: Lessons from the Australian senate. Election Law Journal 13:
559–69.
Farrell, D. and McAllister, I. 2006. Voter satisfaction and electoral systems –does preferential
voting in candidate-centred systems make a difference? European Journal of Political Research
45: 723–49.
12 G. ORR AND T. ARKLAY
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
Gooda, M. 2014. Aboriginal and Torres Strait Islander Social Justice Commissioner, Submission
134 on Electoral Reform Amendment Bill 2013 to the Legal Affairs and Community Safety
Committee, Queensland Parliament. 17 January. URL: <http://www.parliament.qld.gov.au/
documents/committees/LACSC/2013/ElectoralReform2013/submissions/134.pdf>. Consulted 14
June 2014.
Green, A. 2014. Voter ID laws pass their first hurdle in Stafford. Antony Green’s Election Blog. URL:
<http://blogs.abc.net.au/antonygreen/2014/07/voter-id-laws-pass-their-first-hurdle-in-stafford.
html>. Consulted 18 September 2015.
Grofman, B. and Lijphart, A. 1986. Introduction. In B. Grofman and A. Lijphart, eds. Electoral laws
and their political consequences. New York: Agathon Press.
Hershey, M.J. 2009. What we know about voter-ID laws, registration, and turnout. PS 42: 87–91.
Holmes, B. 2014. Voter ID. Australian Parliamentary Library Research Paper, 4 August. URL:
<http://www.aph.gov.au/About_Parliament/Parliamentary_Departments/Parliamentary_Library/
pubs/rp/rp1415/VoterID>. Consulted 18 September 2015.
Howie, E. 2013. Voter ID laws politically motivated. Sydney Morning Herald (online) 21 October.
URL: <http://www.smh.com.au/comment/voter-id-laws-politically-motivated-20131021-2vw6v.
html#ixzz3BTPCrQ6M>. Consulted 16 October 2015.
Hughes, C.A. 1998. The illusive phenomenon of fraudulent voting practices: A review article.
Australian Journal of Politics and History 44: 471–91.
Hughes, C.A. and Costar, B. 2006.Limiting democracy: The erosion of electoral rights in Australia.
Sydney: UNSW Press.
Human Rights Law Centre. 2014. Submission to Queensland Parliament Legal Affairs and
Community Safety Committee on Electoral Amendment Bill 2013. URL: <https://www.
parliament.qld.gov.au/documents/committees/LACSC/2013/ElectoralReform2013/submissions/
133.pdf>. Consulted 14 February 2016.
JSCEM [Joint Standing Committee on Electoral Matters]. 2001.User friendly, not abuser friendly:
Report of the inquiry into the integrity of the roll. Canberra: Parliament of the Commonwealth of
Australia.
JSCEM [Joint Standing Committee on Electoral Matters]. 2015.The 2013 election: Report on the
conduct of the 2013 election and matters related thereto. Canberra: Parliament of the
Commonwealth of Australia.
Lee, S. 2012. Everything you’ve ever wanted to know about Voter ID laws. ProPublica. 5 November.
URL: <http://www.propublica.org/article/everything-youve-ever-wanted-to-know-about-voter-
id-laws>. Consulted 18 September 2015.
Maher, S. 2014. Fair Elections Act will be challenged in Court by Council of Canadians, Federation
of Students. National Post 9 June. URL: <http://news.nationalpost.com/2014/06/19/fair-
elections-act-will-be-challenged-in-court-by-council-of-canadians-federation-of-students/>.
Consulted 18 September 2015.
Mann, T.E. and Wakeman, R.L. 2013. Voting rights after Shelby County v Holder. Brookings
Institute, 25 June 2013. URL: <http://www.brookings.edu/blogs/up-front/posts/2013/06/25-
supreme-court-voting-rights-act-mann-wakeman>. Consulted 18 September 2015.
Mas, S. 2014. Election reform bill an affront to democracy, Mark Mayrand says. CBC News 8
February. URL: <http://www.cbc.ca/news/politics/election-reform-bill-an-affront-to-democracy-
marc-mayrand-says-1.2527635>. Consulted 16 October 2015.
McAlister, I. 2011.The Australian voter: 50 years of change. Sydney: UNSW Press.
McGrath, A. 1996.The frauding of votes. Sydney: Towerhouse Publications.
McGrath, A. 2003.The frauding of elections. Sydney: Towerhouse Publications.
Mercurio, B. and Williams, G. 2004. Australian electoral law: Free and fair? Federal Law Review 32:
365–95.
Murphy, D.J. 1980. The abolition of the Legislative Council. In D.J. Murphy, R.B. Joyce and C.A.
Hughes, eds. Labor in power: The Labor Party and governments in Queensland 1915–57.
Brisbane: University of Queensland Press.
Mycoff, J.D., Wagner, M.W. and Wilson, D.C. 2009. The empirical effects of voter-ID laws: Present
or absent? PS 4: 121–26.
AUSTRALIAN JOURNAL OF POLITICAL SCIENCE 13
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
Neufeld, H. 2013.Compliance review: Final report and recommendations: A review of compliance
with electoral day registration and voting process rules. Ottawa: Elections Canada. URL:
<http://www.elections.ca/res/cons/comp/crfr/pdf/crfr_e.pdf>. Consulted 18 September 2015.
Orr, G. 2010.The law of politics: Elections, parties and money in Australia. Sydney: The Federation
Press.
Orr, G. 2015.Ritual and rhythm in electoral systems: A comparative legal account. Farnham:
Ashgate.
Orr, G. and Levy, R. 2009. Electoral malapportionment: Partisanship, rhetoric and reform in the
shadow of the agrarian strong-man. Griffith Law Review 18: 648–65.
Poilievre, P. 2014. Pierre Poilievre on Fair Elections Act. openparliament.ca, 13 May. URL: <https://
openparliament.ca/debates/2014/5/13/pierre-poilievre-1/only/>. Consulted 16 October 2015.
Queensland Government. 2013.Electoral reform: Discussion paper. January 2013. URL: <http://
www.justice.qld.gov.au/__data/assets/pdf_file/0007/171529/disc-ppr-electroal-reform.pdf>.
Consulted 16 October 2015.
Rayner, J. 2014.Afix for what’s not broken: Why Australia doesn’t need voter ID. Inside Story 4
March. URL: <http://insidestory.org.au/a-fix-for-whats-not-broken-why-australia-doesnt-
need-voter-id>. Consulted 16 October 2015.
Robertson, J. 2015. Queensland ID law meant voters were wrongly turned away, inquiry told. The
Guardian (online Australian edition), 16 April. URL: <http://www.theguardian.com/australia-
news/2015/apr/16/queensland-voter-id-laws-saw-voters-wrongly-turned-away-inquiry-told>.
Consulted 21 October 2015.
Sawer, M. 2003. Enrolling the people: Electoral innovation in the new Australian Commonwealth.
In G. Orr, B. Mercurio and G. Williams, eds. Realising democracy: Electoral law in Australia.
Sydney: The Federation Press.
Seymour, C. [1915] 1970.Electoral reform in England and Wales: The development and operation of
the parliamentary franchise, 1832–1885. Reprint. London: David and Charles Reprints.
Shepherd, R., Stoney, C. and Turnbull, L. 2014. Politicized governance: Hyperbole or measurable
framework? IRSPM Conference.9–11 April, Carleton University, Ottawa, Canada.
Smith, R. 2014.Multiple voting and voter identification: Research report prepared for the NSW
Electoral Commission. February 2014. URL: <http://www.elections.nsw.gov.au/__data/assets/
pdf_file/0006/173373/Multiple_Voting_and_Voter_Identification.pdf>. Consulted 16 October
2015.
The Economist.2012. Counting voters, counting votes: Discouraging citizens voting not good
democracy. 27 October. URL: <http://www.economist.com/news/united-states/21565259-
discouraging-citizens-voting-not-good-democracy-counting-voters-counting-votes>. Consulted
18 September 2015.
The Electoral Commission (UK). 2015.Electoral fraud in the UK: Final report and recommendations.
January 2014. London: The Electoral Commission. URL: <http://www.electoralcommission.
org.uk/__data/assets/pdf_file/0008/164609/Electoral-fraud-review-final-report.pdf>. Consulted 18
September 2015.
The Globe and Mail.2014. We believe that this Act would prove [to] be deeply damaging for elec-
toral integrity within Canada, 19 March. URL: <http://www.theglobeandmail.com/news/politics/
we-believe-that-this-act-would-prove-to-be-deeply-damaging-for-electoral-integrity-within-
canada/article17561359/>. Consulted 18 September 2015.
Wanna, J. and Arklay, T. 2010.The ayes have it: The history of the Queensland parliament, 1957–89.
Canberra: ANU Press.
Wear, R. 2005. Adventure, heterodoxy and knavery: Queensland’s electoral experience. Queensland
Review 12(2): 87–96.
14 G. ORR AND T. ARKLAY
Downloaded by [Professor Graeme Orr] at 18:29 18 June 2016
