Limiting Democracy: The Erosion of Electoral Rights in Australia

BRIEFINGS

BRIEFINGS LIMITING DEMOCRACY

“It is not unusual for those whose party loses elections to say that there are flaws in the electoral system. What has been exceptional in Australia since 1996 is that supporters of the party that has won a series of elections complain they were robbed.” Australia’s electoral system is distinctive, independent and widely respected. For years, parties across the political spectrum have supported its main features – compulsory voting, an independent electoral commission, uniform rules across the nation. Yet, in 2006, all that began to change.

In this book two distinguished political scientists, Colin A. Hughes – Australia’s first electoral commissioner – and Brian Costar, reveal the dangers of the government’s legislation – and of other changes flagged by government MPs. They trace the history of Australia’s admired electoral system, revealing the carefully crafted rules that have ensured a fair and transparent system – a system now endangered by a set of poorly considered changes.

Cover: At Hill End, north of Bathurst, Greg McNamara wheels Sam George and ballot boxes to the polling booth for the 1998 federal election.

UNSW PRESS ISBN 0-86840-948-0

BRIEFINGS A series of short, topical books exploring social, political and cultural issues in contemporary Australia, published in association with Australian Policy Online, www.apo.org.au

9 780868 409481

COLIN A. HUGHES & BRIAN COSTAR

Because Australians have no explicit constitutional right to vote, parliament has wide discretion over how its members are elected. Having won control of both houses of parliament, the Howard government introduced legislation that erodes the right to vote and relaxes controls over political donations, all in the name of “electoral integrity.” Forced through parliament in mid-2006, these measures are dangerous for democracy.

UNSW PRESS

Limiting Democracy

The Erosion of Electoral Rights in Australia Colin A. Hughes & Brian Costar

LIMITING DEMOCRACY THE EROSION OF ELECTORAL RIGHTS IN AUSTRALIA

Colin A. Hughes is emeritus professor of political science at the University of Queensland and was the first Australian Electoral Commissioner (1984–89). Brian Costar is professor of Victorian state parliamentary democracy at Swinburne University of Technology.

BRIEFINGS A series of topical books exploring social, political and cultural issues in contemporary Australia Series editors: Peter Browne and Julian Thomas Australian Policy Online (www.apo.org.au) Institute for Social Research, Swinburne University of Technology

Limiting Democracy The Erosion of Electoral Rights in Australia COLIN A. HUGHES & BRIAN COSTAR

A UNSW Press book Published by University of New South Wales Press Ltd University of New South Wales Sydney NSW 2052 AUSTRALIA www.unswpress.com.au © Colin A. Hughes & Brian Costar 2006 This book is copyright. Apart from any fair dealing for the purpose of private study, research, criticism or review, as permitted under the Copyright Act, no part may be reproduced by any process without written permission. Enquiries should be addressed to the publisher. National Library of Australia Cataloguing-in-Publication entry Hughes, Colin A. (Colin Anfield), 1930- . Limiting democracy: the erosion of electoral rights in Australia. ISBN 0 86840 948 0. 1. Democracy – Australia. 2. Political rights – Australia. 3. Voting registers – Australia. I. Costar, Brian J. II. Title. 324.60994 Cover photograph: At Hill End, north of Bathurst, Greg McNamara wheels Sam George and ballot boxes to the polling booth for the 1998 federal election. Robert Pearce/FairfaxPhotos Printed by Hyde Park Press

Contents Introduction

7

1. Modernising electoral democracy: the 1983 reforms

15

2. Electoral integrity: the rise of the activists

33

3. Electoral integrity: slamming the door

47

4. Real problems, and how to fix them

60

Conclusion

90

Endnotes

92

Introduction

D

uring the three and a half centuries since the end of the English civil war democracy has spread well beyond its earliest outbreaks in western Europe. Free and fair elections are a necessary condition of democracy, but the long evolution of democratic government has meant that arrangements, practices and institutions have varied in different countries. It is more than 180 years since Australians first voted – for the partially elected NSW Legislative Council in 1824 – and 150 years since locally elected politicians gained the power to make electoral laws. Australia’s electoral system reflects a long history of political experimentation, with steadily increasing democracy as the principal theme. But although democracy can grow of itself, it also needs nurturing, and it can be eroded. In this book we show how many of the best features of the Australian electoral system – features that were intact as recently as mid 2006 – have been subverted by recent legislation, and that further erosions are likely. Although this book is primarily about Australia, we refer to comparable problems and solutions in the United States to place what is happening here in perspective. The influence of the United States can be seen very clearly in the recent changes to electoral laws introduced by the Australian government and in the rhetoric of the key advocates of change within the Liberal Party. As Australian political parties and interest groups have increasingly adopted American election campaign methods – negative advertising, excessive expenditure, direct mail, electronic databases and 7

LIMITING DEMOCRACY

push polling – so too have some of their members developed an enthusiasm for elements of US electoral law. We argue that this is a most undesirable development which, in the interest of Australian representative democracy, needs to be resisted. One conservative foreign policy expert, Owen Harries, has declared that “the Bush doctrine – the belief that American power should be used to spread democracy throughout the world – is dead in the bloody water.”1 One reason for this failure is that parts of the US brand of democracy are not up to export standard. Popular mythology holds that the United States is both the finest flower of the democratic experiment and its longest continuous example. Both propositions are contestable.2 The United States was established at a time (1789) when the idea that all the people should have the right to choose their rulers by way of election was still a novelty. Democratic forms were added later (sometimes much later) and in a piecemeal fashion. The Australian nation’s foundation in 1901 was less dramatic than America’s but reflected the advances in democratic theory and practice that had occurred in parts of western Europe and in the Australasian colonies in the second half of the nineteenth century. Australia and New Zealand share the honour of being the oldest continuous, modern democracies. It is regrettable that some members of one major component of Australian democracy, the Liberal Party of Australia, now seem to regard their party as the south-west Pacific version of the Republican Party – or at least its Texan branch – and want to import some of the least democratic elements of the United States’ electoral procedures. At the same time, ironically, quite a few American electoral reformers look to the Australian electoral system as a model worth emulating. The suggestion that the United States has a major democratic deficit may seem like a serious and provocative charge. But the recent bipartisan Carter–Baker Commission on Federal Election Reform, chaired by a former Democrat president and a former Republican secretary of state, argued that “Americans are losing 8

INTRODUCTION

confidence in the fairness of elections.” The twenty-person commission was “united in the view that electoral reform is essential…”3 While we don’t endorse US journalist Gregory Kane’s approving remark that “the United States is a republic, not a democracy,”4 we do argue that America’s current electoral machinery manifests major, overlapping shortcomings which can be categorised under the following heads: fragmentation; voluntarism; partisanship; and excessive judicialism. Its fragmentation is demonstrated by the fact that the United States does not have a single, consolidated, federal electoral law. Section 4 of the US constitution allocates to the 50 states the power to frame electoral procedures for the federal executive and legislature elections – and the states often hand the job over to local counties. The consequence is a bewildering mish-mash of laws and procedures that are open to partisan manipulation and excessive judicial review. In the wake of the controversies surrounding the 2000 presidential election, the US congress passed the Help America Vote Act, the declared objective of which was to facilitate the voting process. Unfortunately the Act also introduced voter identification requirements that have the potential, according to the political scientist Michael L. Goldstein, “to discriminate against poorer and less-educated voters.”5 Australia, by contrast, enacted the federal Franchise Act 1902 and the Commonwealth Electoral Act 1902 to enable uniform national regulation of all elections for federal parliament. Unlike in Australia, both voting and enrolment are entirely voluntary in the United States. This produces chronically low voter turnouts and inaccurate electoral rolls, and – without a reliable database of enrolled voters – makes it impossible to draw accurate electoral boundaries. As we will see, the absence of compulsion leads to widespread disenfranchisement. Moreover, millions of otherwise eligible US citizens are denied the vote because of prior felony convictions. It is not unusual for people to be jailed for voting while disqualified, and the laws are so confusing and inconsistent that many don’t know they aren’t entitled 9

LIMITING DEMOCRACY

to vote. Party activists can challenge the eligibility of would-be voters queued at polling stations by accusing them of felony convictions or launching investigations to scare voters away from the polls on election day.6 Voluntary voting operates more satisfactorily in other countries, which suggests that the development of the electoral administration in the United States is a large part of the problem. Partisanship is embedded in the American electoral system, taking a number of forms that would greatly alarm Australian voters. The task of drawing US House of Representatives electoral boundaries – what Americans call redistricting – rests almost exclusively in the hands of state legislators and governors who, of course, represent political parties and have no qualms about gerrymandering against their opponents.7 While a handful of states have bipartisan redistricting panels, none have non-partisan ones. In most states the top electoral officials are elected representatives of political parties; they are permitted to design ballot papers to benefit their party’s candidates and to make decisions during and after the vote that can and do affect results. In Florida (2000) and Ohio (2004), the two states crucial to the outcome of the presidential election in those years, the top electoral official also served on the Bush–Cheney election committee. The unprecedented difference between the official results of the 2004 presidential election and the usually accurate exit polls has opened the way for allegations that the election may have been “stolen.”8 In Australia all of these functions are discharged by the statutory Australian Electoral Commission, or AEC, whose charter mandates strict impartiality on the part of its officials. The final undesirable feature of American electoral law is the excessive judicialisation of almost the entire process. In Australia the courts play a very minor role in the voting system, but in the United States the approach of a presidential election sees the parties dispatch swarms of lawyers into the key “swing” states, ready to launch litigation to secure every possible advantage for their candidates and to subvert their opponents. Thousands of 10

INTRODUCTION

votes cast in US elections may not be entered into the count because of legal challenges by partisan interests. Most notably, the 2000 US presidential election was decided not at the polling places but in the federal Supreme Court on appeal from a state court.

• In this book we generally take as our starting point the major amendments to the Commonwealth Electoral Act enacted under the Labor government in 1983. While recognising some shortcomings in the 1983 legislation, we judge it to have been generally bipartisan and democratic. We compare that legislation with the Electoral and Referendum Amendment (Electoral Integrity and Other Measures) Act 2006, which we contend was motivated by partisanship and signals a retreat from Australia’s characteristic democracy. The 1983 amendments to the electoral system were the most comprehensive since the electoral Act was consolidated in 1918, and were introduced following a report by the Joint Select Committee on Electoral Reform, or JSCER. This was a committee of great expertise in electoral matters, with members from all parties represented in parliament. Its members worked solidly through the winter and took six months to write their report. At the time it was frequently remarked how unusual it was that, when it had the numbers in both houses to do whatever it wanted with the electoral system – provided the Democrats agreed – the Hawke government was willing to put the matter to a parliamentary committee to achieve the maximum amount of consensus. The outcome was a comprehensive reform of the system. The new, statutory Australian Electoral Commission, or AEC, had more autonomy and responsiblility than the Australian Electoral Office, which it replaced. The Senate grew from 64 members to 76 and the House of Representatives from 125 to 148. The Act ended malapportionment – in other words, electorates with unequal enrolments – and improved the process for drawing electoral boundaries, removing the power of either house of 11

LIMITING DEMOCRACY

parliament to veto redistribution reports. Registration of political parties was introduced, together with public funding for election campaigns and revised disclosure requirements. Voting for the Senate was simplified with the creation of the “above the line” option, intended to reduce the number of informal votes. The qualification to be on the electoral roll was altered from “British subject” to “Australian citizen.” Candidates now appeared in random order, rather than alphabetically, on the House ballot paper. And the redrafting of the Commonwealth Electoral Act also made many administrative alterations to the conduct of elections. Given the historical divisiveness of debates about electoral procedures, the 1983 reforms initially enjoyed a surprisingly high level of cross-party support. This support dissipated, however, as the Coalition, which had been in power continuously from 1949 to 1972 and from 1975 to 1983, lost five consecutive elections (and in one of them, 1990, it polled more votes than Labor).9 Even the Coalition victories of 1996, 1998, 2001 and 2004 failed to quell its alarm as the Senate refused to legislate many of the contentious recommendations of the Joint Select Committee of Electoral Matters, or JSCEM, now dominated by members of the Coalition government. That obstacle was removed when the Coalition acquired an unexpected majority in the Senate on 1 July 2005. The electoral integrity legislation, with which we will be concerned, became law less than a year later. The Act introduces a series of significant changes to electoral practice in Australia. For people who have not voted before, the electoral roll will now close at 8 pm on the day the writs are issued, removing the longstanding seven-day period of grace. Enrollees now need extra proof of identity when making their enrolment. All convicted prisoners have been deprived of the right to vote. Trade unions affiliated with the Labor Party will be subject to a more stringent regime of donation disclosure. And the threshold above which the identity of campaign donors must be revealed has been raised from $1500 to $10,000. 12

INTRODUCTION

In what follows we examine the claims made for and against the 1983 legislation and find that many of the criticisms are partisan rather than evidence-based. We contend that the Coalition’s 2006 legislation aims to solve problems that do not exist, and that its effect is anti-democratic. Where we do identify real problems in electoral law – in the areas of public funding of political parties and disclosure of political donations – we find both an unwillingness to remedy them and too often an enthusiasm for exacerbating them. Beyond the scope of this book is the situation in the six states and two territories, where local electoral systems have generally come to correspond to the federal version, but never entirely so and sometimes with quite significant variations. But the fact that the states and territories run separate but related systems raises an important question. If the federal system is altered in substantial particulars, will the states and territories follow suit? At present, with all eight of those jurisdictions under Labor governments, this seems unlikely. But will there be adverse consequences, for electors or others, when significant differences result? There is no electoral philosopher’s stone equivalent to the corporations power in the constitution that would enable the commonwealth to take over the states’ electoral jurisdictions, though it probably could usurp those of the two territories. Moreover, we might expect the Coalition parties at the state level to be very unhappy if changes in state law were enforced – by commonwealth financial sanctions, for example – thus incidentally depriving them of electoral powers in the longer run. Given that its 2006 amendments create a new gap between commonwealth and state electoral laws, it is ironic that the federal Coalition once justified its refusal to reduce the voting age to eighteen on the grounds that it would force the commonwealth and state electoral rolls out of step.10 Finally, it has to be said that the debate we report, like many others in Australian politics, has been debased by both macromendacity and micro-mendacity. That the incidence of macromendacity – the misrepresentation of major issues to the electorate 13

LIMITING DEMOCRACY

over extended periods – has increased so rapidly over recent decades in part reflects the micro-mendacity – the petty fudging, half-truths, untruths, exaggerations, uncorrected factual errors, misleading statements and selective omissions – in which political practitioners are also seen to engage.11 Such behaviour is, of course, not unique to democracies or these times. It was written of the Roman Emperor Constantius in the fourth century AD that he “sought to give an appearance of authenticity to circumstances which were invented or at best uncertain” and that he “was adept at making a mountain of mischief out of a molehill of evidence.”12 It is not unusual for those whose party loses elections to say that there are flaws in the electoral system. What has been exceptional in Australia since 1996 is that supporters of the party that has won a series of elections complain they were robbed.

A glossary of electoral terms to accompany this book is available at . 14

CHAPTER 1

Modernising electoral democracy: the 1983 reforms

I

n the technical political science literature “electoral system” has a narrow meaning.13 There, it comprises three elements: the electoral formula (how votes are translated into seats); the ballot structure (how voter may mark their ballots); and district magnitude (how many MPs are chosen per district or, to use the Australian term, “division”). But there is also a much wider meaning, which we use in this book. It is the set of rules and practices that determine, among other things, who may participate in elections as electors, candidates or active supporters; what these various individuals are allowed to do in an election campaign, including raising and spending campaign funds; who conducts the election itself and how they do it; who settles disputes about what has happened; how electoral district boundaries are drawn and revised; and even how some people can get into parliament without being elected at all.14 Some of these rules and practices reflect broad principles debated during the evolution of our system. Should all adult men be allowed to vote, or women, or Aborigines, or eighteen-yearolds, or prisoners, or citizens who have left the country? Should important electoral functions be vested in the legislature, in the executive, in the courts, or in an independent, non-partisan body? Some rules and practices appear narrow and technical. What documentary or other evidence should someone produce 15

LIMITING DEMOCRACY

in order to have their name and particulars entered on the electoral roll? How many metres from the entrance to a polling place may campaigning activity take place? As will soon be apparent, both the grand and the narrow can be very controversial. The Australian constitution says relatively little about elections for the House of Representatives and the Senate, though what it does say is very important because of the difficulty of changing the constitutional text. Any move to lengthen current parliamentary terms, for example, requires the approval of the people voting in a referendum. When it was first legislated in 1902 the Commonwealth Electoral Act skilfully combined material from the electoral legislation of the six colonies, especially the progressive South Australian system. That legislation reflected a mix of Britain’s experience in the nineteenth century and Australia’s commitment to democracy, which was stronger, or earlier, than Britain’s. At times a few enthusiasts have looked enviously at US experiments with “direct democracy” – measures that involve the electorate in the legislative process through referendums and initiatives, or allow voters who had put someone into office to take them out again by way of the recall – but never for long and not now. The nuts-and-bolts experience from the half-dozen federal elections after 1902 was incorporated in the consolidated Commonwealth Electoral Act of 1918, but two other important changes were also made at about that time. Preferential voting – what the rest of the world calls the alternative vote – replaced “first-past-the-post” or plurality voting (then and still the British and US way) for both the House and the Senate, and compulsory voting was introduced for elections and referendums. Later, in 1948, the method of voting for the Senate was changed to the single transferable vote form of proportional representation, to which we will return. Originally, federal electoral administration closely resembled other activities of the commonwealth’s governmental machine. Parliament had passed the Act and it was amended, fairly often, 16

MODERNISING ELECTORAL DEMOCRACY

by the usual processes. One minister was responsible for the administration of the Act, but in practice the sensitive character of electoral matters effectively insulated them from partisan politics. Later the most senior electoral officials were made statutory officers, which emphasised their relative independence. Proposals for major changes to the electoral system came from the political realm, and public servants in the ministerial department and its electoral branch advised on their practicability and cost according to the standard Westminster practice. As a result, very little changed. With the advent of the Whitlam government in 1972 came the possibility of a substantial updating of electoral machinery and processes, but for the most part this impulse was frustrated by the known hostility of the Coalition-controlled Senate. Only two changes squeaked through, at the joint sitting following the 1974 double dissolution: legislation to enshrine the principle of one vote, one value, and the creation of two senators apiece for the Australian Capital Territory and the Northern Territory. The Fraser government undid the first of those reforms in 1977. The Hawke government came to office in 1983 with an inheritance of ideas from the Whitlam period, a widespread sense in the media and among the public that elections might matter more than had appeared to be the case during the protracted reign of Sir Robert Menzies and his successors from 1949 to 1972, and some awareness of new electoral technology and regulatory systems developed overseas. Hawke broke with tradition and turned electoral reform over to an all-party parliamentary committee – the Joint Select Committee on Electoral Reform, or JSCER. In 1983 the committee’s wide-ranging review of the electoral system produced a significant set of amendments to the electoral system – the largest renovation of the system since 1918. LINES ON THE MAP Perhaps the most important of the 1983 reforms, and certainly the most distinctive, related to electoral geography. The Hawke govern17

LIMITING DEMOCRACY

ment’s legislation contained a commitment to equal-enrolment electoral districts, reformed the method and timing of electoral distributions, and removed parliament’s right to veto redistribution proposals. These are not just technical issues: given that the primary function of the electoral system for the House of Representatives is to translate votes into seats, exactly where the lines are drawn on the electoral map can influence by whom and by how much an election is won or lost. When elections were introduced in the Australian colonies in the nineteenth century the responsibility for drawing boundaries fell into the hands of the colonial legislatures, which in practice meant the colonial cabinets. Boundary revisions took place frequently, fuelled by rapid population growth and spread, and unhindered by local government, which in some other countries (including Britain) successfully resisted separate national boundaries. New South Wales, drawing on New Zealand’s example, transferred initial responsibility from two intensely political bodies, the government and the legislature, to independent, nonpartisan bodies drawn from the judiciary and the public service. The other Australian colonies and the commonwealth followed suit, vesting responsibility for revisions either in a single commissioner or a small commission. There were three reservations of power. Final approval of the proposed new boundaries (and the names given to electoral districts) remained with the legislature, and was sometimes refused, most recently in 1974 when the Senate declined to pass the proposed redistribution for the House. The government of the day retained the responsibility for initiating the redrawing of boundaries and could delay or hasten the process for partisan advantage. And the independent redistribution authority had to work within a statutory framework drawn up by the government and legislature, which included qualitative and quantitative criteria to be applied in the boundary-drawing process. Such arrangements left electoral boundaries vulnerable to one electoral ill, malapportionment (numerical inequality of voters 18

MODERNISING ELECTORAL DEMOCRACY

among divisions), but relatively insulated from the other ill, gerrymandering (boundaries drawn for partisan advantage). In the United States, by contrast, malapportionment was outlawed from the 1960s but gerrymandering remains endemic. The qualitative criteria for determining electoral boundaries settled down to four regular items – community or diversity of interest, means of communication, physical features and existing electoral boundaries – and others that were less frequently prescribed. There were attempts to weight representation as a way of compensating for an electorate’s relative remoteness or dispersed population, to take account of local government boundaries (an easily identified sort of community of interest) and to anticipate population trends. Practice varied as to how the redistribution authorities presented their recommendations. Initially the commonwealth bodies gave some explanation of how they had applied the criteria; this gradually petered out. But in 1983, in keeping with contemporary reforms in administrative law, a duty to “give reasons” was imposed by statute. Some state bodies produced substantial documents, others got by with the barest minimum. Federally and for all state elections except the Western Australian upper house, the quantitative criterion fixed the permissible deviation from average enrolment per electoral district at the moment of redistribution. Originally, the permissable variation was usually 20 per cent but since 1974 – when the necessary legislation was forced through at a joint sitting of parliament – it has been 10 per cent. A short-lived provision (1977–83) required that all divisions smaller than 5000 square kilometres – urban divisions, in other words – should contain more electors than those of 5000 or more square kilometres. There is also a unique criterion that has been called “fairness.” Where the other criteria are more discreet, this is explicitly political in its intention. Since 1991 the South Australian House of Assembly electoral boundaries are reviewed after every general election to ensure that candidates for the party (including likely allies) that won more than 50 per cent of the two-party-preferred 19

LIMITING DEMOCRACY

vote would be elected “in sufficient numbers to enable a government to be formed” if the same votes were cast at the next election. Put baldly, a majority of votes should produce a majority of seats. The provision has been applied regularly in South Australia and the commissioners’ reports – as they grapple with the technical problems – make interesting reading, but results at the subsequent elections have not matched the intention of the legislation to the extent that had been hoped. Nevertheless, the experience has shown that parties and parliamentarians will put up with frequent redistributions more readily than had been thought likely. This notion of “fairness,” while attractive, cannot be guaranteed in single-member preferential systems, especially at the commonwealth level where redistributions are constrained by the constitutional prohibition of federal divisions straddling state boundaries. When other countries have attempted to match seats to votes, a common approach has been to draw from a pool of additional seats, either built into the system or available when needed, to “top up” the number of seat results to match the corresponding votes. This has worked well in federal Germany, but the objection usually made in Australia has been that “top up” members would be perceived as second-class, and that such a system might fall foul of sections 7 and 24 of the constitution, which mandate the “direct election” of federal parliamentarians. Evidence from the United Kingdom and Germany suggests that there may be trouble over the “second-class” problem but it is not inevitable.15 Commonwealth experience has differed from that of the states in several respects. The federal constitution, which has proved extremely difficult to amend, allocates equal numbers of senators to each state and guarantees a minimum number of lower house members to each of the states that existed at the time of federation, a provision that benefited Western Australia until 1949 and continues to benefit Tasmania. Occasionally, when one side holds all five Tasmanian seats, the other side says the loading is unfair. The provision that no House division may cross state 20

MODERNISING ELECTORAL DEMOCRACY

boundaries means that in states with relatively few lower house members average enrolments per member can vary significantly. The constitution left to parliament itself the question of how territories would be represented in federal parliament; the belated legislation sensibly made the provisions follow those for the states as closely as possible, but the small populations of the two territories means their representation in the House can swing widely. When the Northern Territory gained an extra seat its entitlement doubled from one seat to two, which could halve if it swung back again; recent legislation has sought to damp down the possibility to the limited extent possible while retaining equivalence with the states. The federal constitution permits changes in the number of members of the House of Representatives but requires that it be “as nearly as practicable” twice the number of senators from the states: the so-called “nexus.” Because there are six states and because half-Senate elections are the norm, increases in the House can only be by multiples of twelve. As a result, the federal parliament has increased in size on only two occasions, in 1949 when the number of senators rose from 36 to 60, and in 1984 when the number of senators elected from states (for by then there were two senators from each territory as well) rose to 72. In 1967 a referendum to break the nexus and so allow more frequent increases in the size of the House was easily defeated. With its decision in McKinlay’s case in 1975 the High Court introduced a further constraint. The court held that the constitution, in specifying that the number of members of the House “shall be in proportion to the respective numbers of their people,” required parliament to pay regular attention to the allocation of seats among the states. This requirement was implemented in 1977 by legislation prescribing that the allocation be set early in the life of each new parliament; if a state gained or lost a seat a redistribution of boundaries in that state had to follow. (In mid 2006, for example, Queensland gained a federal seat at the expense of New South Wales. The controversial 21

LIMITING DEMOCRACY

victim was the seat of Gwydir.) At the same time, separate legislation was passed to deal with the problem of unequal representation brought about by relative or absolute changes in enrolments in divisions within a state. According to this new provision, when a quarter of the divisions in a state departed from its average enrolment by more than 10 per cent, there should be a redistribution of that state. A separate but related problem is on the horizon: the growth of the voter population in each division. Excluding Tasmania and the territories (for which special rules apply) the average enrolment per division at the 2004 federal election was 88,000 voters – and a very high average of 95,000 in South Australia. At what point does the population of a division become so large as to constitute problems for local electoral administrators and representatives? There is no definitive answer, but the only solutions are either to increase the number of state senators to 84 and the House to about 170 members, which might encounter popular opposition – “More bloody politicians!” – or to try again to break the nexus by referendum, but there is no evidence that the 1967 defeat would not be repeated. The 1983 Act further tidied up the process. A state redistribution would still take place whenever seats were gained or lost, but now a maximum of seven years would elapse between redistributions; in other words, no more than three general elections could be conducted on the one set of boundaries. And there would be a redistribution if the populations of a third of the divisions in a state deviated from the average by more than 10 per cent. Since 1983 the first two provisions have been activated regularly, but never the third. This is a consequence of an additional innovation intended to maximise electoral equality whenever an election might be called: that the boundaries should be drawn so that the enrolments in divisions in three and a half years’ time – halfway through the seven-year period – will “as nearly as practicable” be equal. Divisions expected to grow rapidly are “handicapped” with below-average enrolments at the start; those expected to fall back 22

MODERNISING ELECTORAL DEMOCRACY

in the field are given starting enrolments above average. If the provision worked perfectly, each division would have a nil deviation over three consecutive elections conducted on those boundaries. Of course so accurate a set of demographic predictions is unattainable, and after a few redistributions it was feared that the qualitative criteria were being sacrificed to numerical equality – as had occurred earlier in the United States. In response, the provision that seemed to require absolute equality was altered by two stages to permit 3.5 per cent deviation. In the enrolment statistics published each month, only 2 to 3 per cent of divisions show a deviation from the state average of more than 10 per cent, and the greatest deviation is about 12 per cent. Deviations between 5 and 10 per cent are also uncommon (about 10 per cent of divisions at any one time). So this part of the system seems to be working well. But if the increasingly conservative US Supreme Court were to retreat from its commitment to the principle of one vote, one value, which has had considerable influence on Australian electoral thinking since the late 1960s, it is possible that the debate would be reopened in Australia as well. The redistribution process makes generous provision for public input, which usually involves political players – parties, members of parliament – and community interests seeking to place their supporters to best advantage in securing representation. Originally, proposed boundaries had only to be advertised and objections considered. Later, submissions were invited at the beginning of the process, then the public was given the chance to comment on the submissions prior to the initial drawing of provision boundaries – which could still be contested. These three classes of submission were published with the final recommendations. In 1983 the process underwent a series of related changes. To encourage uniform interpretations of the criteria, the Australian Electoral Commissioner would sit on every state and territory boundary commission. Second, when the time came to consider objections, the commissions would be enlarged to include the two 23

LIMITING DEMOCRACY

other members of the Australian Electoral Commission, the chair (a judge of the Federal Court) and a senior public servant (who in practice has always been the Commonwealth Statistician). Public hearings were introduced to allow those who had lodged written submissions to discuss and be examined on their views. (In 1987 another round of public input was added in cases where the revised boundaries are significantly different from initial proposals.) Significantly, the federal parliament abandoned its power to reject proposals and require new ones. The decisions taken by the second-stage bodies were now final and implemented forthwith. Of the 1983 changes to the process, the depoliticisation of redistributions is probably the most at risk of reversal. In 1983 one Liberal parliamentarian, Philip Ruddock, opposed the abolition of the parliamentary veto because, he argued, hostile state governments might manipulate the appointments of the two state members of the redistribution committees – their auditors-general and surveyors-general – to influence redistribution outcomes.16 A National Party MP, Evan Adermann, thought that recourse to the courts if the process strayed from the Act was inadequate protection: “How can a third party, even if it is a court, be the best authority of what the Parliament has said, written or intended?”17 But it is important to remember how, a decade earlier, the Senate, with the National Party the prime mover, had defeated redistribution proposals that might have been thought primarily a matter for the House. It did so, not because it disliked the boundaries that had been drawn per se, but because it disapproved of the new legislation under which they had been drawn – and specifically the reduction of the permissible variation in average enrolment from 20 per cent to 10 per cent. And the Senate took this action despite the fact that the legislation had been passed by the most demanding procedure the commonwealth constitution provides – a joint session of the two houses following a double dissolution. After 1983 a substantial degree of agreement settled over redistribution procedures. Major reversions, such as restoring the 20 per cent deviation from average enrolments, seem unlikely. 24

MODERNISING ELECTORAL DEMOCRACY

But where proportional representation is involved, which it is for the Senate, four state upper houses, one state lower house and one territory unicameral legislature, the choice of the number of members to be elected from a district has been and remains a politically sensitive issue. It is often said that the decision to enlarge the Senate in 1983 partly rested on a belief within the major parties that it would be harder for minor parties to win a seat if there were six vacancies at a half-Senate election instead of five. That seems counter-intuitive and certainly proved to be wrong. Senator Eric Abetz, the main proponent of the 2006 legislation, claims that the enlargement of the House and Senate was designed by Labor for its own advantage and that Senator Graham Richardson has admitted as much. As we’ll see, Senator Abetz’s quotation from Richardson was selective, to say the least. A much clearer illustration of the importance of “district magnitude” is what happened to the Tasmanian House of Assembly when the number of members in each district was reduced from seven to five before the 1998 election. In 1996 the Greens, with 11.1 per cent of the vote, won four seats; in 1998, with 10.2 per cent, they won only one. The Liberal share of the total vote dropped by 3.1 percentage points, from 41.2 per cent to 38.1 per cent, but the party lost more than a third of its seats, sixteen down to ten. With its vote up by 4.0 percentage points to 45.5 per cent, Labor won the same number of seats as before, fourteen, but now that number gave it a comfortable majority of seats. When Victoria finally opted for proportional representation for its upper house in 2003, the district magnitude of five places was sufficiently small to make it difficult for minor parties to secure representation when the first election is held under the new regime on 25 November 2006. For the “major” minor parties the numbers of representatives they have elected can translate into cold cash as well as parliamentary clout. When the composition of the small-party vote in the Senate changed in 2004, both the Greens and the Democrats lost staff and entitlements on what was, for them, a serious scale.18 25

LIMITING DEMOCRACY

THE COMPULSION TO VOTE Australia is not the only nation to operate compulsory voting, but it is one aspect of our system that is well known overseas – notably in the United States – and has been extensively analysed here. Compulsion was never in any doubt when the JSCER was reviewing the Commonwealth Electoral Act, and the 1983 legislation retained both compulsory enrolment (introduced in 1911) and compulsory voting (1924). The amended Act mandated compulsory enrolment for those qualified to be enrolled, and stated that “It shall be the duty of every elector to vote at each election.” Compulsion certainly works: 94 per cent of enrolled Australians vote at each federal election and referendum compared with a 35-year average of only 54 per cent for the US House of Representatives. Compulsion has always offended libertarians at both ends of the political spectrum, who regard it as “a form of paternalism”19 that denies citizens the right to withhold their votes as a censure of all candidates. But voters generally support compulsion, with a 2005 opinion poll registering 74 per cent approval. At the level of political theory, compulsory voting can be defended by invoking social contract theory.20 The Electoral Act prescribes penalties for any attempt to hinder a person’s right to vote or to “mislead or deceive an elector in relation to the casting of a vote.” This prevents partisans from discouraging specific groups of citizens from voting by spuriously challenging their eligibility or, as occurred in some counties at the 2000 US presidential election, establishing bogus roadworks to prevent voters getting to polling places. More positively, compulsion provides the Australian Electoral Commission with a clear mandate to conduct electoral education programs and run vigorous enrolment campaigns. Until recently compulsory voting enjoyed the support of all the parliamentary political parties. But a minority of Liberals has broken ranks on this issue as well. Historically, the Liberal Party has been an advocate of compulsion; its Victorian division, for 26

MODERNISING ELECTORAL DEMOCRACY

example, declared in 1956 that “there is no conflict with Liberal principles in maintaining compulsory voting.”21 But in the JSCEM’s report on the conduct of the 1993 election the Coalition members recommended a move to voluntary voting. Two current Liberal ministers, Senator Abetz and Senator Nick Minchin (both of whom have served as special minister of state), and a recent JSCEM chairman, Tony Smith, are vocal critics of compulsion. The JSCEM’s report on the 2004 election devoted a chapter to the issue but merely proposed “that voluntary and compulsory voting be the subject of a future inquiry by the JSCEM.” Nevertheless, the majority of Liberal parliamentarians appear to support the status quo and the Nationals do so strongly. Prime Minister John Howard, himself a long-time supporter of voluntary voting, has ruled out any change in the life of the current parliament. His reticence is partly the result of growing uncertainty about the impact of a reversion to voluntary voting. (No serious player is advocating voluntary enrolment.) It used to be contended that compulsory voting has advantaged Labor because members of its traditional “working class” constituency might be less likely to vote if they had a choice. But times have changed, and data from the 2001 census show that of the 30 federal divisions recording the highest rates of relative socioeconomic disadvantage, eleven are held by Labor, eighteen by the Coalition and one by an independent.22 On this basis, a fall in voter turnout might harm the Liberal Party and the Nationals more than Labor. There is another factor as well: the discussion of voluntary voting has centred on its impact on the House of Representatives and which party wins office, ignoring the real possibility that lower rates of voting in safe seats could significantly affect Senate results. One relevant body of fairly recent data has, as far as we know, been ignored by critics and defenders of compulsory voting alike. Voting in 1997 for delegates to the Constitutional Convention was not compulsory, although the opportunity to vote received substantial publicity. At the individual division level the substan27

LIMITING DEMOCRACY

tial fall in turnout lacked any obvious pattern, but statewide data, broken down by age of electors (see table opposite), confirms the potential of voluntary voting to disadvantage whichever party is more favoured by younger voters. THE BATTLE BEGINS Despite some occasional grumbling, the 1983 Act seemed to become settled policy over the next decade. But since the early 1990s it has come under attack from a number of directions. The first warning shots were fired as early as 1988, when the JSCEM released its report on the 1987 election. Coalition members had their suspicions: We believe that there is widespread concern in the community that the voting system has become so user friendly as to be susceptible to widespread abuse. Although the Committee has not received much hard evidence to support this, we believe that the existing situation is creating a lack of confidence in the integrity of the electoral system and thereby damaging our democratic processes.23

Some political scientists also voiced concerns – Joan Rydon bemoaned “the complex electoral laws”24 – and the electoral activist Amy McGrath alleged that the 1983 amendments had opened up the electoral roll to widespread fraud.25 By 1993 the bipartisanship of 1983 had well and truly evaporated. In 1996 McGrath became a founding member of the H. S. Chapman Society, which describes itself as “a non-political watchdog of the integrity of our voting system.” The Chapman Society has given evidence to successive JSCEM inquiries and its claims of widespread electoral fraud have been eagerly promoted by radio personalities like Alan Jones, who launched McGrath’s book The Frauding of Votes in 1999. According to the society, most of the flaws it sees in the electoral system can be attributed to the 1983 changes. This is a view that appears to be shared by the minister responsible for electoral legislation from 2000 to 28

MODERNISING ELECTORAL DEMOCRACY

Voluntary voting for the 1997 Constitutional Convention: participation by age

NSW Victoria Queensland W Australia S Australia Tasmania ACT NT Australia

18–25

26–35

36–45

46–55

56–65

65+

32.4 41.3 30.5 27.9 30.0 36.3 38.1 29.3 33.9

35.5 45.2 35.6 33.3 24.2 40.0 44.0 35.8 38.0

41.8 50.0 42.0 41.4 42.0 46.4 52.8 41.9 44.2

47.4 54.1 49.1 48.2 51.1 52.5 59.0 47.9 50.1

56.1 62.2 59.4 58.0 62.2 60.1 66.5 49.4 59.2

57.1 60.0 60.1 60.0 61.7 59.6 68.3 49.8 59.2

Source: Australian Electoral Commission, 1997 Constitutional Convention Election Report and Statistics, 1998

early 2006, Senator Eric Abetz (Liberal, Tasmania), who has participated more actively in public debate than did his three predecessors in the Howard government. Critics of the 1983 legislation gained more ammunition from legislation that required the Australian Electoral Commission to conduct trade union elections. When evidence emerged of corrupt practices in these ballots some commentators simply assumed that the same practices must be influencing parliamentary elections. For example, over a quarter of McGrath’s book Corrupt Elections 26 deals with trade union elections. Marshall Cooke, a former Liberal member of the lower house who had unearthed fraud in union elections in Queensland as a state royal commissioner, supposed that such practices would have spread to state and federal elections: Why would you stop? If you think it’s all right to get power in a union election by those sort of means, why not where the prize is bigger? I would be very suspicious about two or three of the Hawke elections and perhaps one of the Keating elections.27

But he provided no evidence. Nor did the former Liberal leader 29

LIMITING DEMOCRACY

John Hewson when he asserted that “we all know of examples of fraudulent enrolment and fraudulent voting” and pointed to the chief culprit: “the ALP has a continuing vested interest in electoral fraud.”28 No stronger on evidence was a speech that Senator Abetz gave at the Sydney Institute on 4 October 2005 entitled “Electoral Reform: Making Our Democracy Fairer for All.” In this first public overview of the impending 2006 legislation, Senator Abetz, who was the special minister of state at the time, identified a series of milestones in the development of the federal electoral system. “But,” he continued: it is an historic fact that the biggest suite of changes in the Commonwealth Electoral Act came about in 1984 [when the 1983 changes came into effect], when the Labor Party and the then Special Minister of State, one Kim Beazley, sought not to strengthen Australia’s democracy, but rather to alter it to suit the Australian Labor Party’s own political advantage. To quote former Senator Graham Richardson, in his book Whatever It Takes, these changes were made so: “… Labor could embrace power as a right and make the task of anyone trying to take it from us as difficult as we could.”29

If the intent of the speech was the demonisation of the 1983 changes, bringing forth the Demon King himself was a plausible start. The trouble begins when the conscientious reader follows up the speech’s footnote and reads what Richardson said next: “For three decades, conservative governments had written and rewritten the electoral laws to suit themselves.” Read a little further and we find the National Party, rather than Labor, pushing what would become, for the Liberals, one of the more contentious elements in the Bill: Ralph Hunt, a genial, nice man who was the National Party representative on the [JSCER], secretly approached Dick Klugman [its chairman] with the idea that the Parliament should be expanded. 30

MODERNISING ELECTORAL DEMOCRACY

He and Ian Sinclair could see how much their party was in decline and they knew that the only losers from the proposal would be the Liberal Party.30

In the event, the National Party secured five of the additional seats created, enough to maintain its numbers for the time being (albeit in an enlarged House of Representatives), and the allparty committee recommendations passed through both houses of parliament overwhelmingly intact. In his Sydney Institute speech Senator Abetz hailed the lowering of the voting age from 21 to eighteen as a significant improvement to Australian democracy, in implicit contrast with the reforms of 1983. In fact, it was the Whitlam Labor government that lowered the voting age in 1973, after years of resistance by successive Coalition governments – though a handful of more liberal Liberals were said to have agreed with the measure. The Snedden-led opposition had the good grace to welcome the change, albeit with an interesting warning that Senator Abetz may well have had in mind when he drew up the 2006 legislation: The Opposition parties recognise the great impact the 18-year-old vote will have on future elections. The voters to be added to the rolls as a result of this measure will be of such numerically significant proportions that they will be able to determine to a large measure the political complexion of future governments.31

FUNDING AND THE FRANCHISE By mid 2006, proposals for changes in the electoral procedures could be divided into three categories: a few already launched, rather more still on the way, and a number strongly suspected from hammering noises coming from behind the scenes. The Howard government had signalled its electoral law intentions by the passage of the Electoral and Referendum Amendment (Electoral Integrity and Other Measures) Act 2006, which includes measures to close the electoral roll significantly earlier than has 31

LIMITING DEMOCRACY

been the norm, to make enrolment less straightforward, to deprive all prisoners of the right to vote, and to make donations from trade unions to the Labor Party more transparent and many other political donations less so. None of these measures is justified by the evidence the government has advanced. The early closure of the electoral rolls, for example, potentially disenfranchises as many as 300,000 Australians who are otherwise eligible to vote. Under the 1983 legislation, two types of voter – first-time voters and those who had voted before but whose residential details had changed – were given seven days’ grace after the calling of a federal election to bring their enrolment up to date with the AEC. The numbers can be quite large: in the seven-day period after John Howard called the 2004 election, 78,908 firsttime voters enrolled and 255,000 voters changed their details. These numbers are often advanced as proof that it is impossible for the AEC to adequately check the validity of the changes in the time available, despite the fact that it has been doing so for twenty years. Enrolment changes are processed at AEC divisional offices, so we need to divide the numbers by 150 to assess the real workload – which cuts around 78,000 new voters to a less intimidating average of 520 per office and deflates the threat accordingly. The 2006 legislation reduces the period of grace to three days for existing voters, and no time at all for first-time voters. Convicted prisoners, meanwhile, will have no right to vote at all. To understand how these significant curtailments of democratic rights could take place in a country with a longstanding commitment to compulsory voting, we need to look back over the debate about “electoral integrity” that has been simmering since the 1980s. Although there is no evidence that the integrity of the Australian electoral process has been compromised, the critics continue to argue that elections can – and have been – rigged by electoral conspirators allied to the Labor Party. Their view has won influential converts and helped shape the 2006 legislation.

32

CHAPTER 2

Electoral integrity: the rise of the activists

A

t first glance “electoral integrity” seems like one of those pious slogans on which all parties can agree. The piety is sufficiently attractive to have secured the term a parenthetical place in the short title of the 2006 Act – “Electoral Integrity and Other Measures.” Since the 1983 changes allowed eligible citizens to enter their names on the electoral roll observed by only one witness and with no other proof of identification, the term “integrity” has become politically charged. Reflecting American concerns, the Chapman Society, some conservative politicians and radio “shock jocks” claim that the openness of the roll is an invitation to electoral fraud. For some, “integrity” is code for tightening the enrolment process in a crusade against alleged fraud. Two preliminary questions have to be settled. Who is eligible to be on the roll and vote (a matter discussed previously)? And what sort of a roll is to be compiled? A new roll for each election (as Canada had until recently)? A new roll once a year (like the United Kingdom)? A multitude of rolls compiled and maintained by the constituent states of a federation applying their own laws (like the United States)? A national roll compiled as and when required from another general database of citizens like an internal passport system (as some European countries have)? A single, national, continuous roll compiled for the purpose (as Australia 33

LIMITING DEMOCRACY

began with at federation, which has gradually incorporated the various pre-existing state rolls into the “common roll”)? The last option seems to be settled policy here. While each state certainly retains the constitutional power to restart its own, separate roll if it had reason to do so (for example, if the commonwealth continues to restrict easy access to the common roll), the cost and the inconvenience it would cause make this undesirable. Computerisation allows slight variations in enrolment rules between jurisdictions by flagging those individuals, such as prisoners, who raise special problems because of different eligibility rules among the commonwealth and the states. The third question, and this is where the controversy about roll integrity is to be found, is how do electors get their particulars entered onto the roll? The information may be copied from another database, and this has recently become a possible starting point for the enrolment process. Otherwise there are two main alternatives: the government takes the initiative by sending forms to every household or address (like the United Kingdom) or sending “enumerators” (a representative from each of the two major parties in that electoral district, working in pairs, as Canada did); or the initiative is left to electors, who go to a prescribed place (as used to happen in some parts of the United States) or collect a form and return it filled in (as in Australia). There can be hybrids, as in Australia for most of the twentieth century: although voters were required to take the initiative, casual electoral staff would make house calls (the “habitation review”) to check the accuracy of current records and add eligible electors who had not exercised their initiative. Because it had been around so long, the habitation review attracted little attention from those who declared that corruption of the roll was rampant – apart from one passing allegation that it was illegal and “enabled the AEO [the Australian Electoral Office, which preceded the Australian Electoral Commission] to enfranchise dwellings and disenfranchise people, including persons temporarily overseas, and itinerant workers, for example.”32 That 34

THE RISE OF THE ACTIVISTS

critic, the electoral activist and High Court litigant Helen Berrill was also bothered by the fact that the habitation reviews involved creating a modified version of the roll, arranged by address rather than surname; she argued that this was also illegal under the Act. Subsequently one of the 1983 amendments came under attack from the Chapman Society on the ground that it made that version of the roll available to registered political parties (who were entitled to check roll accuracy). According to the society, “crooked party workers” could then replicate the house-to-house inquiries of the AEC’s employees, note who had moved away but were still on the roll, and then arrange their impersonation on polling day.33 But it was not quite as simple as that. Anyone could acquire a copy of the traditional surname roll and convert the data into an address-ordered list with a minimum of effort on their computer’s part, so the potential to match names to homes had long existed. And regardless of how it was obtained, actually using the list would never be straightforward: the “crooked party workers” would need to check their door-knocking data against the certified lists after the close of roll to make certain names had not subsequently been deleted. The habitation review made a brief return in 2004 in the Victorian division of Isaacs. The JSCEM had recommended to the AEC that “the integrity of the Electoral Roll be tested by a total habitation review of a sample electoral division in a State which has not had an election in the 12 months preceding the habitation review” – in other words, well outside the burst of enrolment activity that the calling of an election brings on. The methodology and the findings of the exercise are available in the subsequent report, published in April 2005, which is accessible on the AEC’s website.34 The report concluded that the results of the survey of over 50,000 residences and almost 90,000 electors “demonstrated that the roll in the Division of Isaacs is one of high integrity in terms of accuracy and completeness.”35 It should be noted that the review cost approximately $280,000 and only caught one moment in time. In the less than three months between completion of the 35

LIMITING DEMOCRACY

review and the close of rolls for the 2004 federal election there were 847 new enrolments, 776 re-enrolments of people who had dropped off the rolls, 1706 transfers in from other Victorian divisions and 174 from interstate, 1637 transfers out to other Victorian divisions and 221 to other states, and 159 deaths.36 These changes highlight the dynamic nature of a continuous roll and provide hard data instead of feverish speculation. The important thing is that the roll should be accurate and complete on polling day rather than at any random point between elections. To hope that education and law enforcement will keep the roll accurate at all times is to ignore the role of state and local government elections in maintaining accuracy.37 Two criteria are likely to influence which method(s) are selected to compile the roll: cost and practicality, and whether the system is intended to encourage or discourage enrolment. Changes in urban life – all members of families at work during regular hours, fear of casual violence in “tough neighbourhoods,” large apartment buildings that cannot be entered, declining knowledge as to who lives next door – undercut the habitation review, and other arrangements were always necessary for rural areas. Nevertheless cost and practicality ought to be capable of objective assessment most of the time, as evidenced by the lack of controversy over the AEC’s decision to abandon habitation reviews gradually and switch to mining other data sources, retaining habitation review–type activities for a sample of small areas or to investigate suspect data (several different surnames at a single address, for example). Objectively assessing the ease or difficulty of enrolling is harder. To try for a moment to distance our discussion from current controversies in Australia, it is useful to look at another country and a past century for a case of misapplied reformist zeal. In the United States between the end of the civil war (1865) and the start of the first world war (1914) a series of measures were introduced by self-proclaimed reformers of blameless reputations, ostensibly in the name of roll integrity and the avoidance 36

THE RISE OF THE ACTIVISTS

of electoral fraud generally. These reforms were regarded at the time and for quite a while afterwards as having been A Good Thing. But a later generation of historians concluded that the reforms had, at best, frustrated the emergence of disciplined political parties and, at worst, prevented the enfranchisement of immigrants who were Catholics, or came from southern and eastern Europe, or were blacks and poor whites, depending on the part of the country concerned.38 Australia inherited the British approach, which saw the franchise as primarily a property right. As such, the right to vote was eminently suitable to be tested through a judicial process. In the Commonwealth Electoral Act 1902 this took the form of revision courts, composed of a stipendiary magistrate or a pair of justices of the peace, in which suspect enrolments might be challenged. By 1918 the right to challenge had been reduced to a right to lodge an objection with the divisional returning officer, the public servant responsible for the roll, with the discouragement of a five shilling deposit, returnable if the enrolment was thrown out, and payment to the person objected to of up to five pounds if the officer thought the application frivolous. The deposit to accompany an objection, subsequently $2, remained, the costs payable to a successful objectee disappeared, but the main point is that the right to object was never exercised. In 1984 a review of the situation found that the prescribed forms for lodging such objections were out of print, but a reprint failed to attract any business. In colonial times in Queensland elector-initiated challenges were sufficiently common for the name of one partisan and prolific objector to become a verb: to “Bulcock” a roll meant to get one’s supporters on and to get one’s opponents off. But the electorinitiated challenge has never been popular in relation to federal rolls. Why? Explanations tend to rely on the argument that it is too hard or too expensive. As one Chapman Society stalwart, Allan Viney, asked rhetorically: “Would anyone present like to try to organise enough people with the skills and dedication that would be required for a successful ‘door knock’ for even just one marginal 37

LIMITING DEMOCRACY

electorate?”39 By 2000 the cost of lodging an objection to have a specific name struck from the roll had been transmuted into a purely imaginary cost for looking at any one name on the roll, which was in fact free and widely available. Drawing on Chapman Society material, an editorial in the Sydney Morning Herald erroneously stated that “Individuals who uncover fraudulent names on the roll, for instance, have to pay the AEC $2 to examine each name. This is costly if hundreds of names are involved.”40 The answer to Mr Viney’s rhetorical question is that if, as alleged, defective enrolments are sufficiently common to “fix” the results of marginal electorates, a small sample in one division would prove the case. Any political party that can spend $50 million on an election campaign can afford that cost; one anonymous benefactor making a $10,000 donation would be sufficient. Senator Minchin is even more pessimistic about the prospect of unearthing fraud: “The Liberal Party has always been concerned that deliberate fraudulent enrolment is almost impossible to detect.”41 Senator Minchin and the Chapman Society seem to be saying that the very absence of evidence is itself suspicious – an attitude reminiscent of Lieutenant General John L. DeWitt’s notorious comment about Japanese internees in the United States during the second world war: “The very fact that no sabotage has taken place to date is a disturbing and confirming indication that such action will be taken.”42 Speculation thus becomes hard evidence: if one false enrolment is possible, goes the argument, then hundreds or thousands can be registered. The fact that the machinery for challenging the roll is not used may well be a good thing. The process could potentially be misused to discourage targeted populations of electors by posting them registered letters which they refuse to accept, then using the non-delivery to justify lodging challenges to the addressees’ enrolments shortly before polling day. The hope and expectation would be that the enrollees would not bother to go to their precinct polling places to clear the matter up – as is alleged to have happened in Ohio during the 2004 US presidential election. 38

THE RISE OF THE ACTIVISTS

PRESELECTION AMMUNITION The JSCEM’s 2000–01 inquiry into the integrity of the federal roll was triggered by three cases of enrolment fraud intended to influence Labor Party preselection ballots in Queensland. To cast a vote in a Labor preselection ballot in that state, party members must be on the electoral roll for the relevant division. The trouble began when AEC divisional staff in Herbert, who were processing enrolment forms, noted suspicious irregularities; they turned the evidence over to the federal police, and eventually prosecutions were brought under the Crimes Act, which offered more substantial penalties than did the Commonwealth Electoral Act. It is worth noting that AEC officers did detect the miscreant enrolments, which undermines the argument that roll fraud is easily achieved. One of the three accused, who had pleaded guilty to 24 counts of forging and 23 counts of uttering, alleged that electoral fraud was widespread. The notoriety of the cases led to a preliminary inquiry by independent counsel initiated by the state’s Criminal Justice Commission. That commission was followed by the Shepherdson Inquiry into four Labor preselection ballots and one parliamentary byelection. The inquiry suggested two main lines of reform. One was “better procedures for identifying people” when they applied to enrol for the first time and thereafter for changes of address; the other was to bring the party preselection process under a legislative regime with penalties for malpractice. Shepherdson identified two main categories of false enrolment: forgery, where people are falsely enrolled without their knowledge, and consensual false enrolment, where people participate with full knowledge and consent. But he concluded: The information gathered during the Inquiry clearly established that the practice of making consensual false enrolments to bolster the chances of specific candidates in preselections was regarded by some party members as a legitimate campaign tactic. No evidence, 39

LIMITING DEMOCRACY

however, was revealed indicating that the tactic had been generally used to influence the outcome of public elections [our emphasis]. Where it was found to have been used in public elections, the practice appeared to have been opportunistic or related to the family circumstances of particular candidates rather than systemic or widespread… The inquiry uncovered evidence of forgery, but there was great difficulty in obtaining evidence to establish who was responsible.43

The inquiry found that the state’s deputy premier, Jim Elder, had signed two false enrolments as a witness, and that a handful of enrolments had been lodged using his home address. In other words, the enrolments that destroyed Mr Elder’s political career involved a few relatives in an electoral district where his majority had never been fewer than 3000 votes. By an interesting coincidence, the parliamentary by-election involved in this affair appeared to be exactly the sort of event to trigger serious concern about roll fraud. At the 1992 general election Labor easily won the redrawn and renamed electoral district of Mundingburra (11,753 to 8033, a majority of 3720), but at the next election, in 1995, it only retained the seat by the narrowest of margins (9308 to 9292, a majority of sixteen). That result was set aside by the Court of Disputed Returns for reasons that did not involve roll integrity or partisan malpractice. The subsequent by-election was won by a Liberal candidate (10,261 to 9177, a majority of 1084), which gave the opposition a majority in the Assembly, bringing on a change of government. Faced with this momentous outcome, the Shepherdson Inquiry arranged for the Criminal Justice Commission to investigate the approximately 140 names added to the state district roll between the decision of the Court of Disputed Returns (when it became clear there would be a by-election) and the close of rolls prior to the by-election. All but three names checked out. Two of the three were Labor supporters who indicated they had been asked to enrol, improp40

THE RISE OF THE ACTIVISTS

erly, so they could vote in this vital by-election; the other was a friend of the successful candidate and said it was his own idea. The Criminal Justice Commission also examined instances where there were more than four electors at one address, or electors with different surnames at one address, or where the enrolled address of an elector was not their postal address, and found “no impropriety.” At the 1998 general election the seat returned to Labor (9994 to 8486, a majority of 1508). The size of those swings over a period of six years shows how very difficult it would have been to fine-tune the local roll to achieve a particular result, and the same could be said of most electoral districts that fall into the “highly marginal” category at a particular election. It is important to remember that voting fraud can come in either of two forms. One is that a ballot paper marked for A is altered, or miscounted, so that it goes to the credit of B. Allegations that this has happened in Australian elections are extremely rare because of the non-partisan character of polling booth staff, though it is a relatively common complaint in some other countries. It has been very common in the United States recently and that experience points to some of the serious difficulties that will arise if Australia goes down the path of electronic voting – especially if the technology does not generate a paper ballot that can later be checked.44 The second variety of fraud occurs if a ballot paper recording a vote for B is introduced to the count when the supposed voter is not properly on the roll for that district and should not have been given a ballot paper in the first place. The first fraud closes the gap between A and B by two votes, A losing one and B gaining one; the second closes the gap by only one vote, that which B has received. This second category of fraud requires twice as many deceptive acts to achieve the intended objective, and when that is to alter the result of a general election the numbers become very large and the fraud difficult to conceal. In a recent newspaper article the former Liberal leader, John Hewson, another proponent of fears about fraud, quoted Senator Abetz as saying that 41

LIMITING DEMOCRACY

nine seats had been held by a “margin of 1 per cent” (presumably at the 2004 election), intending to show how easy it would be to turn an election around.45 If that 1 per cent margin means winning by no more than 50.5 per cent over 49.5 per cent then there were four such seats, three won by Labor and one by the Liberals. If it means winning by 51 per cent, then there were eleven such seats, six won by Labor and five by the Liberals. At the 2004 election the Coalition won 27 seats more than Labor, 87 to 60, with three independents. The total winning margin in the eleven seats fell a whisker short of 9000 and the average margin was 818. That gives some idea of the size of the underground army of conspirators required to translate Dr Hewson’s rhetoric into reality. Other inquiries were taking place in parallel with Shepherdson and the Criminal Justice Commission. The Queensland parliament’s legislative, constitutional and administrative review committee was looking at the best way to minimise electoral fraud at state elections. Commencing in May 2001, the Australian National Audit Office, or ANAO, undertook a performance audit of federal roll-keeping. And the JSCEM then conducted an inquiry into the ANAO’s audit. Together, the five inquiries provide 600 pages on the subject of electoral fraud, which ought to be enough to get a grip on any problems that existed. Given the ANAO’s expertise and political neutrality, it might be sufficient to look there for any evidence of roll fraud. The audit defined “roll integrity” as having four elements: accuracy – the information about individuals is correct and upto-date; completeness – the roll includes all individuals who are eligible to be on it; validity – it includes no one who is ineligible; and security – the roll is protected from unauthorised access and tampering. Its principal check of integrity was comparison of the 12.6 million roll entries with an even larger database, the more than 18.4 million Medicare records. The audit’s conclusions were that, “overall, the Australian electoral roll is one of high integrity, and that it can be relied on for electoral purposes.” It went on: 42

THE RISE OF THE ACTIVISTS

We concluded that the AEC is managing the electoral roll effectively. AEC policies and procedures can provide an electoral roll that is accurate, complete, valid and secure. In particular, the AEC has mechanisms in place to provide assurance that the names and addresses on the electoral rolls are legitimate and valid; and that people who are eligible to vote are registered properly. At the same time there are areas of AEC management of the roll that can be improved; in particular by better targeting and expansion of the data sources currently used to update the roll, by strengthening strategic relationships with key stakeholders, and by better identification and management of risks to the integrity of the roll.46

The JSCEM’s inquiry was critical of the ANAO’s matching exercise because, while it tested name and date of birth, it did not test address and it was essential to know that the person was at the right address in the right division in the right state. The JSCEM made a series of recommendations to the AEC. It wanted the AEC to adopt performance indicators more widely and to report more about implementation of the various recommendations and about specific enrolment matters (such as the cases where enrolment had not been accepted immediately). It also wanted the AEC to improve continuous roll update procedures – by drawing on a larger range of databases in New South Wales and Victoria, for example – and conduct both “a total habitation review” (which became the Isaacs exercise already described) and random spot checks of a sample of addresses. The JSCEM’s 2001 election report reviewed progress on improvements in roll integrity and heard AEC responses to various submissions, including proposals to revive regular habitation reviews and issue general mail-out warnings a year before a general election is due. As usual most criticisms focused on securing maximum accuracy at one moment in time, and the roll administrators pointed to the volume of changes that were always taking place. 43

LIMITING DEMOCRACY

SIMPLICITY UNDER CHALLENGE Instead of seeking to reactivate the machinery for elector-initiated challenges, those who claim enrolment fraud has taken place or could take place have turned their attention to making enrolment more difficult. Following the passage of the 2006 “electoral integrity” legislation, documentary evidence of eligibility is required when an adult initially applies for enrolment and at subsequent changes. Enrollees must now produce a driver’s licence as proof of identity or, if they don’t have one, some other identity document, or they must have their application countersigned by two electors who can attest to their identity. As well as ignoring the extent of licence fraud, this requirement constitutes a poll tax (the cost of a driver’s licence or equivalent) by the back door, unless citizens are to be issued a free identity document as has been recommended for the United States – which does not seem to be the intention for the proposed Australian “Smart Card.”47 The rhetorical flourish used by Senator Abetz and others to make the case for tightening enrolment was that getting on the roll should not be easier than hiring a video or joining a junior football team. Unfortunately for the accuracy of the comparison, neither of those activities is compulsory or near-universal among the adult population, whereas enrolment for the federal roll has been so since 1911. These arguments are almost a mirror image of the principles nailed to the mast by the all-party JSCER in 1983. In those days its members – from all parties – were more concerned with getting previously disadvantaged potential electors onto the roll than keeping suspects off: The Committee believes as a matter of principle that procedures to enable electors to fulfil their obligation to enrol should be as simple as possible, and non-compliance with the more technical and formal requirement is not a proper basis for invalidating an elector’s claim.48 44

THE RISE OF THE ACTIVISTS

Critics would later use the point about simplicity as part of their weaponry. They argued that “simple” meant easily abused and, because the case for simplicity allegedly derived from the compulsion in compulsory voting, the abolition of compulsory voting would remove the need for simplicity. As the dissenting minority of Coalition members in the 1993 JCSEM election report put it: The determination of the AEC to resist attempts to reduce the possibility of fraud is to a large extent based on the premise that, because the law forces people to vote, the Electoral Act should make it as easy as possible for people to enrol and vote.49

The use of the simplicity argument by critics of the 1983 reforms required some nimble stepping from “obligation to enrol” to “obligation to vote.” One of the many paradoxes wandering through the debate is that compulsory enrolment is usually ignored, or when it is mentioned nodded through as helpful to roll accuracy. In principle, opponents of compulsory voting should be just as opposed to compulsory enrolment. But the issue has probably been left alone because many parliamentarians find the roll useful for identifying constituents and, more recently, as the foundation for their local databases for constituency services and re-election campaigning. Additionally, rolls have always been useful to commerce, to small businesses like door-to-door salesmen and to big businesses who, it was alleged, used to have the printed rolls shipped overseas to data-entry havens. Computerisation and the supply of electronic rolls to members of parliament and registered political parties, which also made commercial uses much easier, led to legislative amendments aimed at curbing nonpolitical use. When it was originally introduced by the Fisher Labor government in 1911, compulsory enrolment was justified on the basis of the usefulness of the list for a number of government services; law enforcement and taxation authorities continue to be 45

LIMITING DEMOCRACY

substantial users. In an era of terrorism it is unlikely that any Australian government would propose following the United States as far as voluntary enrolment. Neither is it likely they might adopt one recent remedy for the sad state of US rolls, enrolment at the polling place on polling day immediately before getting your ballot paper. Senator Abetz and Senator Minchin remain enthusiasts for, at least, voluntary voting, but the JSCEM review of the 2004 election did not recommend the change and Prime Minister Howard is not enthusiastic.50

46

CHAPTER 3

Electoral integrity: slamming the door

T

he critics put forward another, more narrowly focused, proposal for roll integrity, which involved changing the time when the roll is closed prior to an election or byelection. Once more the story begins in 1983, just prior to the general election. Established practice had been for the prime minister to announce the date of a forthcoming election to the parliament or, if parliament was not in session, via a public statement. Some days later, the writs for the election would be issued and the rolls would close for both additions and alterations at 6 pm that day. The period between the election announcement and the issue of writs varied, but since 1940 the average gap between the calling of the election and the closure of the rolls has been just over nineteen days. Prior to 1983, the shortest intervening period between the announcement and the close of rolls was eight days in 1980. But in 1983 the writs were issued on 4 February, the day after the election had been called, and consequently the rolls closed at 6 pm that day, effectively excluding many thousands of citizens from voting. Four cases were brought in the High Court by people who had been placed on the New South Wales roll after the commonwealth roll had closed and had been refused the right to vote in the election. They relied on section 41 of the Australian constitution, which appeared to guarantee a federal voting right to all people who had the right to vote in a state election. Six justices 47

LIMITING DEMOCRACY

held that section 41 was a transitional section covering the short period before the commonwealth passed the first federal Franchise Act in 1902.51 Justice Lionel Murphy, in a solitary dissent, pointed out that the prime minister had been saying there would not be an early election and, closer to the event, that he was exploring the possibility of an election in 1984, and observed that at other elections there had usually been “some reasonable warning that the election is imminent.” Not to apply section 41 any longer would leave it “open to the Federal Parliament to restrict the federal franchise,” he wrote – by disqualifying people in receipt of charity, for example, an option that had been common to several of the federating colonies. In addition to constitutional and human rights questions there was a matter of practical politics. One of us wrote at the time: It is impossible to say how many new voters were prevented from getting on the rolls, though the media spoke vaguely of hundreds of thousands. As the Morgan Gallup Poll surveys during February showed the ALP doing about 10 percentage points better with the under-35 age group than with the over-35s, and as the under-35s would have been more affected by the sudden closing of the rolls, it is reasonable to assume that the ALP was disadvantaged. It is most unlikely that this was a consideration in the Prime Minister’s mind when he rushed through his double dissolution.52

The JSCER subsequently considered that “the closing of the rolls almost immediately an election is announced as occurred in February 1983, is not in the best interests of parliamentary democracy” and recommended a proclamation by the governorgeneral to announce the election date a minimum of seven days before writs are issued.53 For an unknown reason, a decision was made at the drafting stage of the 1983 legislation to deal with the issue of the writs and the closing of the rolls separately, and this has allowed critics to claim that the new provision was an inno48

SLAMMING THE DOOR

vation when it was no more than a guarantee that previous practice would be observed. The JSCER’s recommendation went into the Act, and the JSCEM said little about the matter in its first two election inquiries. The 1987 report noted that “one submission alleged that large numbers of dummy enrolments were made before the close of rolls” but that, while individual cases had been cited, “no evidence of an organised scheme was presented.” It suggested the AEC examine enrolment changes in two marginal divisions, one from each side, for the months before the next election and compare them with the habitation reviews after the election.54 It also suggested a substantial increase in penalties for electoral fraud. More generally, the committee considered options including “a computerised election system” (which may have meant having the certified lists available online across the country to prevent multiple voting in one name), more proof of identity at enrolment and polling, and “locality” voting, but decided they were not warranted yet. The JSCEM’s 1990 report returned to ignoring the matter. By 1993, however, the issue had become prominent. There had been some 160,700 new enrolments and 296,000 other changes in the period after the writs were issued, and the JSCEM reported that it had received submissions proposing that rolls should close on the day the election was announced “to prevent false names being registered for multiple voting purposes.” The committee noted that the AEC “has conceded that there is virtually no scope for any detailed check” of the enrolment forms during that period. Following the 1990 election and pursuing the 1987 report recommendation, the AEC had audited six divisions – two government marginals, two opposition marginals and, as controls, two safe seats. It had investigated the 23,240 new enrolments and found only 72 cases of wrongful, not necessarily fraudulent, enrolment; these amounted to 0.02 per cent of the total enrolments in those divisions. It had also monitored enrolment movement prior to the 1993 election and found that new 49

LIMITING DEMOCRACY

elector enrolments were fewer in marginal divisions than in nonmarginal, and that they constituted a smaller proportion of total roll activity as well. But the battle lines had now been drawn. The Labor majority on the committee opposed early closing of the rolls and considered its consequences – out-of-date enrolments and the denial of the franchise to new electors – “would be highly regrettable.” The Coalition members thought that the system was “vulnerable” in the period, and suggested a modest corrective step. Rather than encouraging the last-minute rush (when it advertised heavily to bring them in), the AEC should advertise extensively over the six-month period before the life of the parliament ended and move to a continuous roll review, a computerised process drawing on a range of databases.55 By the 1996 report the balance of numbers on the JSCEM had changed in favour of the Coalition. This time there had been just under 101,000 new enrolments and 328,000 changes after the election was announced. The AEC said detailed checking was “virtually impossible” and acknowledged that “those concerned about the integrity of the rolls” often called the seven-day period into question. Accordingly the Coalition majority concluded that the rolls should close for new enrolments at 6 pm on the same day the writs are issued, and for changes on the third day after that. The Labor minority was against the changes, arguing that it was the advent of an election campaign that jogged memories rather than “an orchestrated campaign to fill rolls by political activists,” and that a “very significant portion of the electorate” was involved. The minority concluded that “under the guise of outlandish, unsubstantiated claims about the feasibility of fraud, vast numbers of Australian citizens will be deprived of a vote.”56 The Coalition still lacked the numbers in the Senate, however, and though a number of changes proposed in the 1996 report were implemented by the Electoral and Referendum Amendment Act 1998, the controversial section 155 remained. Prior to the 1998 election the late enrolment rush was somewhat smaller, a total just short of 352,000 – only 82 per cent of the 50

SLAMMING THE DOOR

1996 rush. Because the general election was premature and there had been another major electoral event in the interim, the Constitutional Convention election, the rolls had been in better shape to begin with. The Coalition majority remained concerned about “potential inaccuracies” and reiterated its 1996 recommendations to “preserve the integrity” of the rolls. But it did acknowledge the AEC’s effort to enhance the accuracy of the rolls through a new address register listing “a range of attributes for each address including a land use code, occupancy status, an enrolment limit, the last review date and whether the address is habitable and valid for enrolment.” According to the JSCEM the register promised to become “an increasingly powerful tool for the AEC to detect fraudulent or inaccurate enrolment by identifying addresses incorrectly described or duplicated on the Register, those that have a high number of enrolments or re-enrolments, and those that have two or more groups of electors resident with different family names.”57 THE CRITICS ENCOUNTER RESISTANCE The JSCEM’s 2001 report displayed a striking change of opinion. The Coalition majority once again acknowledged concerns about the difficulties of checking details during the close of rolls period, which “would be especially pertinent in marginal electorates where small numbers of votes can affect the outcome.” But it also found that “there seems to be no persistent pattern of high close of roll enrolments in marginal seats.” High enrolment activities occurred in inner-city seats where they might have been expected. The majority noted the submission of one of the present writers, which supported the existing seven-day period “on the grounds that it essentially formalised what had been the practice before 1983.”58 This appears to have been the first time that previous practice had been recognised in the early-closing debate. The report acknowledged two major developments. The AEC conversion to an address-based enrolment system, and in 1999 habitation reviews had finally been abandoned as the main check 51

LIMITING DEMOCRACY

on roll integrity – though a limited number still took place. The AEC had also introduced continuous roll update, which had generated a third of enrolment forms in the previous financial year. The committee compared the process of ordinary enrolment activity with what happened in the rush period: While there are anomalies, changes or additions are not made to the rolls until these are resolved. This applies at all times including during the close of rolls. Applicants for enrolment are therefore not added to the roll during the close of rolls period until verification of eligibility is complete, with one exception. If verification is completed after the roll close, the elector is added to the notebook (“additions list”) and advised on election day to cast a provisional vote. Moreover, insofar as there are residual doubts with respect to enrolments at the close of rolls, the Committee believes that these are addressed by its recommendations relating to proof of identity and address for first time enrolments, re-enrolments, transfers and provisional voters. In the light of these, the existing checking processes for the close of rolls period, and the prospect of unnecessarily disenfranchising voters by foreshortening the close of rolls, the Committee recommends that the existing seven-day period between the issue of the writs and the rolls closing should be retained.59

The report was tabled in June 2003. When the Howard government responded in October it did not support that recommendation. The law required electors to update their enrolments, the government said, and an early close would “ensure that the AEC has sufficient time to verify details provided by applicants for enrolment, which will preserve the integrity of the roll.”60 Despite the government’s view, there were problems for advocates of either instant closing for new enrolments and a brief respite for changes to existing enrolments, or instant closing for 52

SLAMMING THE DOOR

both. First, there was the possibility that the next JSCEM might give the same dusty answer to another round of submissions arguing all was not well. This might help explain a change in the composition of the committee. The 2001 chairman, Petro Georgiou, a former Victorian state secretary of the Liberal Party, was replaced for 2004 by Tony Smith, a Victorian lawyer, and another Victorian, John Forrest, was replaced by a Queenslander, Steven Ciobo. Gone were the sole National Party member, Forrest, and the moderate chair, Georgiou. Second, the old argument that the AEC could not possibly run adequate checks in the rush period appeared to have been overtaken by the major improvements in the checks that it was running at all times. The supporters of the early roll closure responded to these problems in several ways. One was to step up the blameworthiness of potential victims, a standard political tactic of the time. Sixty per cent of transactions during the close of roll period should have been attended to earlier when the entitlement or change arose, they said. Allowing late enrolments encouraged voters to neglect their obligations. Time and again, taxpayers’ funds had been spent to try to persuade the defaulters to get their enrolments into order. They had acted unlawfully. And, the minister would add subsequently, they compared badly with the 65,000 seventeen-year-olds who had sought provisional enrolment, of whom 13,803 were thereby enabled to vote when they turned eighteen by polling day. A second response was to praise rather than vilify the potential victims of early closure. In this vein, the new chairman of the JSCEM wrote that it is “condescending and wrong” to suppose that potential electors would ignore a change in the legislation and miss out on the right to vote.61 The supporters of the early roll closure also argued that sticking with the existing system would “decrease the accuracy of the roll during non-election periods,” though the consequences of that problem were left unsaid; perhaps mail returned to MPs was the worst. The obverse proposition – that maximum accuracy during election periods 53

LIMITING DEMOCRACY

was important, and the period of grace played a critical role in that process – vanished without trace. The attempt by Senator Abetz and his supporters to vilify the potential victims of their proposal ignores the exigencies of modern life, as a recent revelation from within their own party shows. In the process of seeking preselection for the blue riband seat of Kooyong in 2006, Joshua Frydenberg, a former national security adviser to Prime Minister Howard, revealed that for the past ten years “he had failed to update his electoral enrolment several times as he moved between Canberra and Melbourne” to fulfil his employment duties.62 He had, however, been continuously enrolled (though perhaps not at the correct address) and voted at each intervening election. We were not told whether the announcement of those elections jogged his memory, but that would certainly be the case for tens of thousands of law-abiding citizens. A third tactic in this phase of the debate was to go back to the beginning, ignore the other side’s arguments, say time was needed to check applications (as the government had done in its response to the 2001 report), sternly affirm the old argument that there are villains out there, and shift the balance of proof: The Committee also agrees that the current close of roll arrangements present an opportunity for those who seek to manipulate the roll to do so at a time where little opportunity exists for the AEC to undertake the thorough checking required ensuring roll integrity. The Committee believes that those who argue for the retention of the seven day close of rolls and who promote the argument that there is no proof that enrolment fraud is sufficiently widespread to warrant any action, have missed the point. The fundamental issue facing this Committee is to prevent any such fraud before it is able to occur. Failure to do so would amount to neglect. While the risk exists that fraud sufficient to change the result of an election might occur, we are failing in our duty to protect and preserve the integrity of our electoral system and our democratic processes and principles.63 54

SLAMMING THE DOOR

Thus the Committee had two bob each way by backing both “any fraud” and “fraud sufficient to change the result of an election” in successive sentences, and then went on to break any link between “democratic processes” and voting rights by saying that the negligent forfeit their rights. In October 2005 the minister was still bolstering this option by claiming that the statutory provision “was introduced as part of Labor’s self-serving changes in 1984” (emphasis added) thereby ignoring the history and offering his own revisionist version: So what the Government is proposing is not unusual, and it is not unprecedented – indeed, it was the system for the first 83 years of our Federation.

A fourth tactic has been to claim popular support for the proposal, rather like the proposition that disenfranchising prisoners would be supported by the average person. Senator Abetz’s version of that was: The rolls close on the day the writs are issued in Tasmania and New South Wales (our two oldest states), and yet no-one has ever claimed that there is a democratic deficit in those States.64

This is misleading, to say the least. Senator Abetz justifies this statement by citing “relevant” sections of the two state acts to the effect that rolls close when the writs issue. In the case of New South Wales, legislation determines that elections now take place on “the fourth Saturday in March.” There is no democratic deficit because there cannot be a snap election. In Tasmania, the electoral Act contains what the JSCER originally proposed for the commonwealth Act, a requirement that the dissolution of the Assembly must be proclaimed not less than five days before the rolls close – or an equivalent period if the effluxion of time had ended its life. No snap election could close the roll immediately, which is why there are no claims of a democratic deficit 55

LIMITING DEMOCRACY

there either. If Senator Abetz, a Tasmanian lawyer and former president of the Tasmanian division of the Liberal Party, thinks a proclamation five or six days before close of rolls is preferable to what has been in the Commonwealth Electoral Act since 1983 then that would no doubt be acceptable to all parties to the argument. A fifth tactic is to go to the question of motive, and deny that the intention is to exclude voters who, on balance, might be expected to vote against the Coalition parties. One variation is to ask: who can say where an advantage might lie? For example, one of the present writers made the following point in a submission to the JSCEM inquiry on roll integrity: A reasonable estimate now is that blocking last minute enrolments by 18–22 year olds is worth about 150 votes to the Coalition on average in every commonwealth electoral district. The additional benefit from frequent movers cannot be estimated, but is I think real. In 1983, though the Coalition lost the election, the sudden close certainly saved the seats of Cook and probably Riverina and Bruce.65

A variation of this tactic is to say that the evidence has been changing or new evidence is available. The evidence Senator Abetz relied on when he argued this point comes from a paper presented at the Youth Electoral Study Workshop in Canberra, June 2005, which found that more “young people” (aged 18–30) voted Liberal rather than Labor, 41 per cent against 32 per cent. Rather like the senator’s selection from state statutes, there is a bit more to the data than that. Those figures are first preferences only, and as the paper’s author, Professor Clive Bean, says, the evidence shows considerable support by young electors for minor parties, especially the Greens, which would generally assist Labor. In mid 2006 opinion polls were reporting that on a twoparty-preferred basis 18–25-year-olds prefer Labor 64 per cent to the Coalition 36 per cent.66 56

SLAMMING THE DOOR

Demographic evidence reveals that the high propensity of 20–35-year-olds to move house can be matched with their propensity to vote differently from the over-50s.67 The likely advantage for the Coalition of an early close of roll is the only plausible explanation for that side of politics having seized upon the question and pursued it with such dogged tenacity for so long. While other items from the Chapman Society stable, like sub-divisions or numbered ballot papers, were quickly or belatedly put aside, instant close of rolls has been the polestar to which the Coalition has long steered – save in 2001, and that steersman (Petro Georgiou) was soon put over the side. PRACTICAL PROBLEMS Now that the 2006 legislation will enforce a prompt closure of the roll, what will become of the would-be electors and the changers caught out by it? Might not some, unaware of the legislative change, still get an enrolment card from a post office after an election is announced, fill it in and send it off, and then turn up at a polling place on election day thinking all was well? Their acknowledgement form would have told them that they would be on the commonwealth roll in due course, and were already on their state roll, but they were not entitled to vote at the forthcoming election. But if they had only moved within the division, they would still be entitled to two votes, House and Senate; if they had moved out of the division but still within the same state, they would be entitled to vote for the Senate. The computer that writes their acknowledgement card might address their particular case, but general advertising about this special problem is going to be complicated. Moreover, being feckless souls who had left the business so late anyway, are they likely to read an acknowledgement card that carefully? At best, there will be protracted arguments and explanations at the issuing table inside the polling place, and the need to document such cases; at worst, there will be uncertainty about how many eligible electors may have been mistakenly denied 57

LIMITING DEMOCRACY

access to the right ballot papers. A good many applicants for a ballot paper may end up insisting on time-consuming provisional ballots and these will have to be dealt with during the scrutiny. In close outcomes there could be challenges in the Court of Disputed Returns. The simple but important act of voting risks being turned into a shambles. There are also a couple of constitutional flies in the government’s ointment. One that has been neglected, probably because it immediately directs attention to the extraordinary events in 1983 – which the peddlers of the ointment would prefer are never mentioned – is the central role of the governor-general. When Prime Minister Malcolm Fraser first went to Yarralumla to seek the dissolution intended to keep Bob Hawke out of the Labor leadership and, incidentally, keep a number of young, eligible people off the roll, he was told to go away as the governor-general was unavailable. The governor-general, Sir Ninian Stephen, subsequently asked for a more substantial statement of Fraser’s reasons. It was only then that the prime minister secured his double dissolution, albeit still in time to close the rolls prematurely. If the snap close of rolls requires a snap dissolution, the prime minister will have to be first certain of his governorgeneral. Presumably there will not have been an ongoing political or constitutional crisis, for that would have alerted interested parties and enrolment campaigns would already have been initiated. The thunderbolt of a snap closing is best launched out of a clear blue sky. The argument for prompt closure relies on the existence of a large number of people standing ready to lodge shonky enrolment papers so that either they, or others, can vote fraudulently once an election has been announced. But might not that chain of events be set in motion by other means? This had always been a major flaw in the case for snap closing: that a huge underground army existed and could be mobilised, but only if the rocket had already gone up with an election date. Might not the generalissimo of this vast conspiracy merely hire someone to sit in a car at 58

SLAMMING THE DOOR

the gates of Yarralumla watching for the prime minister to visit the governor-general and texting “Red Alert!” at every sighting, at which point the holders of the applications would start moving to their local divisional office to await the next signal, “Lodge ’em!” It is a nonsense, but no more so than the belief that an underground army of political villains is out there waiting for their call. The second fly in the ointment is section 12 of the constitution. While the governor-general may dissolve the House whenever he or she thinks fit and issues the writs for the House and the four territory senators, it is the state governors who issue the writs for the Senate. With all state governors acting on the advice of Labor premiers at present, and perhaps still at the next federal election, the prime minister may only be able to use the snap close of roll device for a House-only election. Senate-only elections (such as occurred in 1964, 1967 and 1970), with their likely eruption of minor party senators, would then follow. Alternatively, one or more premiers might be inclined to advise their governors not to issue Senate writs for, say, seven days after the announcement of the election, thereby keeping open the rolls in those states, with unknown consequences. The possibility of later High Court challenges to the validity of the election, especially if some premiers went along with the prime minister and others did not, cannot be discounted. The bottom line is not that the end of the world as we have known it is nigh, but that the current and previous special ministers of state, presumably with the blessing of more senior ministers, appear to have led Australia into a potential constitutional minefield from which no one will emerge with credit, and have done so for a cause that is so flimsy that only misleading evidence can be deployed to support it.

59

CHAPTER 4

Real problems, and how to fix them

I

n addition to embracing representative democracy earlier and more enthusiastically than many western societies, Australia has been at the forefront of electoral system innovation. There are few respectable electoral practices that have not been experimented with at one time or another by either the commonwealth or one of the states and territories. Victoria pioneered the official printed ballot paper which became such a hit in north America and much of Europe from the 1880s.68 Australia adapted the preferential method (the alternative vote) to local conditions and is only one of three countries to employ a single transferable vote version of proportional representation for national elections. Despite the 2006 electoral legislation, Australia remains a world leader in democratic best practice. But that doesn’t mean there is no room for improvement, and this chapter looks at several real areas of concern. DONATIONS AND DISCLOSURES The 1983 renovations of the Commonwealth Electoral Act made major changes to the regulations governing election campaign expenditure and disclosure, and introduced a system of public funding to political parties for campaign purposes. The first reform was motivated by the outdated condition of the existing 60

REAL PROBLEMS

procedures and their lack of enforcement. These problems had been highlighted at the 1979 Tasmanian state election when breaches of similar expenditure laws led to the result in one electoral district being overturned and new elections ordered. Alarmed that the experience might be repeated at the 1980 federal election, the Fraser Coalition government effectively removed all limits to campaign expenditure – which created another set of problems which the 1983 amendments sought to address.69 The second reform, public funding of election campaigns, was introduced in New South Wales in 1981 after it had been operating in US presidential campaigns for just over a decade. The commonwealth followed in 1983, but implementation has not been uniform; Western Australia, South Australia, Tasmania and the Australian Capital Territory still remain outside the tent. Assessments of the effectiveness of the federal funding and disclosure regime vary widely. While the JSCEM’s report on the 2004 election confidently declared that “after 20 years, Australia’s funding and disclosure scheme is achieving its major goals,” others argue that the system is deeply flawed and that reform is “long overdue.”70 We would contend that public funding has failed to achieve its stated goals and that private funding of parties and candidates lacks an appropriate degree of transparency. While some limits on campaign expenditure may appear desirable, the fungibility of money means that the only “grand solution” worth pursuing is disclosure of donations.71 When it recommended the introduction of public funding in 1983, the JSCER claimed that it would: 1. remove the necessity or temptation to seek funds that may come with conditions imposed or implied; 2. help parties to meet the increasing cost of election campaigns; 3. help new parties or interest groups compete effectively in elections; 4. relieve parties from the “constant round of fund raising” so that they could concentrate on policy problems and solutions; and 61

LIMITING DEMOCRACY

5. ensure that no participant in the political process was “hindered in its appeal to electors nor influence in its subsequent actions by lack of access to adequate funds.”72

Regrettably these goals have not been achieved by public funding, or have brought with them unintended, negative consequences, or have come about for completely different reasons. This is clear when we examine the JSCER’s list of benefits one at a time. First, the system is not sufficiently transparent to determine which, if any, donations come with conditions attached, and the policy influence trail is usually too twisted to confirm any connection post facto. Second, the cost of campaigning has soared since 1983 and shows no sign of abating; public funding has fed rather than restrained the trend. Third, despite a few continuing newcomers (mainly Greens and independents), the vast majority of current federal parliamentarians represent the same parties as they did in 1983. This is hardly surprising given that funding is allocated according to the number of votes won, which means the major parties receive most of the money. One unforeseen problem has arisen recently: an unsuccessful candidate, having made only a modest campaign outlay, can win a large share of the vote and profit handsomely from public funding. In the most notorious example, one unsuccessful independent candidate for the Senate in 2004, Pauline Hanson, received nearly $200,000 in public funding yet appears to have spent frugally on her campaign. The fourth declared objective of public funding has failed miserably since we now live in the age of “continuous campaigning” (another mainly US import) and consequently continuous party fundraising. It is now common practice, regardless of which party is in government, for ministers to make themselves available – at social functions and even when out jogging – to those able to pay often thousands of dollars into party coffers to be used for election campaigning. The bait is the opportunity to lobby the minister in person. As the political scientists Peter van 62

REAL PROBLEMS

Onselen and Wayne Errington ask, “Is it legal? Technically. Is it ethical and proper? Absolutely not.”73 The meaning of the fifth of the JSCER’s justifications for public funding is opaque. What is clear, however, is that public funding has not discouraged the major parties in their pursuit of private money. In 2000–03 the Labor Party attracted $132 million in funding, of which only 18.8 per cent was public funding; the equivalent figures for the Liberal Party were $120 million and 17.5 per cent.74 While the official public funding scheme is non-partisan in the sense that it benefits all parties and candidates who poll the required level of votes (4 per cent), two “unofficial” schemes of assistance that have expanded recently clearly favour incumbents, whether they are governments or individual sitting members. Federal and state governments have the right, indeed the duty, to convey to the public information about government services and entitlements, but not to use taxpayer-funded advertisements to boost the electoral stocks of the political party that happens to be in government. The Australian National Audit Office developed guidelines for government advertising in 1998, but they have never been adopted. Instead, federal and some state government television advertisements now routinely carry an authorisation so as not to breach section 328 of the Commonwealth Electoral Act, which regulates “electoral advertisements.” Between 1996 and 2005 the Coalition government spent a record $693 million on advertising, including $95 million in the 2004 election year. But the former special minister of state, Senator Abetz, has insisted that the pre-election blitz “was a legitimate use of public money to advertise programs that people stood to gain benefits from.”75 It is also a duty of all senators and lower house members to communicate directly with their constituents, and for this they receive an annual postal and printing allowance. Since 1 July 2005 the printing allowance has been $125,000 per year and the postage allowance a maximum of $44,042 per year. In addition to staff and a fully maintained and equipped electorate office, these allowances 63

LIMITING DEMOCRACY

accord incumbents significant advantages over challengers. The advantages of incumbency are now so strong that Australia is approaching the situation in the United States, where “congressmen are more likely to die or be indicted than they are to lose a seat.”76 Setting aside the landslide election of 1996, nineteen incumbent lower house members were defeated at the 1993 federal election and 21 in 1998; but in 2001 only six were defeated and in 2004 ten. The system also has the capacity to advantage one party over others, especially since the finance department relaxed the guidelines to permit the postal allowance to be used to distribute campaign material for sitting members and other members and candidates. In Queensland, for example, the Coalition currently holds 21 of the state’s 28 federal divisions and seven of its twelve senators; Labor has six divisions and four senators; and there is one independent member of the lower house and one Democrat senator. With current printing and postal entitlements, the Coalition has available each year in Queensland $3.5 million dollars to Labor’s $1 million to supplement the party organisations’ increasingly important direct mail campaigning. Graeme Orr has convincingly argued that “the single greatest issue confronting elections in the developed world is the influence of private money.”77 While it remains difficult to trace the influence of donations on political parties and politicians, Australian election campaigns are now very expensive operations. At the 2004 federal election Labor and the Coalition each received $40 million in public funding together with very significant amounts from private donors. Surprisingly perhaps, Labor received $46.4 million to the Coalition’s $40.8 million. This discrepancy is the result of some large corporations’ splitting donations between both parties to appear impartial, and the willingness of mega–trade unions to donate generously to Labor because of their hostility to the Coalition government’s industrial relations policies. The largest individual donor in 2004 was Lord Michael Ashcroft (a British citizen) who gave $1 million to the 64

REAL PROBLEMS

Liberal Party, thereby drawing attention to the fact that Australia, unlike the United States, does not prohibit foreigners donating to political parties. Such prohibitions are, however, easily avoided by utilising local intermediaries to make the donation to a party. During its 2004 inquiry the JSCEM considered placing limits on the amounts individuals and groups are permitted to donate to political parties and “associated entities” for campaign purposes, limits such as operate in Canada. But for good reasons – caps don’t work – and bad – they allegedly infringe on freedom of political expression – the committee majority did not accept the advice. Other options to reduce the costs of election campaigns, including limiting political advertising on television, appear attractive but would probably only encourage parties to find other ways to spend their money. What is much more likely to work is a well-designed and rigorously administered regime of disclosure of private donations. Unfortunately, the Electoral and Referendum (Electoral Integrity and Other Measures) Act 2006 has gone in the opposite direction by raising the threshold at which donations have to be declared from $1500 to $10,000. Evidence from other countries, including the United States, shows that political donations may have the characteristics of the biblical mustard seed. One dollar given to a politician or a party may lead to a hundred dollars’ worth of government contracts and the ten or twenty dollars of profit that come from the contracts. Because the major parties are separate legal entities federally and in the states and territories, an individual or corporation may now donate a total of $90,000 to a party without an obligation for it to be disclosed and this sum is certainly large enough to raise concerns about secretive influence-peddling. Canada, by contrast, requires all donations in excess of C$200 to be declared. Senator Abetz has repeatedly said that there is no evidence to support suggestions that these levels of secret donation could corrupt the political system. But the dangers are evident in a recent scandal involving corporate funding for the election cam65

LIMITING DEMOCRACY

paigns of certain “pro-development” candidates at the 2004 Gold Coast municipal elections. In a scathing report on attempted purchasing of influence, the Queensland Crime and Misconduct Commission said, “It must be seriously questioned whether the integrity of any electoral process could withstand the barrage of secrecy, deceit and misinformation that this inquiry has found occurred during the Gold Coast City Council election of 2004.”78 Soon after the report was handed down the Australian Financial Review reported that many property developers on the Gold Coast “have vowed to stop donating to political campaigns.”79 The lack of effective regulation of political money is the most serious problem afflicting Australia’s electoral system. We can only hope that corporations, fearful of adverse publicity or even the possibility of prosecution, will become less enthusiastic about donating large sums to political campaigns. But we also need the tightest possible rules – far tighter than in the government’s 2006 legislation – to keep the federal electoral process as free as possible of the influence of money. PROBLEMS WITH PREFERENCES The Australian colonies initially adopted Britain’s first-past-thepost (plurality) system of voting, in which the candidate (or candidates) with the most votes won an electoral district’s seat (or seats). In many countries the emergence of disciplined political parties in the early twentieth century led to the adoption and modification of a device still used overseas to deal with cases where no candidate secures a majority (50 per cent + 1): the “second ballot” or “run-off election” held a week or two later. Run-off elections prevent a candidate of one political persuasion defeating several candidates of other persuasions who have collectively received more votes; at the second round of voting their support would go to the most popular of the original candidates. This system remains popular elsewhere and was adopted in former Soviet bloc countries when democratic elections were introduced recently. But in rural Australia a second trip to the 66

REAL PROBLEMS

polls so soon after the first would have been an unpopular imposition on the electors. Starting in Queensland in 1893, voters were given the option of recording on the same ballot paper the name(s) of the candidate(s) they would vote for next if their favourite candidate failed to win. The two candidates with the most primary or first preference votes would go immediately into a second and final count in which the “second or (any) subsequent preferences” of voters who had backed first-count losers were distributed between the two remaining candidates. The idea has some support in the United States currently as an instant run-off election, but most of the world calls the use of preferences “the alternative vote” and prefers to use “the second ballot.” In the twentieth century the early preferential model was abandoned for a new version that required voters to assign every candidate a sequential preference if their ballots were to be counted; in Australia that model became known as “preferential voting.” The early model is better called “contingent voting” to distinguish it from “optional preferential voting,” which is now used in New South Wales and Queensland state elections. In those two states voters need record only as many preferences as they wish, but candidates are excluded sequentially rather than in a single batch of all but the top two. The switch from contingent voting to preferential voting had a negligible effect on election outcomes. Between 1919, the House of Representatives’ first election with preferential voting, and 2004, the latest, out of 4000 contests at general elections there have been ten instances (0.25 per cent) when the last two candidates in the count had not been the two leading candidates at the first count. Twice an independent candidate won the seat (including the most recent case – Kennedy in 2004), and the other eight were won by Coalition candidates. Invariably the exception was caused by a candidate originally in third place getting into the final two-candidate count. Interest in the phenomenon heightened when a leading candidate 67

LIMITING DEMOCRACY

subsequently defeated on preferences was a party leader, Pauline Hanson of One Nation, in Blair in 1998. As for by-elections, there have been only two cases out of the 113 held since 1919, Wentworth in 1956 and Blaxland in 1996, where the change in system affected the outcome, and both occurred in unusual circumstances. The pattern is very similar at state elections with lower-house preferential voting: 26 cases spread over five states. The appearance of four of these at one election – when One Nation polled strongly in Queensland in 1998 – suggests that the intervention of a fourth major party can make a difference and maybe the Greens are on the verge of doing this at the expense of Labor, but for the time being a return to contingent voting would be pointless and no one is promoting it. Where preferential voting does seem to have mattered has been in the growing proportion of divisions where preferences needed to be distributed to find the winner and in the advantage conferred when they have been distributed. If the 1919–2004 figures for the House of Representatives are separated into three periods – 1919–54, which might be called “two-party,” 1955–74 “one influential minor party,” and 1975–2004 “multi-party” – the proportion of divisions in which preferences had to be distributed to find a winner has risen from 20.5 per cent to 26.5 per cent to 36.8 per cent. But the proportion of all divisions in which the outcome was thereby changed has varied little, at 6.6 per cent, 6.5 per cent and 5.2 per cent respectively. A look at who benefited most when the outcome was changed shows that invariably it was the Coalition, with 75.6 per cent in the two-party period, 95.3 per cent in the one influential minor party period (it was what that minor party, the Democratic Labor Party, or DLP, was about), and 63.5 per cent in the multi-party period. If the argument for preferential voting is that it helps candidates other than those chosen by the major parties to win seats, the evidence is that outcomes at a general election are rarely changed so that an independent or minor party candidate wins. The figures for the three periods are 2.6 per cent, 1.6 per cent and 3.5 per cent of the 68

REAL PROBLEMS

small number of changed outcomes, for a grand total of six members of the lower house since 1919. What of the possibility that preferential voting leads to the election of candidates who had little primary support and have merely assembled sufficient second, third or even more trifling preferences to win a seat? The possibility exists, but the occurrence is relatively rare. The case of the Country Party’s Arthur Hewson winning the seat of Macmillan in 1972 with only 16.6 per cent of the primary vote has entered electoral folklore but is atypical. Across the six electoral systems – federal and five states (excluding Tasmania because of its single transferable vote system) – between two-thirds and three-quarters of successful candidates had already polled 45 per cent or better on first preferences and needed only a bit of a shove to get across the line; 20–30 per cent had already scored 48 per cent to begin with. The proportion winning with a start below 30 per cent of first preferences never gets into double figures in any of the six jurisdictions. Nowadays more divisions require distributions of preferences to determine the winner because of the steep increase in the number of candidates on offer, which is coupled with the decline in the share of first preferences won by the Coalition and Labor combined. This has occurred mainly in the multi-party period 1975–2004. At the first election (1975) 53.3 per cent of divisions had only two or three candidates and 0.7 per cent – one solitary division – had seven or more; at the most recent election (2004) no division had only two or three candidates and 66.5 per cent had seven or more, the largest number being fourteen candidates in one division. Put another way, the most frequent number of candidates in 1975 was three and in 2004 it was seven. Such an increase in the number of candidates demands more of every elector’s decisionmaking capacity. The major parties are well known figures on the political landscape; they produce the most advertising and are the focus of media attention during an election campaign. Deciding whether to prefer Coalition or Labor is relatively easy, 69

LIMITING DEMOCRACY

but what of the other, typically now five, candidates? What do they stand for? Who are their leaders and spokespeople? Do they ever get on television or into the newspapers? In particular, where should they be ranked on a comprehensive scale? Even getting the numbers right has become more of a challenge for an increasing number of electors. Voters are entitled to surrender a ballot paper they know they have spoiled and get a new one, and they are reminded of their right to do this as much as possible. But there is both a temptation to put a mistake into the ballot box to escape embarrassment and a likelihood the error has not been spotted. On this basis, it must be asked why it is necessary to provide a complete set of preferences if the voter does not have them or does not wish to record them. It used to be said that some electors were unwilling to record any preference for a Communist candidate, and certainly in the past when the only candidates were Labor and Communist the informal vote shot up. Compulsory voting requires voters to attend a polling place or else obtain a postal ballot, which interferes with the electors’ physical freedom to some extent but does not demand they have an opinion or pretend to have one. But to compel voters to provide a full set of preferences interferes with freedom of opinion by insisting that if electors wish to vote (as the great majority do) and have their vote counted, they have to record preferences they may not actually believe in. During 1955–74, when the DLP was influential, the average proportion of House ballot papers ruled informal and excluded from the count was 2.5 per cent. In the multi-party period the figure was 3.6 per cent but that conceals two different sets, 1975–83 when the figure was only 2.2 per cent, much as it was in the DLP period, and 1984–2004 when the proportion swelled to 4.4 per cent. And that in turn conceals a potentially significant trend line that starts high (6.8 per cent in 1984) and steadily reduces (3.0 per cent by 1993), then steadily rises again (5.2 per cent in 2004). It should be remembered that with average divisional enrolments of around 80,000, a mere 1 per cent of 70

REAL PROBLEMS

informal votes equals 800 electors. An informal vote of 10 per cent (or more), as sometimes happens now, means that approximately 8000 electors in a division went to the polls and tried to vote but their ballot papers were not admitted into the count. More detailed analysis of informal vote patterns shows that it is highest in two types of division – those with a substantial proportion of residents whom the census reveals to be “not fluent in English” and those with larger numbers of candidates – so the political impact varies. Two states, New South Wales starting in 1981 and Queensland starting in 1992, have abolished the requirement that a full set of preferences must be recorded for a ballot paper to be valid at state elections and allow their electors to express as few preferences as they like. Initially relatively few voters availed themselves of the opportunity but, as knowledge of the new possibility spread, the proportion rose to approximately 50 per cent of those whose preferences had to be distributed to achieve a result. The evidence also suggests that the change has had the effect of keeping down the proportion of informal ballots in high “not fluent in English” and “candidate numbers” divisions. Over a dozen elections in the two states there is only one indisputable case (and a couple of possible cases) of a candidate who would have won the seat with a mandatory full set of preferences losing it because of the change. Once again any partisan fall-out has been minimal, suggesting that adopting the method for federal elections, which we favour, would not be as disruptive as some parties have feared. The JSCEM considered a number of possibilities at its 2004 election inquiry and recommended retaining compulsory preferential voting and emphasising its requirements in AEC election advertising in those states where optional preferential voting prevails at state elections. It also recommended an oblique attack on the problem of candidate numbers through a tighter party registration process, starting with a thorough purge of all currently registered parties that had failed to win federal parliamentary representation, which was incorporated into the 2006 legislation. 71

LIMITING DEMOCRACY

This attempt to reduce the number of candidates has a number of implications. Requiring past or current federal parliamentary representation would shut out parties like One Nation that have secured and still retain seats in a state house. Some countries only register parties that have nominated candidates in a substantial proportion or a majority of seats, or sometimes in all seats; by not taking up this option the recommendation suppresses parties with small followings but generous funds to pay deposits. Requiring currently registered parties to re-register and prove significant membership numbers may also launch a new sub-profession of party-member recruiters, analogous to the recruiters of initiative-petition signatures in the United States. There was a time within living memory when party agents scoured the Australian countryside signing up whole families for party membership, thereby producing healthy branch memberships, sound bank-order-based party funds and good commissions for themselves. It might be worth considering a further provision that, to retain their registrations, parliamentary parties should have a branch membership at mid-point between elections equal to, say, 10 per cent of their vote at the previous general election. It could only make for a healthier democracy – providing the branch stackers can be held at bay. IN PROPORTION The single transferable vote, or STV, is a form of proportional representation used at the national level in only Australia, Ireland and Malta. The earliest versions of proportional representation gave party managers rather than voters the power to decide which of its candidates would be elected – the “list” method. Such enthusiasm for proportional representation as there was in the Australian colonies was qualified by a wish to allow electors to say which of their party’s candidates they preferred and ensure that those candidates were elected. STV was created for that purpose. As a method of voting it starts on the same basis as preferential voting. Each voter has to record a set of preferences (for some or 72

REAL PROBLEMS

all of the candidates, as the legislation prescribes). The voter need not have regard for the order in which the parties may have set out their nominations. Under STV voters can assign their first preference to the third candidate on List B, their second to the last candidate on List C, their third to the first candidate on List A, and so on. A few countries have adopted a hybrid version under which a voter may record preferences but only within one party list; that is a version of the list system, not of STV. The essence of STV is the freedom it allows voters to move both between party lists and within any one list as well. Turning votes into seats begins by determining “the quota” required to win a seat. Again several choices are available but Australian electoral laws use “the Droop quota,” named after H. R. Droop, an English enthusiast. The number of valid votes is divided by the number of seats to be filled in a multi-member district plus one, and then one is added to that figure. For example, in New South Wales at the 2004 federal election 3,974,565 valid votes were recorded to fill six vacancies, and that number of votes divided by seven, with one vote added, produced a quota of 567,796. The Labor team scored 2.5460 quotas and eventually won three seats, the Coalition team scored 3.0883 quotas and eventually won three seats; no other team won a seat and the best performance was by the Greens with 0.5140 of a quota. Compare that with Victoria, where the quota was 428,085; the Labor team, with 2.5282 quotas, a whisker less than in New South Wales, won only two seats, and the Coalition team, with 3.0869 quotas – again a whisker less than in NSW – still won three. Although the Greens, with 0.6157 of a quota, performed better than in NSW, they again failed to win a seat. The Family First team, with a bare 0.1317 of a quota, won the sixth Victorian seat. Sometimes the “churning” of preferences after the first count matters, and occasionally it matters a great deal. A minor consequence of a relatively elaborate system of counting, one often complained about, is that authoritative counting of the votes must wait until all ballot papers that are eligible to be 73

LIMITING DEMOCRACY

counted have been received. A correct quota must be calculated from the start. Australia’s generous provisions for voting anywhere in the country or overseas ensure the delay can last a couple of weeks. The use of computers has greatly speeded up counting and has become possible because of the introduction of “ticket voting,” which is used by 95.8 per cent of voters, with only Tasmania and the Australian Capital Territory – with their greater experience of STV voting at local Assembly elections – down at about 80 per cent. Marking a single box for that purpose allows rapid data-entry for the ballot papers that have used it, leaving only 5 per cent of ballot papers showing idiosyncratic sets of preferences that have to be keyed in much more slowly, at least until number-reading machines are deemed sufficiently reliable to cope with below-the-line votes. The JSCEM’s 2004 inquiry recommended that voters be required to number all boxes above the line (assuming they chose to vote above the line), but that option could have two unexpected consequences: a delay in getting a final result and a significant increase in informal votes. STV was first introduced in Tasmania for lower-house elections in 1909 and has been used there ever since – which is why any mention of preferential voting for lower-house elections in Australia has to say “federal and five states,” or more recently “federal, five states and one territory (Northern Territory).” It also had a brief life in New South Wales for lower-house elections in the early 1920s but was soon abandoned for preferential voting. STV was adopted for the federal Senate in 1949, since when it has been adopted for upper-house elections in most states – New South Wales, South Australia, Western Australia and most recently Victoria, though the first STV election will not take place there until November 2006. As Queensland is unicameral and Tasmania has the reverse order of STV for the lower house and preferential for the upper, STV with minimal variations is now the characteristic electoral system for Australian upper houses. If the main justification for STV was that it empowered electors to pick and choose among the candidates their preferred party had nomi74

REAL PROBLEMS

nated for the Senate, that power has not been used. Most voters, somewhat fewer in Tasmania because of their experience at Assembly elections, were already accustomed to marking their ballot papers to conform to their preferred party’s “how-to-vote” card and they carried on doing so after STV was adopted. Where STV has shown that it can have a significant impact on election outcomes has been in the transmission of (a) the major parties’ “surplus” preferences, left over after the initial group of candidates that received them has made as many quotas as it is going to get and secured that many seats, and (b) preferences from the excluded candidates of groups that never made a quota. Since the start of the period of DLP influence half a century ago, with one exception every Senate election has returned at least two senators who were not either Coalition or Labor. (The exception was the double dissolution in 1974, when political allegiances were exceptionally polarised and the DLP expired.) About 90 per cent of places secured by the big two are won by achieving a quota on the first count or from the immediate flow-on of “surplus” preferences down the party ticket; “other” candidates do about as well as them on first preferences but not surpluses, and have to depend on the distribution of preferences from excluded candidates who were never near a quota. With large numbers of small and very small parties and a fair few independent candidates who do not stand as part of a group, the latter process is extremely complicated, which is why so many voters abdicate their free choice to their preferred party through the ticket system. Australia’s enthusiasm for the majoritarian doctrine embodied in preferential voting for the lower house has periodically run into difficulties because of the “other” house – the Senate. After the transition to the new, STV system of electing the Senate had been completed by the double dissolution election of 1951, for a brief period the Coalition controlled both chambers. Once the DLP had secured enough seats in the Senate to hold the balance of power there, there was potential for trouble, but the DLP’s 75

LIMITING DEMOCRACY

readiness to support the Coalition government postponed the crunch. The advent of a Labor government in 1972 showed the dangers of a hostile and powerful upper house – dangers that had been barely hinted at over 1949–51 when the houses were last controlled by different major parties. While the most spectacular eruption of that period – withholding supply from a government with a majority in the lower house – may not recur, the possibility of extensive frustration of a government’s legislative program remains open. Between the 1972 and 1974 elections the Senate rejected six bills twice and from 1974 to the 1975 election no fewer than 22 bills. From July 1981 when the Fraser Coalition government lost control of the Senate until July 2005 when the Howard Coalition government secured control of the upper house, the “balance of power” was held by a group of minor parties of which the Australian Democrats was usually the largest. While the Senate majority often made life difficult for both Labor and Coalition governments, supply bills, except for a brief scare in 1990, went unmolested. Whether such restraint would characterise a future Senate controlled by the opposition remains to be seen. Two schools of dissatisfaction with the operation of STV in the Senate (and to a lesser extent in the state upper houses) have emerged in recent years. One concerns the possibility of candidates being elected when they and their party have received a minuscule proportion of the total vote. The Droop quota to elect a senator is approximately 14.3 per cent of the total vote. But at a double dissolution election, when all Senate positions are contested, the quota drops to 7.7 per cent and small parties have a much greater chance. At the 1987 double dissolution election a controversial small party of the left, Nuclear Disarmament, succeeded in electing a senator in New South Wales who had started with only 1.5 per cent of the state’s vote, not much higher than the party’s 1.1 per cent national support. In 2004, as already noted, another controversial small party, this time of the right, Family First, succeeded in electing a senator in Victoria having 76

REAL PROBLEMS

started with only 1.9 per cent of that state’s vote and a national 1.8 per cent. Before all the results were finalised it looked as if the Howard government might have to depend on that senator for its Senate majority, the attendant publicity drove the point home, and eventually it happened in December 2005 when one Coalition senator broke ranks and crossed the floor. Experience with the DLP and later the Australian Democrats had conditioned electors to the possibility of substantial minor parties having considerable influence in the Senate and thus on public policy generally, but more recently several things appeared to be going wrong. Candidates were being elected whose primary support had been so small as to raise doubts about whether they should have been elected at all. The device to prevent that would be a threshold below which no candidate could win a seat; when such a measure has been proposed the suggested threshold has been around 5 per cent of primary votes.80 Another concern was that preferences were moving around in directions which it seemed did not match the “real intentions” of voters. As the JSCEM 2004 report said: It is often difficult for ordinary voters to understand how preferences will actually flow from their “1” vote to other parties. The complexity of the Group Voting Ticket does not assist voters in understanding where their vote will eventually rest. In reality, the effect of the Group Voting Ticket is that only very few above-theline electors who bother to inquire will have the faintest idea where their Senate preferences are going.81

Whereas, one might add, they used to know such things when they painstakingly copied their party’s how-to-vote card onto the Senate ballot paper – and often made mistakes that caused their vote to come to rest in the limbo for informal votes. There is one other concern about the way STV operates. Some small parties have clearly been formed with vote-catching names merely to snare unsuspecting voters. If these voters follow that 77

LIMITING DEMOCRACY

party’s how-to-vote card then their preferences are distributed to another party which – by accumulating 1 per cent, or less, here and there – is eventually made a serious contender by the final count. Every minor party candidate who wins a place in the Senate does so by a different route, but a revealing example is the way the Nuclear Disarmament Party’s Robert Wood in 1987 grew 48,238 first preferences into a final, seat-winning vote of 288,434, thereby overshooting the quota of 245,883 comfortably and out-performing the only remaining other candidate, with 200,634 votes, even more so. Early in the count small numbers moved his way. There was only one other candidate on the Nuclear Disarmament ticket; she scored 760 first preferences, and built up to only 877 before she was excluded. Of these 677 went to Wood and 200 elsewhere, indicating that some of her second preferences began to stray from the start. Surpluses from elected candidates did little: only 208 votes from Labor, 144 from the Coalition and 73 from the Democrats. The exclusion of 22 independent and minor party candidates supplied another 204, barely 1300 in total. Then things started to improve for Wood: stronger independents added 4249 as they went, then the last of the Greens 25,975, the independent at the head of the ballot paper 21,769 and the last of the Democrats 39,564. It might be thought that the last but one of those represented the “donkey vote,” moving to the next available recipient down the ballot paper. But in fact the candidate at the top of the paper had started with 30,117 first preferences and accumulated up to 43,200 when excluded; his votes were divided almost equally between Elaine Nile of Call to Australia (Fred Nile Group) – who was ahead of Wood on the ballot paper – with 20,173, and Wood with 21,769, indicating that there was a lot more than donkey voting at work. By this point Wood had about trebled his initial vote. Next to go was the last Coalition candidate, whose preferences favoured Nile 59,707 to 13,410, while the last exclusion, the remaining Labor candidate, went more strongly the other way – Wood 133,813, Nile 78

REAL PROBLEMS

4494. This was quite an achievement for someone who started with around a fifth of a quota. Much simpler is the route to the Senate taken by another group of people, those who manage to get into parliament without being elected at all. Although their numbers have been growing in the recent past, the phenomenon has its roots in the late nineteenth century. When the Australian colonies were moving towards federation and looking abroad for ideas about federal government, senators in the United States were still being chosen by the state legislatures rather than popular election. Democratic ideas had progressed sufficiently by the 1890s that the framers of the Australian constitution never seriously considered the American example. In fact, firsthand experience of undemocratic and obstructionist upper houses in the Australian colonies led them to adopt a common franchise for the House and the Senate. But US influences were evident in what became section 15 of the constitution, which governs “casual vacancies.” In the event that “the place of a Senator becomes vacant before the expiration of his term of service” he or she is replaced by a person (since 1977, a person of the same party) chosen by a joint sitting of the relevant state’s two houses of parliament (or, in Queensland, the Legislative Assembly). Recent research reveals that at 30 June 2005, fully 31 of the 76 serving senators, or 41 per cent, were originally section 15 appointments.82 In 1993 Michael Tate resigned from the Senate five days into a new term to accept a diplomatic post; his Labor replacement served almost six years in the Senate before facing the voters (successfully) in 1998.83 On the other side of the chamber, Santo Santoro was appointed to the Senate in October 2002 by the parliament of Queensland, having managed to lose his blue riband Liberal seat at the 2001 state election, and became minister for ageing in February 2006. Before a constitutional amendment in 1977, these “section 15 senators” were required to face the voters at the subsequent federal election, either for the Senate or the House; since then they serve 79

LIMITING DEMOCRACY

out the entirety of the retiring senator’s term. Given the general advantages of incumbency and that above-the-line voting has ensured STV operates as a de facto list system, party managers have effectively replaced voters in launching the parliamentary careers of many senators. DISENFRANCHISING PRISONERS There is no explicit right to vote embedded in the federal constitution, and on the few occasions when the High Court has been asked to find an implicit right the results have been disappointing to the plaintiffs. The necessary qualifications to get on the electoral roll and hence to obtain voting rights are entirely a matter for the parliament. Among the 1983 amendments to electoral law was a liberalisation of the restriction on prisoners’ rights to vote. The Commonwealth Electoral Act 1902 had denied the vote to any person “attainted of treason, or who has been convicted and is under sentence or subject to be sentenced for any offence punishable… by imprisonment for one year or longer.” The 1983 amendments excluded anyone “serving a sentence of 5 years or longer for an offence against the law of the Commonwealth or of a State or Territory” and anyone who “has been convicted of treason or treachery and has not been pardoned.” In the states the relevant legislation varies: Victoria, Queensland, Western Australia and the territories disqualify prisoners serving five years or more; Tasmania draws the line at a three year sentence and New South Wales at one year; and South Australia does not disqualify any prisoners at all. Since 1996 attitudes to prisoner voting rights have hardened and American influences are again discernible. When it comes to prisoners and voting, the United States has by far the harshest laws of any liberal democracy. In only two states are prisoners permitted to vote; the other 48 disenfranchise all prisoners convicted of a felony. Thirty-two do so even if the person is on parole; and thirteen deny convicted felons the right to vote for all time – unless 80

REAL PROBLEMS

they have been pardoned. Currently nearly six million Americans cannot vote because of felony convictions.84 These laws stem from the period of post–civil war reconstruction and have had a major partisan impact because of the racially skewed character of the US prison population. Thirteen per cent of black males have lost the right to vote. Afro-American voters favour the Democratic Party by a factor of at least eight to one and, as a consequence of the burgeoning prison population since 1970, felony disenfranchisement now helps determine many US elections – including the presidential. Recently a new question has been put on the political agenda: should these numerous disenfranchised individuals continue to be counted as residents of their states for the purpose of apportioning House seats after each census? At the Australian constitutional convention of 1891 the Victorian premier, James Munro, proposed that any person convicted of a felony should lose the right to vote in federal elections for all time. It would be “a disgrace to the federation” to permit them to vote, he contended.85 While Munro’s proposal was not accepted in 1891, a watered-down version did gain the approval of the parliament in 2004. In the session prior to the announcement of the 9 October 2004 federal election, the government and the opposition combined in the Senate to shrink Australia’s inclusive franchise. The Electoral and Referendum Amendment (Prisoner Voting and Other Measures) Act 2004 reduced the disqualifying length of a prison sentence from five to three years. How and why did this reduction of existing rights occur? In 1993 the JSCEM had recommended the removal of all limits on the right of prisoners to vote. This would have placed Australia among the small number of countries – Israel, South Africa, Sweden, Switzerland and Denmark – that allow all prisoners to vote. The Keating Labor government did not adopt the recommendation. In 1988, reflecting the altered political balance of its membership, the JSCEM recommended that all prisoners be denied the vote. Lacking a Senate majority, the Howard government failed in its attempts to legislate to that effect. 81

LIMITING DEMOCRACY

Following the JSCEM report on the 2001 federal election, the government introduced two pieces of legislation, one of which contained a proposal to strip all prisoners of the right to vote. When the Senate debated the Bill on 24 June 2004 the clause proposing to disenfranchise all prisoners seemed destined for defeat until Labor’s upper house leader, Senator John Faulkner, moved an amendment to remove from the electoral roll any person imprisoned for a full term of the House of Representatives. Given that the historical average for a House term is two years and seven months, that would have effectively doubled the existing electoral penalty on the imprisoned. In a near empty chamber, Faulkner’s amendment was supported by the government and passed without division. When pressed later in the debate by the Greens’ Bob Brown, Faulkner conceded, “I did not expect that the opposition’s amendment would be carried but it was…”86 The AEC then advised the government that Faulkner’s amendment was unworkable. This led to the 2004 Bill, which removed all reference to the term of the House of Representatives and instead prevented all people serving a full-time sentence of three years or more from enrolling or voting. Labor supported the Bill, which passed the Senate 30 votes to five – the five being Senators Bartlett and Murray (Democrats), Senator Harradine (independent) and Senators Brown and Nettle (Greens). The defence of the right to vote did not attract a single supporter from the major parties. In the absence of any other proffered explanation, it may be assumed that Labor supported the Bill so as not to be accused by the government of being “soft on crime” so close to an election. Indeed, in an earlier debate Labor’s Senator Robert Ray, a JSCER veteran, had warned Senator Brown that he would be “blackguarded on talkback radio” for supporting such “an emotive issue” as voting rights for prisoners.87 What is of concern is the weakness of the arguments put forward to justify this diminution of the right to vote. The communications minister, Senator Helen Coonan, asserted that “people who commit offences against society sufficient to 82

REAL PROBLEMS

warrant a full-time prison term for an electoral cycle or longer should not have any entitlement to vote and elect the leaders of this society.”88 Senator Abetz concurred: If you’re not fit to walk the streets, as deemed by the judicial system in this country, then chances are you’re not a fit and proper person to cast a vote in relation to the future of your country.89

Sloganeering of this sort is the stock-in-trade of right-wing commentators and bloggers in the United States. Writing in the Los Angeles Times on 8 March 2005 Jonah Goldberg asserted that “criminals tend to vote Democrat.” “Lawbreakers shouldn’t be lawmakers,” he went on, and “I oppose voting by ex-cons because voting should be harder, not easier – for everyone.” Back in Australia, in a letter to the Canberra Times, a Dr Peter Phelps of Queanbeyan castigated one of the current authors for wanting “killers, rapists, drug dealers, bashers and recidivist property offenders voting for our politicians…”90 Dr Phelps didn’t feel the need to tell readers that he was Senator Abetz’s chief of staff.91 There are, we contend, good civic reasons for giving prisoners the right to vote. First, in liberal democracies such as Australia offenders are imprisoned as punishment not for punishment – other than that which is incidental to their loss of liberty. A person in prison has the right to marry, for example, and to pursue civil law actions. So long as the potential for rehabilitation remains part of our penal system, and given that almost all prison inmates will secure release, it is bad public policy to alienate prisoners further from mainstream society by stripping them of the right to the franchise. Second, such proposals are inevitably discriminatory since the prison population, though rising, is not a mirror of society as a whole: most prisoners are males (94 per cent), and are aged between 20 and 35 (56 per cent). Notoriously, the imprisonment rate of Aborigines is fifteen times that of the non-Aboriginal population, and increasing. Third, denying prisoners the vote 83

LIMITING DEMOCRACY

runs counter to recent developments in comparable jurisdictions. The Canadian Supreme Court invalidated a similar law in 2002, and in March 2004 the European Court of Human Rights ruled that the section of the United Kingdom’s Representation of the People Act 1983 that denied all prisoners voting rights, was in breach of the European Convention on Human Rights. No such considerations influenced the majority of the members of the JSCEM, whose report on the 2004 election was tabled in September 2005. It endorsed the Liberal Party’s submission and recommended that “persons sentenced to a period of full-time imprisonment should not be allowed to [sic] a vote during that time…”92 The only supporting evidence for the recommendation was a quotation from Senator Abetz. Labor members of the committee submitted a minority report which did not mention the issue of prisoner disenfranchisement. Only the one Democrat member, Senator Murray, opposed the recommendation. The JSCEM’s recommendation was incorporated in the government’s 2006 legislation, with the variation that prisoners could remain on the electoral roll, and is now law. Before stripping an unpopular minority of a major civil right, Australia’s legislators might have heeded High Court Justice Michael Kirby’s argument: Prisoners are human beings. In most cases, they are also citizens of this country, “subjects of the Queen” and “electors” under the Constitution. They should, as far as the law can allow, ordinarily have the same rights as all other persons before this Court. They have lost their liberty whilst they are in prison. However, as far as I am concerned, they have not lost their human dignity or their right to equality before the law.93

Excluding all convicted prisoners from voting is out of sympathy with Australia’s hitherto inclusive, democratic culture. The law passed in 2006 will strip the vote from about 20,000 people in random and discriminatory ways. For example, a fine-defaulter in 84

LIMITING DEMOCRACY

prison will retain the vote. People sentenced to imprisonment and immediately released because they were imprisoned for an equivalent period while on remand will be able to vote – as they were when in prison on remand. An offender who has been convicted but not yet sentenced pending a sentencing report may vote. Offenders whose health or family responsibilities get them a suspended sentence or periodic detention may vote. The same offence may carry a mandatory detention in one state but not in another, and even among federal offenders, according to the Australian Law Reform Commission, “there is compelling evidence of inconsistency in the sentencing… across Australia.”94 Finally, the disenfranchisement happens only if you are so unlucky as to be in jail when an election occurs, once every two and a half years. Can a law containing so many anomalies really be a good law? Moreover, if moral turpitude is to be the target of community indignation (the “pub test”), what of those who had destroyed the lives of employees and shareholders and are disqualified from acting as company directors for periods that may equal the lives of several parliaments yet don’t have to go to prison? Australia, unlike Britain, Canada and New Zealand, lacks a national Charter of Rights which might permit citizens judicial redress – though there is likely to be a High Court challenge after the next election on the grounds that to deny a whole class of citizens, convicted prisoners, access to the franchise breaches the “directly chosen by the people” requirement of sections 7 and 24 of the constitution. THE ALSATIA OF CANDIDATE SELECTION In eighteenth century London the space behind the Savoy prison was called Alsatia. It became a notorious haunt for all types of rogues and thieves, so dangerous that no law enforcers dared to enter it. Political party preselections in Australia have traditionally been every bit as unexplored and potentially dangerous, which is hardly surprising given the importance of the decisions made by small groups of party members in the safest seats. 85

LIMITING DEMOCRACY

A great many electoral districts are so safe for one major party or the other that polling day is pretty much a formality, a popular anointing of whichever candidate secured the endorsement of that party. At recent federal elections upwards of a third of districts might be classed as “safe” because the winning candidate secured more than 60 per cent of the two-party-preferred vote. Realistically, the proportion of seats is usually larger than that. In those districts the winner is effectively chosen by a small number of people who may or may not be electors of that district and may, in some cases, not even be Australian citizens. Significantly, it is in the preselection process – in which the number of participants is much smaller than in an election itself and outcomes are correspondingly easier to influence – that the scant evidence we have of roll fraud has emerged. More evidence comes with the occasional disclosure of worrying transactions that are not expressly prohibited under the relevant laws. For example, a candidate (in a 1986 case, the broadcaster Alan Jones) might withdraw from a party preselection process after his employer (in that case, Kerry Packer) offers him a substantial raise in salary so that another protégé (John Hewson) can have a clear run (and, in Hewson’s case, go on to become the party leader).95 More commonly, an ageing incumbent and a rising challenger do a deal so that the former gets one more term in parliament and then goes quietly. Very occasionally the contest for a major party’s endorsement, usually in a safe district, becomes a much bigger news story than the subsequent election could ever be. A good example occurred in the Sydney division of Wentworth prior to the 2004 general election, when a couple of thousand Liberal Party members voted in the preselection instead of the usual number of, at most, a couple of hundred. Intense competition in this contest can cause extraordinary expansion in local membership as contestants recruit potential supporters for the dominant party’s ballot. “Branch-stacking,” especially when new members are recruited from particular ethnic groups in the electorate, can be both newsworthy and litigation-prone. 86

REAL PROBLEMS

Behind these controversies are some real concerns for the democratic process. When a decision that is eventually attributed to the 80,000 voters in an electorate is actually taken by a few dozen or at best a few hundred party activists – or worse, the conscripted foot soldiers of such activists – the appearance of democratic decision-making is hard to sustain. A recent and notorious example was the 2006 Labor Party preselection for the Victorian seat of Hotham, which was decided by fewer than 300 votes when the party’s 2004 vote in the division was 40,000. It was alleged that 30 per cent of the preselectors came from an ethnic minority which constituted less than 2 per cent of the division’s population.96 Then there is the degree to which party policy can be distorted when the few who are strongly motivated about political issues usurp the decision-making function of the many who have rather different interests or degrees of interest. Left stalwarts will choose candidates from the left end of their party’s opinion spectrum, right stalwarts the reverse. The result can be a legislature much more polarised and dogmatic than the electorate it is supposed to represent, causing dissatisfaction with representative institutions among the bulk of the population in the middle. In the late nineteenth century some electoral jurisdictions of the United States introduced what became known as “primaries,” in which a much wider section of the electorate, those who had registered as partisans of one party or the other for this purpose, selected its candidates – and occasionally, in “cross-over primaries,” the other party’s candidates too – for the subsequent general election. Sometimes a right to “write in” additional candidates who had not come through the parties’ preliminary vetting was thrown in as well. The intention of primaries was to protect the process from shadowy figures who might promote candidates willing to pass helpful legislation and prevent unhelpful measures. But primary elections have succumbed to the same ills as regular elections, relying on mass media coverage, massive expenditure and large donors. There have been academic advo87

LIMITING DEMOCRACY

cates of primary elections in Australia, but never much enthusiasm where it might have counted. The introduction of laws governing party registration has opened the door for laying down the procedures a party must use to choose the candidate(s) who display its label on their ballot papers. For example, we could require that anyone participating in the preselection process must meet the age, citizenship and residence requirements of electors. But this would not be straightforward, especially in cases where a mix of locals and central party nominees make the choice or where affiliated bodies like trade unions or primary producer groups have an historic right to participate in candidate selection. It would also be possible to require that the AEC (or a state commission with an equivalent power) conduct preselection ballots for a party as it currently does for trade unions. But neither a commission (with its resources already stretched and the risk of becoming involved in intra-party disputes) nor a party (jealous of its autonomy) is likely to be enthusiastic about the idea. Given all these problems, a more straightforward option would be to ensure that the penal provisions against inducements or threats at ordinary elections – which already apply in trade union elections – also apply to preselection ballots. It is difficult to see why unions and political parties should operate under different rules, especially when the parties benefit from public funding. And there is an Australian precedent: Queensland’s Electoral and Other Acts Amendment Act 2002 introduced elements of such a regime. Registered parties must have a “complying constitution” that disqualifies anyone convicted of specified electoral offences from party membership for a period of ten years, and ensures its preselection ballots satisfy “the general principles of free and democratic elections.” The state’s electoral commission has a supervisory role over party preselection ballots, including auditing ballot papers.

• 88

REAL PROBLEMS

Even the most democratic of electoral systems needs fine-tuning to take account of changing political circumstances and advances in democratic theory and electoral technology. We have drawn attention to a series of real problems, and we encourage the government and parliament to deal with them. Disappointingly, the trend of legislative change during 2006 ran in the opposite direction, producing a less democratic and more partisan set of electoral procedures.

89

Conclusion

I

n this book we have sought to report on the condition of the contemporary Australian electoral system. Generally we have found it to be in reasonably good shape – especially when contrasted with the decrepit state of the US system. But we have drawn attention to the shortcomings of the Coalition’s legislative changes in mid 2006, and the poor quality of the arguments adduced to defend the legislation. Those changes are disappointing enough, but the activists within the Coalition appear to have more in mind. Some – notably Senators Abetz and Minchin – are motivated by a long-term aim of abolishing compulsory voting, but shifts in the socioeconomic makeup of Coalition support in recent years mean that the party leadership is unlikely to go that far. Nevertheless, the activists will keep pushing for measures that break the longstanding link between adult citizenship and the right to vote, and measures that reduce the independence of the AEC. Already, in its 2004 report, the JSCEM has proposed that the Act should allow communication between the AEC and the relevant minister “during the election period” (our emphasis).97 At the time of writing, the government had not yet responded formally to the committee’s recommendations. Much of what we describe as “hammering behind the scenes” will become more obvious once they do so. The Howard government’s standard response to critics of its electoral law changes is to claim that they are “politically moti90

CONCLUSION

vated and ill informed.”98 Having made the accusation that the 2006 legislation was motivated by the desire for partisan advantage, we may anticipate being accused of having the same motivation and, if the current state of political debate in this country is any guide, probably more. The reader will note that neither of the major parties would agree with all the points made in this book; for example, we would not expect the Labor Party to embrace fully transparent disclosure of campaign donations and its record on prisoner enfranchisement is spotty. Moreover the electoral integrity legislation was opposed in the parliament not only by Labor but also by the three independent members, one of whom, Peter Andren, moved extensive, positive amendments which were not supported by either the Coalition or Labor.99 In the other chamber, not a single non-government senator voted for the legislation. We stand by our claim that the purported rationale for the changes to the electoral laws “is so flimsy that only misleading evidence can be deployed to support it.” As democrats, we would be happy to let the people decide.

91

Endnotes 1. Australian, 19 January 2006. 2. Journalist Greg Sheridan’s assertion that “the US is the oldest continuous democracy” (Australian, 22 January 2005) is notably ill-informed. Certainly before the passage of the Voter Rights Act 1965 the United States would be better called an incomplete democracy. 3. Jimmy Carter and James A. Baker III, Building Confidence in U.S. Elections: Report of the Commission on Federal Election Reform, Center for Democracy and Election Management, American University, September 2005, p ii. 4. Baltimore Sun, 18 March 2006. 5. Michael L. Goldstein, Guide to the 2004 Presidential Election, CQ Press, 2003, pp 108–09. 6. Preserving Democracy: What Went Wrong in Ohio, Status Report of the House Judiciary Committee Staff, 2005, pp 43–47. 7. See Richard L. Engstrom, “Revising Constituency Boundaries in the United States and Australia: It Couldn’t Be More Different,” Democratic Audit of Australia, Aug 2005, . (Only five of the 50 states have bipartisan, but not non-partisan, redistricting panels, Washington Post, 29 October 2005.) 8. Robert F. Kennedy Jr, “Was the 2004 Election Stolen?,” Rolling Stone, 1 June 2006; Steve Freeman and Josh Mitteldorf, “A Corrupted Election,” , 15 February 2005. 9. Peter Brent, “Preferential Treatment,” Australian Business Monthly, July 1994, pp 86–89. 10. Australian Parliamentary Debates, House of Representatives, 13 March 1973, p 478. 11. Race Mathews, “Lying and Public Trust,” Public Sector Ethics Seminar, Monash University, 22 September 1995, pp 112–13. 12. Ammianus Marcellinus, The Later Roman Empire, Penguin, 1986, p 231. 13. David M. Farrell and Ian McAllister, The Australian Electoral System: Origins, Variations and Consequences, UNSW Press, 2006. 14. Graeme Orr, Australian Electoral Systems: How Well Do They Serve Political Equality?, Democratic Audit of Australia, 2004. 92

ENDNOTES

15. Thomas Carl Lundberg, “Second-Class Representatives? MixedMember Proportional Representation in Britain,” Parliamentary Affairs, 59:1, January 2006, pp 60–78. 16. Australian Parliamentary Debates, House of Representatives, 9 November 1983, p 2536. 17. Ibid, p 2539. 18. Norm Kelly, “Determining Political Parties: A Real Status Symbol,” Democratic Audit of Australia, December 2004. 19. Derek Chong et al, “It’s an Evil Thing to Oblige People to Vote,” Policy, Summer 2005–06, p 16. 20. Sydney Morning Herald, 29 March 2005; Lisa Hill, “Compulsory Voting: Residual Problems and Potential Solutions,” Australian Journal of Political Science, 37:3, 2002, pp 437–38. 21. “Compulsory Voting: Victorian Division’s Report, 21 September 1956,” Liberal Party of Australia, Victorian Division Records, Melbourne University Archives, Box 72/32. 22. Current Issues Brief No 1, 12 July 2004, Parliamentary Library, Parliament of Australia. 23. JSCEM, Report on the Conduct of the 1987 Election, 1989, p 111. 24. Joan Rydon, “The Federal Elections of 1987 and Their Absurdities,” Australian Quarterly, Spring & Summer 1987, pp 357–65. 25. Charles Copeman and Amy McGrath (eds), Corrupt Elections: Recent Australian Examples and Experiences of Ballot Rigging, Towerhouse Publications, 1997; Amy McGrath, The Frauding of Votes, Tower Books, nd. 26. Copeman and McGrath, op cit. 27. Courier-Mail, 3 November 2000. 28. Australian Financial Review, 22 June 2001. 29. Senator Eric Abetz, “Electoral Reform: Making Our Democracy Fairer for All,” Sydney Institute, 4 October 2005. The link with Beazley has been carried on by Senator Abetz’s successor as special minister of state: see Australian Parliamentary Debates, House of Representatives, 10 May 2006, p 88. 30. Graham Richardson, Whatever It Takes, Bantam, 1994, p 146. 31. Phillip Lynch, Australian Parliamentary Debates, House of Representatives, 13 March 1973, p 473. 32. Helen Berrill, “No Minister!,” Quadrant, 29:7, p 23. 33. C. Griffith, “Labor Squads Hit Rural Poll Booths,” Courier-Mail, 4 November 2000. 93

LIMITING DEMOCRACY

34. Australian Electoral Commission, Report on the 2004 Full Habitation Review of the Division of Isaacs, . 35. Ibid, p 4. 36. Ibid, p 18. 37. Ian Campbell (Australian Electoral Commissioner), evidence to the Senate Finance and Public Administration Legislation Committee, Canberra, 7 March 2006, p 3. 38. See, for example, Paul Kleppner, The Third Electoral System, 1853–1892, University of North Carolina Press, 1979. 39. Copeman and McGrath, op cit, p 30. 40. Editorial, Sydney Morning Herald, 12 July 2000. 41. Copeman and McGrath, op cit, p 15. 42. D. P. Moynihan, Secrecy: The American Experience, Yale University Press, 1998, p 132. 43. The Shepherdson Inquiry: An Investigation into Electoral Fraud, Queensland Government Printer, 2001, p xiv. 44. Because the US elects multiple officials on the same day and on the same ballot paper, “voting machines” (many now electronic) have long been necessary. By contrast the most complicated task a typical Australian federal voter faces is to rank a number of House candidates, place a single mark above the line on a separate Senate ballot paper and very occasionally indicate YES or NO on a referendum paper. A problem raised recently is that sight-impaired voters must convey their choices to another person to be able to cast a valid vote. Computer touchscreens could protect the secrecy of their votes, but would be expensive to install at all polling places. 45. Australian Financial Review, 3 February 2006. 46. Australian National Audit Office, The Integrity of the Electoral Roll: Australian Electoral Commission, Canberra, 2002, p 11. 47. Carter and Baker, op cit, p 80. 48. Joint Select Committee on Electoral Reform (JSCER), First Report, 1983, pp 107, 112. 49. Joint Select Committee on Electoral Matters (JSCEM), Report on the Conduct of the 1993 Election, 1994, p 157. 50. Sydney Morning Herald, 5 October 2005. 51. R v Pearson; Ex parte Sipka (1983), 152 CLR 254. 52. Colin A. Hughes, “An Election about Perceptions,” in Howard Penniman (ed), Australia at the Polls: The National Elections of 1980 and 1983, 94

ENDNOTES

George Allen & Unwin, 1983, p 284. 53. JSCER, First Report, 1983, p 110. 54. JSCEM, The 1987 Federal Election: Inquiry into the Conduct of the 1987 Federal Election and 1988 Referendums, 1989, p 7. 55. JSCEM, Report on the Conduct of the 1993 Federal Election, 1994, pp 34–35. 56. JSCEM, Report on the Conduct of the 1996 Federal Election, 1997, pp 14, 122–23. 57. JSCEM, Report on the Conduct of the 1998 Federal Election, 1999, p 19. 58. JSCEM, Report on the Conduct of the 2001 Federal Election, 2002, p 62. 59. Ibid, pp 62–63. The “anomalies” are explained in the report’s note 141. 60. Gary Nairn, “ALP’s Electoral Reform Claims Have to Be Taken with a Grain of Salt,” Canberra Times, 23 February 2006, p 15. 61. Tony Smith, “Voting Changes Will Strengthen Our Democracy,” Age, 9 November 2006. 62. Australian, 24 March 2006. 63. JSCEM, Report on the Conduct of the 2004 Federal Election, 2005, pp 35–36. 64. Abetz, op cit. 65. Colin A. Hughes, Submission No 49 to JSCEM, The Integrity of the Electoral Roll: Review of ANAO Report No 42, 2001–02, Canberra, 2002. 66. “Federal Poll: Shorten Raises Little Interest,” ACNielsen AgePoll, Age, 22 May 2006. 67. Graeme Hugo, “Nomads? On the Move in Australia,” Dialogue 24, 2005. 68. Peter Brent, “The Australian Ballot: Not the Secret Ballot,” Australian Journal of Political Science, 41:1, March 2006, pp 39–50. 69. Colin A. Hughes, “A Close-Run Thing,” in Penniman, op cit, pp 236–38. 70. Bryan Mercurio and George Williams, “Australian Electoral Law: ‘Free and Fair’?,” Federal Law Review, 32, 2004, p 31. 71. Colin A. Hughes, “Election Finance Controls: Is There an End Game?” in Karl-Heinz Nassmacher (ed), Foundations for Democracy: Approaches to Comparative Political Finance, Nomos Verlagsgellschaft, 2001, pp 206–21. 72. JSCER, First Report, 1983, pp 153–55. 73. Peter van Onselen and Wayne Errington, “Political Donations and Party Fund Raising: Buying Time with a Minister,” AQ, 76:6, Nov–Dec 2004, p 6. 95

LIMITING DEMOCRACY

74. Dean Jaensch et al, Australian Political Parties in the Spotlight, Democratic Audit of Australia, 2004, p 32. 75. Age, 24 March 2005. 76. Jeffrey Toobin, “Drawing the Line,” New Yorker, 6 March 2006, p 35. 77. Orr, op cit, p vii. 78. Independence, Influence and Integrity in Local Government: A CMC Inquiry into the 2004 Gold Coast City Council Election, Crime and Misconduct Commission, Brisbane, 2006. 79. Australian Financial Review, 13–14 May 2006. 80. See, for example, Helen Coonan, “Safeguard or Handbrake on Democracy?” in Brian Costar (ed), Deadlock or Democracy? The Future of the Senate, UNSW Press, 2000, pp 12–28. 81. JSCEM, Report on the Conduct of the 2004 Election, 2005, p 213; George Brandis, “Dealing Underhand,” Courier-Mail, 17 October 2005. 82. John Nethercote, “Senate Vacancies: Casual or Contrived?,” in Upholding the Australian Constitution, Samuel Griffith Society, 2005, p 109. 83. Ibid, p 110. 84. Canberra Times, 26 October 2005. 85. Official Record of the Debates of the Australasian Federal Convention, Sydney, 1891, p 657. 86. Australian Parliamentary Debates, Senate, 24 June 2004, p 2460. 87. Ibid, 15 February 1999, p 1813. 88. Ibid, 12 August 2004, p 2615. 89. Sydney Morning Herald, 15 June 2004. 90. Canberra Times, 15 July 2004. 91. Media Watch, ABC Television, 9 August 2004. 92. JSCEM, Report on the Conduct of the 2004 Federal Election, 2005, p 129. 93. Muir v The Queen (2004), HCA 21 (2 April 2004), p 7. 94. Same Crime, Same Time: Sentencing of Federal Offenders, Australian Law Reform Commission, 2006, p 14. 95. Tom Burton, “Alan Jones: Why I Quit the Liberal Race,” Sydney Morning Herald, 30 May 1990. 96. Australian, 24 February 2006. 97. JSCEM, Report on the Conduct of the 2004 Election, 2005, pp 76–77. 98. Nairn, op cit. 99. Australian Parliamentary Debates, HR, 10 May 2006, pp 89ff. 96

BRIEFINGS

BRIEFINGS LIMITING DEMOCRACY

“It is not unusual for those whose party loses elections to say that there are flaws in the electoral system. What has been exceptional in Australia since 1996 is that supporters of the party that has won a series of elections complain they were robbed.” Australia’s electoral system is distinctive, independent and widely respected. For years, parties across the political spectrum have supported its main features – compulsory voting, an independent electoral commission, uniform rules across the nation. Yet, in 2006, all that began to change.

In this book two distinguished political scientists, Colin A. Hughes – Australia’s first electoral commissioner – and Brian Costar, reveal the dangers of the government’s legislation – and of other changes flagged by government MPs. They trace the history of Australia’s admired electoral system, revealing the carefully crafted rules that have ensured a fair and transparent system – a system now endangered by a set of poorly considered changes.

Cover: At Hill End, north of Bathurst, Greg McNamara wheels Sam George and ballot boxes to the polling booth for the 1998 federal election.

UNSW PRESS ISBN 0-86840-948-0

BRIEFINGS A series of short, topical books exploring social, political and cultural issues in contemporary Australia, published in association with Australian Policy Online, www.apo.org.au

9 780868 409481

COLIN A. HUGHES & BRIAN COSTAR

Because Australians have no explicit constitutional right to vote, parliament has wide discretion over how its members are elected. Having won control of both houses of parliament, the Howard government introduced legislation that erodes the right to vote and relaxes controls over political donations, all in the name of “electoral integrity.” Forced through parliament in mid-2006, these measures are dangerous for democracy.

UNSW PRESS

Limiting Democracy

The Erosion of Electoral Rights in Australia Colin A. Hughes & Brian Costar